Charcot-Marie-Tooth Disease, Demyelinating, Type 1a

A number sign (#) is used with this entry because Charcot-Marie-Tooth disease type 1A is caused by duplication of, or mutation in, the gene encoding peripheral myelin protein-22 (PMP22; 601097).

Deletion of the PMP22 gene characteristically results in hereditary neuropathy with liability to pressure palsies (HNPP; 162500). Point mutations have also been described in the PMP22 gene in patients thought to have hypertrophic neuropathy of Dejerine-Sottas (145900).

Description

For a general phenotypic description and a discussion of genetic heterogeneity of Charcot-Marie-Tooth disease type 1, see CMT1B (118200).

CMT1A is the most common form of CMT. The average age of onset of clinical symptoms is 12.2 +/- 7.3 years. Slow nerve conduction velocity (NCV) less than 38 m/s is highly diagnostic and is a 100% penetrant phenotype independent of age (Lupski et al. (1991, 1992)).

Clinical Features

Bird et al. (1983) and Dyck et al. (1983) reported families of typical CMT1 except that linkage to the Duffy blood group locus (Fy) on chromosome 1, where CMT1B maps, was excluded. Whereas Dyck et al. (1983) could discern no phenotypic differences between the linked and unlinked forms, Bird et al. (1983) suggested that slowing of nerve conduction is less marked and onion bulb formation on sural nerve biopsy less conspicuous in the Duffy-unlinked form.

Berciano et al. (1994) observed that clinically normal adult CMT1A patients are rare, but do exist. They referred to 1 duplication-positive woman who had normal neurologic examinations at least up to the age of 31 even though her motor nerve conduction velocities were 30 meters per second in the median nerve. This patient had a clinically affected 4-year-old son. Berciano et al. (1994) stressed the importance of doing not only neurologic examinations but also electrophysiologic studies or DNA studies to exclude the diagnosis of CMT1A. Hoogendijk et al. (1994) reviewed the clinical and neurographic features of 44 affected individuals, aged 8 to 68 years (mean 34 years), from 6 families with chromosome 17p duplication. Motor nerve conduction velocity and, to a lesser extent, compound muscle action potential amplitude were inversely related to clinical severity. Neither clinical severity nor NCV was significantly related to age. They interpreted the findings as suggesting that the primary pathologic process is not active, or only slightly active, after childhood. Garcia et al. (1995) found remarkable concordance of nerve conduction velocities in each of 2 pairs of male homozygotic twins with a type 1A duplication. There was also congruity between the left and right side of each twin as well as between twin brothers. However, there was marked dissimilarity in the clinical severity in each of the twin pairs, as well as asymmetric clinical involvement of each affected individual. Palpable nerve enlargement was greater in the less affected twins than in their more severely affected brothers. The marked discrepancy between nerve conduction velocities and clinical weakness suggested that other factors must be responsible.

Lupski et al. (1993) studied 2 unrelated patients with both CMT1 and neurofibromatosis type I (NF1; 162200). Since both of these disorders map to the pericentric region of chromosome 17, they investigated whether this might be a contiguous gene syndrome. In both patients, however, the CMT1A was inherited from the father, who did not have NF1. Furthermore, molecular analysis showed that the CMT1A duplication was stable in the 2 patients. One patient transmitted both disorders to her daughter. Thus, this was a chance concurrence of 2 common disorders. Bosch et al. (1981) had also reported the concurrence of these 2 conditions.

Pyramidal dysfunction due to compression of the cervical spinal cord by hypertrophied nerve roots, resembling radicular neurofibromas, had been reported in several individuals with type 1 Charcot-Marie-Tooth disease by Rosen et al. (1989). Butefisch et al. (1999) reported an individual with compression of the cervical spine and the cauda equina, similar to the cases described by Rosen et al. (1989). By demonstrating a duplication of the PMP22 gene, they confirmed that this individual had CMT1A.

Liehr et al. (1996) identified mosaicism for the CMT1A duplication by 3 different and independent techniques. Mosaicism was supported by the clinical features of the patient. The 25-year-old woman reported painful sensations in the shoulders, which increased after exercise. She had markedly reduced motor and sensory nerve conduction velocities and showed bilateral pes cavus. She showed a mild distally pronounced muscular weakness of the arms and legs. Muscle stretch reflexes were absent. Sensory disturbances of the limbs were located distal to the elbow and knees. Vibration sense was reduced at the malleolus internus. Sural nerve biopsy showed a marked reduction of myelinated fibers with signs of demyelination and onion bulb formation. Four different tissues were investigated successfully, yielding different patterns of mosaicism.

Dematteis et al. (2001) diagnosed sleep apnea (107650) and CMT1A in 1 family member and prospectively investigated 13 further members not previously suspected of having neuropathy or sleep apnea. Eleven of the 14 family members had the autosomal dominant demyelinating form of CMT with PMP22 gene duplication. In addition, all 11 individuals had sleep apnea syndrome with a mean apnea-hypopnea index of 46.6 per hour (28.5) of sleep (normal value less than 15 per hour). The remaining 3 family members were free from neuropathy and sleep apnea syndrome. Sleep apnea and neuropathy severity were highly correlated; the compound muscle action potential amplitude of the median nerve was inversely correlated with the apnea-hypopnea index. The severity of neuropathy and of sleep apnea was higher in male CMT individuals and correlated with age and body mass index. No wake or sleep diaphragmatic dysfunction was shown. Dematteis et al. (2001) concluded that sleep apnea syndrome is related to a pharyngeal neuropathy.

Patients with CMT disease are particularly susceptible to vincristine neurotoxicity (Weiden and Wright, 1972; Griffiths et al., 1985). Naumann et al. (2001) reported a 31-year-old woman with recurrent Hodgkin lymphoma (236000) and unrecognized HMSN I who developed severe motor neuropathy 3 weeks after the first cycle of treatment including 2 mg of vincristine. After the fact, the patient was found to have bilateral pes cavus deformity since early childhood, contractions of ankle joints, and shortened Achilles tendons. Her brother and mother had areflexia and moderate foot deformity. The diagnosis of HMSN IA was confirmed by molecular analysis.

Swan et al. (2007) found no differences in disability, as measured by a CMT neuropathy score, between 44 previously pregnant women with CMT1A compared to both 47 affected men or 15 affected women who had never been pregnant. Statistical analysis revealed no difference in severity between men and woman overall. Approximately 50% of women who had been pregnant noted a worsening of symptoms during pregnancy, mainly weakness, changes in balance, and alterations in sensation. Nine of these women reported a permanent worsening, and 13 felt it was temporary, lasting about 2.5 months after giving birth. However, symptom scores between these 2 groups were not significant. Swan et al. (2007) concluded that men and women are equally disabled by CMT1A and that neither gender, pregnancy, nor plasma progesterone levels significantly contribute to the severity of neuropathy in women with CMT1A.

Diagnosis

Matise et al. (1994) referred to the tandem duplication underlying CMT1A as resulting in segmental trisomy. The search for the CMT1A disease gene was misdirected and impeded because some chromosome 17 genetic markers that are linked to CMT1A lie within the duplication. Matise et al. (1994) demonstrated that the undetected presence of a duplication distorts transmission ratios, hampers fine localization of the disease gene, and increases false evidence of linkage heterogeneity. They devised a likelihood-based method for detecting the presence of a tandemly duplicated marker when one is suspected.

Schiavon et al. (1994) devised a rapid, informative, economical, and easily interpretable nonradioactive test for detection of the CMT1A duplication based on a microsatellite polymorphism. They found the CMT1A duplication in 76% of 56 unrelated patients.

King et al. (1998) described a patient with CMT1A caused by duplication of the PMP22 gene through an unusual mechanism: unbalanced translocation of 17p to the X chromosome. This finding further supported the hypothesis of gene dosage as the basis of CMT1A. The case highlighted the importance of fluorescence in situ hybridization as an alternative molecular technique in the diagnosis of CMT1A. The duplication would not have been detected by standard commercial methods based on identification of a novel junction fragment by pulsed field gel electrophoresis.

Aarskog and Vedeler (2000) described a quantitative PCR method for detecting both duplication and deletion of the PMP22 gene in CMT1A and HNPP, respectively. Their method of real-time quantitative PCR is a sensitive, specific, and reproducible method allowing 13 patients to be diagnosed in 2 hours. It involves no radioisotopes and requires no post-PCR handling.

Saporta et al. (2011) were able to find a molecular basis for 527 (67%) of 787 patients with a clinical diagnosis of CMT. The most common CMT subtypes were CMT1A in 55%, CMT1X (302800) in 15.2%, HNPP (162500) in 9.1%, CMT1B (118200) in 8.5%, and CMT2A2 (609260) in 4.0%. All other subtypes accounted for less than 1% each. Eleven patients had more than 1 genetically identified subtype of CMT. Patients with genetically identified CMT were separable into specific groups based on age of onset and the degree of slowing of motor nerve conduction velocities. Saporta et al. (2011) concluded that combining features of the phenotype and physiology allowed for identification of patients with specific subtypes of CMT, and proposed a strategy of focused genetic testing for CMT.

Clinical Management

In a double-blind, randomized, placebo-controlled pilot study, Sahenk et al. (2005) found that subcutaneous administration of the nerve growth factor neurotrophin-3 (NT3; 162660) promoted peripheral nerve regeneration and sensory improvement in 4 patients with CMT1A compared to 4 untreated patients. Similar results were observed for 2 mouse models of CMT: 1 with the common PMP22 duplication and 1 with a PMP22 point mutation. Sahenk et al. (2005) concluded that NT3 improved mutant Schwann cell survival and differentiation, resulting in increases in the available Schwann cell pool, which in turn increased the number of myelinated fibers.

Selles et al. (2006) reported that the Rotterdam Intrinsic Hand Myometer (RIHM) demonstrated excellent reliability in the measurement of intrinsic hand muscle strength in patients with CMT.

Shy et al. (2008) reported use of the 'CMT Neuropathy Score' (CMTNS) and the 'Neuropathy Impairment Score' (NIS) to determine the rate of disease progression in patients with CMT1A over time. The scoring systems could be used to monitor disease progression and standardize the efficacy of certain treatments.

Videler et al. (2008) presented evidence that loss of motor axons is the major cause of hand dysfunction in patients with CMT1A. Evaluation of fine motor skills of the hand in 48 patients with CMT1A showed a correlation between decreased pinch strength, clawing of the fingers, and decreased manual dexterity, and motor axon loss as measured by compound muscle action potentials and motor unit number estimation. There was no significant correlation between motor and sensory impairment.

Mapping

Middleton-Price et al. (1989, 1990), Nicholson et al. (1989), and Vance et al. (1989) presented evidence that one form of Charcot-Marie-Tooth disease is determined by a mutation on chromosome 17. In all 3 reports, high lod scores were obtained for linkage to D17S58 and D17S71. The disorder mapping to chromosome 17 was referred to as Charcot-Marie-Tooth disease type 1A or hereditary motor and sensory neuropathy type I (HMSN I).

In studies of 7 families, Chance et al. (1989, 1990) found a high probability of linkage to chromosome 17 markers in 5. Of the other 2, linkage to the Duffy blood group was suggested in 1 and excluded in the other. In the 2 families that did not show linkage to chromosome 17, the disease was more severe than in the chromosome 17 families.

In a multigenerational family in Belgium, Raeymaekers et al. (1989) excluded chromosome 1 as the site of the mutation and demonstrated linkage to D17S58 and D17S71. By further linkage studies in this family, Timmerman et al. (1990) demonstrated that the CMT1A gene is located in the chromosomal region 17p12-p11.2 between marker D17S71 and the gene for myosin heavy chain polypeptide-2 of adult skeletal muscle (MYH2; 160740). In a large French-Acadian kindred, Patel et al. (1990) confirmed the localization of CMT1A to the pericentromeric region of chromosome 17. McAlpine et al. (1990) provided linkage data on 5 Caucasian families which excluded linkage of CMT1A to the Fy area of chromosome 1 and demonstrated close linkage to D17S58, located at 17p11.2-p11.1; maximum lod = 10.828 at theta = 0.0. The CMT1A locus appeared to be proximal to MYH2, which maps to 17p13. By differential Alu-PCR of a rodent-human hybrid cell containing only chromosome 17 and a rodent-human cell containing only chromosome 17 with a deletion of the p11.2 band, Patel et al. (1990) isolated a marker that showed linkage to CMT1A with a peak lod score of 3.41 at a recombination fraction of 0.12. By multipoint linkage analysis, Vance et al. (1991) localized the CMT1A gene to 17p11.2 and identified flanking DNA markers. Lebo et al. (1992) studied the order of markers in the region of the CMT1A gene by means of multicolor in situ hybridization which they showed could resolve loci within 0.5 Mb on early-metaphase chromosomes.

Molecular Genetics

Common 1.5-Mb Duplication on Chromosome 17p12-p11

See 601097.0001 for discussion of the work of Lupski et al. (1991) and others indicating that a DNA duplication on chromosome 17 in the p12-p11.2 region is frequently the basis of CMT1A.

Hoogendijk et al. (1992) found that 9 of 10 sporadic patients had the duplication in chromosome 17 as a de novo change. Hertz et al. (1994) also demonstrated that a sporadic case of CMT1A was due to de novo duplication of the 17p12-p11.2 region. In all 12 de novo CMT1A duplications reported to that time, the duplication was of paternal origin. Sorour et al. (1995) described a case of CMT1A with molecular duplication of 17p12-p11.2 and inheritance of the duplication from a mosaic father. Whereas the patient had typical clinical features, the father had minimal findings of CMT1A.

To investigate the frequency of de novo CMT1A duplications, Blair et al. (1996) examined 118 duplication-positive CMT1A families. In 10 of these families it was demonstrated that the disease had arisen as the result of a de novo mutation. They estimated that 10% or more of autosomal dominant CMT1 families are due to de novo duplications. Using polymorphic markers from within the duplicated region, they showed that 7 of the duplications were of paternal and 1 of maternal origin. This was the first report of a de novo duplication of maternal origin. Bort et al. (1997) reported that the prevalence of de novo mutation in duplication-positive CMTA1 families was 18.3%. They reported that the ratio of maternal to paternal origin of the duplication was 1:8 in their study.

Weterman et al. (2010) identified a heterozygous 186-kb deletion on chromosome 17p12 with breakpoints within the common 1.5-Mb duplication but not involving the PMP22 gene in 6 probands with CMT1A. The duplication segregated with the disorder in 2 families and was absent in more than 2,000 control chromosomes. Haplotype analysis of 2 families suggested a founder effect. The junction breakpoints were located in a repeat-rich region, located 90-kb from the proximal CMT repeat region on 1 side and 3-kb upstream of PMP22 between PMP22 and TEKT3 (612683) on the other side. The junctions created by this duplication were located outside of any known genes or open reading frames, and there was no indication for the involvement of genes located within the duplication. Weterman et al. (2010) postulated that this duplication affects PMP22 expression levels through an as yet unidentified mechanism. Weterman et al. (2010) noted that the 186-kb duplication would not be detected in most diagnostic assays.

Point Mutations in the PMP22 Gene

In a Dutch kindred with CMT1A, Valentijn et al. (1992) identified a point mutation in the PMP22 gene (601097.0002). Thus, either duplication or point mutation in the PMP22 gene can result in CMT1A.

Fabrizi et al. (1999) reported a family in which 4 individuals over 4 generations had severe CMT1A with focal myelin thickenings with a regular fusiform contour (tomacula) or a coarsely granular appearance. Ultrastructural examination disclosed uncompacted myelin and redundant irregular myelin loops. All affected patients had a heterozygous mutation in the PMP22 gene (601097.0016). Fabrizi et al. (2000) noted that myelin outfoldings have been described in other autosomal dominant CMT patients with mutations in MPZ (159440.0023), EGR2 (129010.0004), and PMP22, and that the finding is not restricted to CMT4B (see CMT4B1; 601382).

Kleopa et al. (2004) reported a family from Cyprus in which 4 affected individuals had features of HNPP and/or CMT1A. One patient presented with typical HNPP, which later progressed to severe CMT1, 2 patients had HNPP with features of CMT1, and 1 patient had a chronic asymptomatic CMT1 phenotype. All 4 patients had the same heterozygous point mutation in the PMP22 gene (601457.0019). Kleopa et al. (2004) emphasized the broad phenotypic spectrum resulting from mutations in the PMP22 gene, as well as the phenotypic overlap of HNPP and CMT1A.

Genotype/Phenotype Correlations

Suter and Patel (1994) reviewed and discussed the curious finding that gene dosage and point mutations affecting the same gene can lead to a similar phenotype. They pointed to a possibly identical situation with Pelizaeus-Merzbacher disease (312080) in which either deletion of the entire locus encoding proteolipid protein (PLP1) (Raskind et al., 1991), as described in 300401.0006, or duplication of the PLP1 locus (Cremers et al., 1987) can cause Pelizaeus-Merzbacher disease.

Gabreels-Festen et al. (1995) compared the histology of peripheral nerve in patients with duplication of the PMP22 gene to those with point mutations. In the duplication cases, onion bulbs developed gradually in the first years of life, and the ratio of the axon diameter versus the fiber diameter was significantly lower than normal. In contrast, in patients with point mutations in PMP22, nearly all myelinated fibers had a high ratio of axon diameter versus fiber diameter, and onion bulbs were abundant from an early age.

Pellegrino et al. (1996) illustrated how it is possible in some instances to determine the genetic basis of clinical features in chromosomal rearrangements. They reported a child with monosomy 10q and dup(17p) resulting from an apparently balanced maternal translocation t(10;17)(q26.3;p11.2). Manifestations of both the duplication and the monosomy were present; however, the overall development was better than that previously reported in either syndrome. The patient's motor development was significantly more impaired than cognitive development, and signs of a peripheral neuropathy were found and attributed to duplication of 17p. Indeed, the patient was found to be trisomic for the PMP22 gene resulting in demyelinating neuropathy. An elevated serum alpha-fetoprotein had been detected at 16 weeks of gestation. The infant showed bilateral inguinal hernias and hydroceles at birth, and echocardiogram demonstrated ventriculoseptal defect (VSD) and bicuspid aortic valve. There was gastroesophageal reflux requiring Nissen fundoplication with gastrostomy tube. The VSD closed spontaneously. Hypoplastic corpus callosum was demonstrated by MRI. Terminal deletions of 10q had been reported in 26 patients, resulting in a definite phenotype (Wulfsberg et al., 1989). The manifestations included postnatal growth retardation, microcephaly, downslanting palpebral fissures, clinodactyly, syndactyly, congenital heart disease, and urogenital anomalies, all of which were present in the patient reported by Pellegrino et al. (1996).

Gouvea et al. (2010) reported an unusual case of a father and daughter with CMT who had 2 mutations in the PMP22 gene: the common 1.4-Mb duplication and S72L (601097.0007). Restriction analysis indicated that the S72L mutation was only present in 1 of the 3 PMP22 genes for both father and daughter. Both patients had a relatively mild form of the disease, manifest mainly as generalized pain without significant motor or sensory defects. The findings suggested that presence of 2 mutations in the PMP22 gene results in an attenuated form of the disease rather than a more severe form. Gouvea et al. (2010) hypothesized that the increased dosage resulting from the 1.4-Mb duplication offset the toxic gain-of-function effects of the S72L mutation on intracellular trafficking.

In 2 unrelated patients with a severe form of CMT1A, Liu et al. (2014) identified a 1.4-Mb triplication of the PMP22 gene (601097.0022). Each individual was part of a family with autosomal dominant CMT1A in which the other affected family members had the common 1.4-Mb duplication and a more typical CMT1A phenotype that was less severe. In both families, molecular analysis of the triplication indicated that it occurred on the chromosome with the duplication and arose from the duplication during meiosis in the affected mother. Haplotype analysis indicated 2 different mechanisms: in 1 family, the triplication arose via intrachromosomal nonallelic homologous recombination (NAHR), whereas in the other family it arose from intrachromosomal NAHR followed by a gene-conversion event that most likely exchanged alleles between the maternal homologous chromosomes. A review of a database for CMT1A duplication testing identified 13% with duplication and 0.024% with a duplication-to-triplication event. These findings suggested that the rate of duplication to triplication is higher than that of de novo duplication. Liu et al. (2014) proposed that individuals with duplications are predisposed to acquiring triplications and that the population prevalence of triplication may be underestimated. The inheritance pattern in this scenario resembles genetic anticipation and has implications for genetic counseling.

Pathogenesis

Katona et al. (2009) found highly variable levels of PMP22 protein, ranging from below normal to above normal, in skin biopsies from 20 patients with CMT1A due to the common 1.4-Mb duplication in the PMP22 gene. The findings were somewhat surprising, since an overall increase in PMP22 gene expression and protein was expected. In addition, there was no correlation between PMP22 mRNA or protein levels and phenotypic severity in CMT1A. In contrast, 6 individuals with HNPP (162500) due to the reciprocal PMP22 deletion had similarly decreased PMP22 levels. Katona et al. (2009) noted that PMP22 levels are tightly coordinated in the normal state, and that about 90% of translated PMP22 is rapidly degraded and never inserted into myelin (Pareek et al., 1997). Katona et al. (2009) hypothesized that dysregulation and variability of PMP22 expression may cause CMT1A, but noted that the disorder cannot result simply from a loss of normal function, since the phenotypes of CMT1A and HNPP are so different. The findings indicated that phenotypic variability in CMT1A cannot be explained by PMP22 levels in myelin.

Population Genetics

Lupski et al. (1992) stated that CMT in all of its forms is the most common inherited peripheral neuropathy in humans, with a total prevalence rate of 1 in 2,500. In a series of 172 index cases of Italian families in which there was at least 1 subject with CMT1, Mostacciuolo et al. (2001) found that among 170 informative unrelated patients, the frequency of the chromosome 17 duplication was 57.6%. A difference was observed between the duplication frequency in familial (71.6%) as opposed to nonfamilial cases (36.8%). Among the patients without the duplication, 2 had mutations in the PMP22 gene, 12 in the GJB1 gene (304040), 4 in the MPZ gene (159440), and none in the EGR2 gene (129010).

Among 153 unrelated patients with CMT, Boerkoel et al. (2002) found that 79 had a PMP22 duplication.

The 1.5-Mb duplication of PMP22 is the predominant cause of autosomal dominant CMT1, accounting for approximately 70% of all cases (Reilly, 2005).

Among 227 Japanese patients with demyelinating CMT, Abe et al. (2011) found that 53 (23.3%) carried PMP22 duplications and 10 (4.4%) carried PMP22 mutations. These were the most common genetic causes of demyelinating CMT, but the frequency of duplications was less than that observed in Caucasian populations. A molecular basis for demyelinating CMT could not be identified in 111 Japanese patients.

Animal Model

In a review of hereditary motor and sensory neuropathies, Vance (1991) pointed to the autosomal dominant 'Trembler' mutation (Tr) in the mouse as a possibly homologous condition. A hypomyelin neuropathy with onion bulb formation develops in older animals. Because of the extensive homology of synteny between mouse 11 and human 17 (Green, 1989), it is particularly attractive to think that these may be fundamentally the same disorder. In 2 allelic forms of the Trembler mouse, Suter et al. (1992, 1992) demonstrated point mutations in 2 distinct putative membrane-associated domains of a potentially growth-regulated 22-kD protein, peripheral myelin protein-22 (Pmp22). PMP22 is expressed by Schwann cells and is localized mainly in compact peripheral nervous system myelin. Baechner et al. (1995) demonstrated widespread distribution of PMP22 RNA in several mesodermal and ectodermal tissues of developing mice, as well as in the villi of the adult gut, suggesting to them a broader biologic significance for Pmp22 in cell proliferation or differentiation.

Huxley et al. (1996) constructed a mouse model for CMT1A by pronuclear injection of a YAC containing the human PMP22 gene and a large proportion of the region duplicated in CMT1A. They noted that CMT1A represents a unique case in which partial trisomy of a major gene leads to the pathology. Yeast artificial chromosomes are ideal for creating animal models of overexpression of genes since they contain very large stretches of DNA within which not only the structural gene but the long range controlling elements that confer full levels of tissue-specific expression may be present. In 1 transgenic line, about 8 copies of the human DNA was integrated into a mouse chromosome. This mouse developed a peripheral neuropathy closely similar to that seen in human CMT1A, with progressive weakness of the hind legs, severe demyelination in the peripheral nervous system, and the presence of onion bulb formations.

Perea et al. (2001) generated a transgenic mouse model for CMT in which mouse Pmp22 overexpression occurs specifically in Schwann cells of the peripheral nerve and is switched off when the mice are fed tetracycline. Overexpression of Pmp22 throughout life (in the absence of tetracycline) causes demyelination. In contrast, myelination is nearly normal when Pmp22 overexpression is switched off throughout life by feeding the mice tetracycline. When overexpression of Pmp22 is switched off in adult mice, correction begins within 1 week and myelination is well advanced by 3 months, suggesting that the Schwann cells are poised to start myelination. Upregulation of the gene in adult mice (which had previously had normal Pmp22 expression) is followed by active demyelination within 1 week. The authors hypothesized that even adult mice are sensitive to the level of expression of Pmp22 with respect to homeostasis of the myelin sheath.

The steroid hormone progesterone has been shown to stimulate Pmp22 gene expression both in cultured Schwann cells and in adult mice (Melcangi et al., 1999). Using a rat model of CMT1A with extra copies of the Pmp22 gene, Sereda et al. (2003) demonstrated that the progesterone antagonist onapristone reduced overexpression of Pmp22 and improved the CMT phenotype in male mice, as indicated by maintenance of large axons and improved motor performance. Pmp22 mRNA was decreased by 15% in onapristone-treated animals. Sereda et al. (2003) suggested that a reduction in Pmp22 transcription may have a beneficial effect on the disease course.

In mutant mice overexpressing Pmp22, Passage et al. (2004) found that treatment with ascorbic acid resulted in amelioration of the CMT1A phenotype, as measured by improved motor function and increased survival. The treated mice also showed a 10-fold decrease in Pmp22 RNA in sciatic nerves. Passage et al. (2004) noted that ascorbic acid is a promoter of myelination, and proposed a mechanism of Pmp22 suppression via inhibition of cAMP.

In HeLa cell studies, Khajavi et al. (2007) observed that Tr and TrJ mutant human PMP22 proteins accumulated within the endoplasmic reticulum (ER) and induced apoptosis. Treatment of these cells with curcumin, a chemical compound derived from turmeric, decreased apoptosis, similar to that observed in a previous study of mutant MPZ (Khajavi et al., 2005). Oral curcumin treatment of TrJ newborn mice resulted in dose-dependent improved motor performance that correlated pathologically with decreased Schwann cell apoptosis, increased axonal caliber, and increased myelin thickness compared to untreated mice. Bioavailability studies showed low serum levels of curcumin with accumulation in sciatic nerves and brain of treated animals. Treated mice exhibited no side effects. Khajavi et al. (2007) concluded that curcumin reduced the toxic effects of mutant aggregation-induced apoptosis, possibly by enabling a more efficient protein-trafficking process in the ER.

Meyer zu Horste et al. (2007) found that long-term treatment with subcutaneous onapristone in 5-week-old rats with a duplication of the Pmp22 gene resulted in significantly increased muscle strength and muscle mass at 26 weeks. Detailed studies of peripheral nerves indicated that treatment reduced progressive axonal loss but did not alter the visible myelination pathology, suggesting that axonal support and myelin assembly are distinct physiologic functions of Schwann cells, both of which are interrupted in CMT. Pmp22 levels in cutaneous nerve at 9 weeks correlated with hind-limb grip strength at 26 weeks.