Charcot-Marie-Tooth Disease and RelatedNeuropathies: Mutation Distribution and

Genotype-Phenotype CorrelationCornelius F. Boerkoel, MD, PhD,1 Hiroshi Takashima, MD, PhD,1 Carlos A. Garcia, MD,3

Richard K. Olney, MD,4 John Johnson, MD,5 Katherine Berry, MS,5 Paul Russo, MD,6 Shelley Kennedy, MS,7

Ahmad S. Teebi, MD,7 Mena Scavina, DO,8 Lowell L. Williams, MD,9 Pedro Mancias, MD,10

Ian J. Butler, MD,10 Karen Krajewski, MS,11 Michael Shy, MD,11 and James R. Lupski, MD, PhD1,2

Charcot-Marie-Tooth disease (CMT) is a genetically heterogeneous disorder that has been associated with alterations ofseveral proteins: peripheral myelin protein 22, myelin protein zero, connexin 32, early growth response factor 2, peri-axin, myotubularin related protein 2, N-myc downstream regulated gene 1 product, neurofilament light chain, andkinesin 1B. To determine the frequency of mutations in these genes among patients with CMT or a related peripheralneuropathy, we identified 153 unrelated patients who enrolled prior to the availability of clinical testing, 79 had a 17p12duplication (CMT1A duplication), 11 a connexin 32 mutation, 5 a myelin protein zero mutation, 5 a peripheral myelinprotein 22 mutation, 1 an early growth response factor 2 mutation, 1 a periaxin mutation, 0 a myotubularin relatedprotein 2 mutation, 1 a neurofilament light chain mutation, and 50 had no identifiable mutation; the N-myc downstreamregulated gene 1 and the kinesin 1B gene were not screened for mutations. In the process of screening the above cohortof patients as well as other patients for CMT-causative mutations, we identified several previously unreported mutantalleles: two for connexin 32, three for myelin protein zero, and two for peripheral myelin protein 22. The peripheralmyelin protein 22 mutation W28R was associated with CMT1 and profound deafness. One patient with a CMT2 clinicalphenotype had three myelin protein zero mutations (I89N!V92M!I162M). Because one-third of the mutations wereport arose de novo and thereby caused chronic sporadic neuropathy, we conclude that molecular diagnosis is a nec-essary adjunct for clinical diagnosis and management of inherited and sporadic neuropathy.

Ann Neurol 2002;51:190–201DOI 10.1002/ana.10089

In some pediatric neuromuscular clinics and referralcenters, 40 to 50% of patients with a peripheral neu-ropathy have hereditary disease.1,2 Affecting approxi-mately 1 in 2,500 individuals, Charcot-Marie-Toothdisease (CMT) is the most common inherited disorderof the peripheral nervous system.3 On the basis of elec-trophysiological properties and histopathology, CMThas been divided into primary peripheral demyelinatingand primary peripheral axonal neuropathies. The pri-mary peripheral demyelinating neuropathies are charac-terized by demyelination with severely reduced motornerve conduction velocities (NCV), whereas the pri-mary peripheral axonal neuropathies are characterizedby axonal loss and normal or mildly reduced NCVs.4,5

The primary peripheral demyelinating neuropathiesconstitute a spectrum of neuropathy phenotypes, in-cluding Charcot-Marie-Tooth disease type 1 (CMT1,MIM 118200), Dejerine-Sottas syndrome (DSS, MIM145900),� congenital� hypomyelinating� neuropathy(CHN, MIM 605253), and hereditary neuropathywith liability to pressure palsies (HNPP, MIM162500).6,7 At least 15 genetic loci and seven geneshave been associated with these disorders; identified ge-netic causes include altered dosage of peripheral myelinprotein 22 (PMP22) or mutations in one of the fol-lowing genes: PMP22, the gap junction protein !1 orconnexin 32 gene (GJB1), the myelin protein zero gene(MPZ), the early growth response gene 2 (EGR2), the

From the Departments of 1Molecular and Human Genetics and2Pediatrics, Baylor College of Medicine, Houston, TX; 3Depart-ments of Neurology and Pathology, Tulane University Health Sci-ences Center, New Orleans, LA; 4Department of Neurology, Uni-versity of California San Francisco School of Medicine, SanFrancisco, CA; 5Clinical Genetics, Shodair Children’s Hospital,Helena, MT; 6Kaukauna Clinic, Kaukauna, WI; 7Division of Clin-ical and Metabolic Genetics, The Hospital for Sick Children, To-ronto, Canada; 8Division of Neurology, duPont Hospital for Chil-dren, Wilmington, DE; 9Columbus, OH; 10Department ofNeurology, University of Texas-Houston Health Science Center

Medical School, Houston, TX; and 11Department of Neurology,Wayne State University, Detroit, MI.

Received Jul 24, 2001, and in revised form Sep 20. Accepted forpublication Oct 5, 2001.

Published online Dec 13, 2001

Address correspondence to Dr Lupski, Department of Molecularand Human Genetics, One Baylor Plaza, Room 604B, Houston,TX 77030. E-mail: jlupski@bcm.tmc.edu

190� © 2001 Wiley-Liss, Inc.

myotubularin-related protein 2 gene (MTMR2), theN-myc downstream-regulated gene 1 (NDRG1), or theperiaxin gene (PRX). These genes encode proteins ofdiverse functions: compact myelin structural proteins(MPZ, PMP22), a noncompact myelin gap junctionprotein (GJB1), signal transduction proteins (NDRG1,MTMR2), a transcription factor for late myelin genes(EGR2), and a noncompact myelin cytoskeleton-associated protein (PRX). Both dominant (PMP22,GJB1, MPZ, EGR2) and recessive (MTMR2, NDRG1,PMP22, EGR2, PRX) mutant alleles have been de-scribed.

Primary peripheral axonal neuropathies also form acontinuum extending from severe infantile-onset dis-ease to mild adult-onset disease and include Charcot-Marie-Tooth disease type 2 (CMT2, MIM 118210)and giant axonal neuropathy (MIM 256850).6,7 Atleast 11 genetic loci and 5 genes have been associatedwith these disorders; identified genetic causes includemutations in the neurofilament light chain gene(NEFL), the kinesin 1B gene (KIF1B), and the gigaxo-nin gene (GAN1).6,8–10 In addition, reflecting the in-timate trophic interactions between neurons and glia,some mutations in GJB1 and MPZ also present withclinical and electrophysiologic findings of CMT2. Mu-tation of NEFL or KIF1B cause dominantly inheritedaxonal neuropathies, whereas mutation of GJB1 orMPZ can present as genocopies of dominant axonalneuropathies.11 The mechanism by which some GJB1and MPZ mutations cause pronounced early axonal pa-thology has not been delineated, although the patho-logic process likely begins with demyelination.12,13

We report the distribution of CMT mutations andclinical findings among patients referred before theavailability of clinical testing and in addition, CMT-causing point mutations for 16 unrelated families: 6with GJB1 mutations, 5 with MPZ mutations, and 5with PMP22 mutations. Furthermore, our genotype-phenotype correlation identifies another PMP22 muta-tion causing CMT and profound deafness and anotherde novo GJB1 mutation. Remarkably, one-third (6/16)of the point mutations represent de novo events inchronic sporadic neuropathy patients.

Patients and MethodsHuman SubjectsAll patients referred to this study by their primary physicianor neurologist received appropriate counseling and gave in-formed consent approved by the Institutional Review Boardof Baylor College of Medicine. We isolated DNA from theperipheral blood of each patient and established lymphoblas-toid cell lines.

Mutation Distribution CohortTo determine the distribution of CMT-causative genetic al-terations, we identified a cohort of 153 unrelated patientswho were enrolled prior to August 1993 when clinical testing

for the CMT1A duplication became available throughAthena Diagnostics. The distribution of clinical diagnoses inthis cohort was 141 patients with CMT1, 3 with DSS, 1with CHN, 1 with hereditary motor and sensory neuropathyV (HMSNV), 1 with HNPP, and 6 with CMT2. Because ofthe research studies being conducted at that time, this cohortwas strongly biased for a family history of peripheral neurop-athy; of 69 families for whom a mode of inheritance wasrecorded, 63 (91%) displayed dominant inheritance, 2 (3%)recessive inheritance, and 4 (6%) sporadic disease. Nearly allpatients included in this cohort were referred from neurologyor genetics clinics in the United States and were children ofnonconsanguineous unions.

Mutation Screening CohortThe patient cohort screened for mutations in PMP22, GJB1,and MPZ contained 159 unrelated patients who had testednegative for the CMT1A duplication and HNPP deletion.These 159 patients included 53 from the previous cohortplus 106 who were enrolled after August 1993. The distri-bution of clinical diagnoses was 145 families with a primaryperipheral demyelinating neuropathy (112 CMT1, 10 DSS,21 CHN, 2 HNPP) and 14 with a primary peripheral axonalneuropathy (CMT2). Patients included in this cohort werereferred from neurology or genetics clinics around the worldbut predominantly North America; the overwhelming major-ity was children of nonconsanguineous unions.

For both of the above patient cohorts, the referring phy-sician assigned the diagnosis to each proband. When avail-able, we confirmed the diagnosis by a review of accompany-ing clinical information and objective laboratory studies, eg,NCVs and nerve histopathology.

Mutation ScreeningWe identified all PMP22 coding exons by aligning the BACsequence (AC018616) with the human PMP22 cDNA(L03203); MPZ and GJB1 coding exons were derived frompublished sequences (GJB1: NM_000166; MPZ: L24894,L24893). We designed primers for PCR amplification of ex-ons and intronic splice junctions with the Primer v3 program(http://www-genome.wi.mit.edu/cgi-bin/primer/primer3_www.cgi) (Table 1). All forward primers had a "21 M13primer tail (TGTAAAACGACGGCCAGT) and all reverseprimers a M13 reverse tail (CAGGAAACAGCTATGACC).Using 50ng of patient genomic DNA, the appropriate prim-ers, and Qiagen HotStarTaq, we amplified each region asfollows: 15 minutes at 95°C, 40 cycles of amplification(95°C for 30 seconds, 65°C for 30 seconds, 72°C for 1minute) and 7 minutes at 72°C. The primers and PCR am-plification conditions for the PRX gene and the MTMR2gene have been previously published.14,15

Using the Qiagen 96-PCR purification kit (Qiagen, Valen-cia, CA), we purified PCR products amplified from thegenomic DNA of patients, relatives, and control chromosomesand sequenced the products with dye-primer chemistry (Ap-plied Biosystems, Foster City, CA) with an ABI377 automatedsequencer (Applied Biosystems). We aligned resulting se-quences and evaluated mutations with the Sequencher se-quence alignment program (ACGT Codes, Ann Arbor, MI).We numbered the GJB1, MPZ, and PMP22 cDNA sequencesbeginning with the adenine of the presumed initiating methi-

Boerkoel et al: CMT-Causing Mutations and Their Distribution 191

onine and described mutations according to den Dunnen andAntonarakis.16

Paternity TestingPaternity testing was done by Identigene (Houston, TX).

ResultsDistribution of CMT-Causative MutationsPrior to the availability of clinical testing for theCMT1A duplication, 153 unrelated peripheral neurop-athy patients were referred to our laboratory for re-search investigating the molecular bases of peripheralneuropathy; 91% of these patients had a hereditaryneuropathy as determined by family history. For eachpatient, we screened the coding exons of PMP22,MPZ, GJB1, and EGR2 by DNA sequencing and eachcoding exon of PRX, MTMR2, and NEFL by a com-bination of DHPLC and DNA sequencing. The distri-bution of mutations and diagnoses among these pa-tients is summarized in Table 2. Clearly, the CMT1Aduplication accounts for the majority of molecularlydiagnosed patients. Mutations of GJB1, MPZ, andPMP22 accounted respectively for 7.2, 3.3, and 4.0%of patients with an inherited peripheral neuropathy.Based on this cohort, mutations of EGR2, PRX, NEFL,and MTMR2 appear to be minor contributors to he-reditary peripheral neuropathy. Nearly one-third of pa-tients did not have an identifiable molecular cause oftheir disease by denaturing high-performance liquidchromatography (DHPLC) analysis or DNA sequenc-ing of coding regions for genes in which mutationshave been associated with neuropathy.

Mutation Analysis of GJB1, MPZ, and PMP22 inNeuropathy PatientsBy direct sequencing of the coding regions of GJB1,MPZ, and PMP22 in a different cohort consisting of159 unrelated peripheral neuropathy patients who hadtested negative for the CMT1A duplication and for amutation in PRX, EGR2, and MTMR2, we identified 6GJB1 mutant alleles (Table 3 and Fig 1), 5 MPZ mu-tant alleles (Table 4 and Fig 2), and 4 PMP22 mutantalleles (Table 5 and Fig 3). None of these mutant al-

Table 2. Distribution of Neuropathy-Causing Mutationsamong Patients Referred Prior to the Availability of ClinicalTesting for the CMT1A Duplication

Genetic Alteration PhenotypeNo. of

Patients (%)

17p12 duplication (51.6)CMT1 79a

DSS 0CHN 0

GJB1 mutation (7.2)CMTX 8DSS 0CMT2 3b

MPZ mutation (3.3)CMT1 5DSS 0CHN 0

PMP22 mutation (3.3)CMT1 3c

DSS 2CHN 0HNPP 0

EGR2 mutation (0.65)CMT1 1DSS 0CHN 0

PRX mutation (0.65)CMT1 0DSS 1CHN 0

MTMR2 mutation (0)CMT1 0DSS 0CHN 0

NEFL mutation (0.65)CMT1 0DSS 0CHN 0CMT2 1

No mutation (32.7)CMT1 44DSS 1CHN 1HNPP 1CMT2 2HMSNV 1

Total 153

a1 patient with deafness, 1 with spasticity, 1 with seizures.b1 male and 2 female patients.c1 patient with profound deafness.

Table 1. Primer Pairs Used for Amplifying the MPZ,PMP22, and GJB1 Coding Regions

Gene Primer Name Primer Pairs

MPZExon 1 F CCACCTCTCAACTGCACATG

R ATTGCTGAGAGACACCTGAGTCCExon 2 F TCCTCTGTATCCCTTACTGG

R TTTGAAGCACTTTCTGTTATCCExon 3 F GGAGCTAAGCTTTGACAGCTGTG

R ATCCCCTCCCAAACTGCTTCExons 4, 5 F CGGACTAGGAACCACAGATAC

R CTCCCAGGGTTCTCCTTCCCAExons 5, 6 F TGTGTCCGCGGTGCAAGGGGTTC

R CTTTGGGCCTTTGGCGGACTCPMP22

Exon 2 F CTAGTGCGCGGGACCCTCR CTGAACCAGCAGGAGCACGGGCTG

Exon 3 F CATGCAGGGGTGGGCGGTGTGR GGGCTGAGAAACGTGTTACAG

Exon 4 F TGGCCCTTCAGGCCCTGCACCTR CCCACACATACAAGCACCCACCCTCA

Exon 5 F TTCCTACCCAGCAATTGTCAGCR CTTCCTCCCTTCCCTATGTACG

GJB1Amplicon 1 F CCAGCTTTCTGACAGCTTGCT

R CTCAAACAACAGCCGGAACACAmplicon 2 F GCACAAGGTCCACATCTCAGG

R AGGGCAGGGTCGGGGATGGATG

192 Annals of Neurology Vol 51 No 2 February 2002

leles were observed in #160 North American controlchromosomes. Of the 16 mutant alleles for which weprovide clinical information, 7 (GJB1: L108P, N205I;MPZ: I89N$V92M$I162M, G123C, V136E; PMP22:W28R, L71P) are novel, 3 (GJB1: R15W,17–22 E102G,23

R215W 24–26) have been reported previously with little orno clinical information, and 5 (GJB1: R22Q18,20,23,27–30;MPZ: S78L,24,28,31–34 Y82C28,34–36; PMP22: S72L,37–41

G94fsX11042) have been reported with a phenotypic de-scription. Clinical information for patients with GJB1,MPZ, and PMP22 mutations is listed in Tables 3, 4, and5, respectively.

DiscussionCMT and related hereditary peripheral neuropathiesare caused by mutation of one of several differentgenes. By analyzing DNA samples from a mixed cohortof 153 patients with predominantly inherited periph-eral neuropathy who were referred to our research lab-oratory before the availability of clinical testing, we de-termined the relative contribution of mutations invarious genes to the CMT phenotype. Among this pa-tient population who were not selected for a peripheraldemyelinating phenotype, 52% (79/153) had a dupli-cation of 17p12; this fraction rises to 70% (79/113) ashas been previously reported if patients with a periph-eral demyelinating neuropathy are analyzed.18,43,44

Based on a thorough clinical evaluation of probandsand family members, previous studies found that 40%of patients presenting to neuromuscular clinics with aperipheral neuropathy have hereditary disease.1,2 Ourobservation that half of the patients in our cohort havea duplication of 17p12 suggests that approximately20% of patients presenting in adolescent and adultneuromuscular clinics with findings of chronic periph-eral neuropathy will have a duplication of 17p12 as thecause of their disease.

Less than 20% of the patients in the cohort referredbefore the availability of clinical testing had a mutationwithin the coding sequence of GJB1, MPZ, PMP22,EGR2, PRX, MTMR2, or NEFL. Among the 25 pa-tients who did, mutations in GJB1 accounted for 44%,mutations of PMP22 for 24%, mutations in MPZ for20%, and mutations in EGR2, PRX, or NEFL for 2%each. Both within our cohort and in the Inherited Pe-ripheral Neuropathies Mutation Database (http://molgen-www.uia.ac.be/CMTMutations/), transversionsand transitions each account for approximately half ofthe point mutations within PMP22, MPZ, and GJB1.This ratio of transversions to transitions is similar tothat observed for pathogenic mutations in other hu-man diseases.45

Of the patients referred before the availability ofclinical testing, approximately one-third did not have

Table 3. Patients with Myelinopathy Secondary to a GJB1 Mutation

Family (Patient)

HOU103(399)

HOU177(599)

HOU232(697)

HOU232(698)

HOU237(714)

HOU303(860)

HOU337(910)

Mutation 43C#T 643C#T 65G#A 65G#A 305A#G 323T#C 614A#TR15W R215W R22Q R22Q E102G L108P N205I

Clinical diagnosisa CMTX CMT1 CMT2 CMT2 CMT2 CMT2 CMT2 $deafness

Sex M M M F F F FInheritance pattern XD XD XD XD XD Sporadic XDAge at onset (yrs) N/A 32 12 14 N/A 33 10Current age (yrs) 65 49 72 38 52 44 75Motor involvement N/A Distal severe Distal severe Distal moderate Distal severe Distal mild Distal mildAge of walking (yrs) N/A N/A 1 1.1 N/A 1 1Cranial nerve involve-

mentN/A No No No N/A No Yes

Sensory loss N/A Yes No No N/A Yes YesUnsteady gait N/A Yes No No N/A Yes YesDeep tendon reflexes N/A Absent Absent Normal Decreased Decreased DecreasedFoot deformity N/A N/A Pes cavus Pes cavus N/A No Pes cavusMotor nerve conduction

velocity (median orulnar)

N/A Slow 28–29m/s 56.6m/s 38m/s 47m/s 44–47m/s

Peripheral nerve histo-pathology

N/A N/A N/A N/A N/A Axonal loss N/A

aClinical diagnosis is the diagnosis that the patient had at the time he or she was entered in the study; with the results of molecular testing, allof these patients would now be classified as CMTX (Charcot-Marie-Tooth disease X). A diagnosis of CMTX would resolve the apparentdisparate diseases in family 232.

CMT % Charcot-Marie-Tooth disease; M % male; F % female; XD % X-linked dominant; N/A % not available.

Boerkoel et al: CMT-Causing Mutations and Their Distribution 193

an identifiable genetic cause for their chronic neurop-athy; this observation combined with the lack of an-other identifiable major genetic locus strongly suggeststhat mutations in additional genes will cause CMT andrelated inherited peripheral neuropathies.7 However,because we only analyzed the coding region of eachgene, some patients may have had mutations that werenot detected by our analyses (a noncoding region mu-tations, gene duplication, gene deletion, or deletion ofPCR primer annealing sites) and others “nongenetic”or acquired chronic neuropathies. These patients donot have an ascertainable phenotypic difference or aless secure neurologic diagnosis than those patients inwhom we were able to identify a genetic cause of neu-ropathy.

During screening for MPZ, GJB1, and PMP22 mu-tations in our second cohort of 159 unrelated patientswithout the 17p12 duplication or deletion, we ob-served that one-third (6/16) of patients in whom weidentified a mutation had a de novo mutation; 1 had ade novo GJB1 mutation, 3 a de novo PMP22 muta-tion, and 2 a de novo MPZ mutation. Thus de novopoint mutations appear to be an important cause ofsporadic neuropathy. This observation is reminiscent ofprevious reports in which a significant fraction of pa-tients with sporadic peripheral demyelinating neuropa-thy have a de novo the 17p12 duplication or dele-tion.18,46 Based on these findings, we conclude thatmolecular diagnosis is a necessary adjunct for differen-tiating genetic and acquired peripheral neuropathieseven in nonfamilial chronic neuropathy.

GJB1 MutationsMutations of GJB1 can cause a neuropathy with clin-ical and electrophysiologic features of either a demyeli-nating or an axonal neuropathy.27,47 Our patients 663(R238H), 697 (R22Q), 698 (R22Q), 714 (E102G),and 910 (N205I) all carried a clinical diagnosis ofCMT2, whereas patients 399 and 599 had a diagnosisof a demyelinating neuropathy, CMTX and CMT1,respectively. In general, as demonstrated by familyHOU232, male patients with GJB1 mutation usuallyhave a more severe and earlier-onset neuropathy thanfemale patients do and often have clinical and electro-physiologic signs suggestive of a peripheral demyelinat-ing neuropathy, whereas female patients such as 663,698, 714, and 910 often have signs consistent with anaxonal neuropathy,6,7 although animal models suggestthat the primary cause of this axonal pathology is de-myelination.12,13 One would speculate that the severityof disease among women depends on Lyonization aswell as the nature of the mutation, genetic backgroundand environment.

The mechanism by which GJB1 mutations lead todisease has not been fully elucidated, but the gap junc-tions formed by connexin 32 provide a radial diffusionpathway for nutrients, metabolites, and ions throughthe noncompact regions of the myelin sheath.27,47

Those mutations of GJB1 that have been studied haveresulted generally in either a loss of ability to formfunctional gap junction channels or an alteration in thegating properties of the channels.47–49 These two ef-fects have been observed as the result of dominant neg-

Fig 1. GJB1 alterations identified in six families. Family HOU103 exhibited X-linked recessive inheritance; families HOU177,HOU232, HOU237, and HOU337 showed X-linked dominant inheritance; and family HOU303 had sporadic disease. The geno-types are shown below each individual tested. All affected men were hemizygous and affected women heterozygous for their respectivemutations. Black symbols indicate affected status. We confirmed the paternity of patient 860 by analysis of polymorphic repeats.

194 Annals of Neurology Vol 51 No 2 February 2002

ative and loss-of-function mutations.27,50,51 Of themutations we report, GJB1 proteins with mutationR215W do not make functional channels,51 and R22Qmutant GJB1 proteins probably do not because muta-tions R22P and R22G do not.49 GJB1 proteins con-taining mutations E102G and R238H form channelswith altered gating properties.49,52,53 Functional stud-ies have not been reported for mutations R15W,L108P, and N205I.

GJB1 is expressed in Schwann as well as oligoden-drocytes cells and many other nonneural tissues54;however, clinically evident dysfunction from GJB1 mu-tations is generally limited to the peripheral nervoussystem. Recent studies of brainstem-evoked responsesamong patients with GJB1 mutations have identifiedcommon subclinical abnormalities of these responses,suggesting that the central nervous system is involved,and other reports have suggested that rare patients haveCNS demyelination.19,55,56 These observations andsegregation of deafness with the GJB1 mutationR142Q suggest that some GJB1 mutations can causeCNS involvement such as clinical deafness as well asperipheral neuropathy.57 In light of this, the severe andearly onset deafness observed in patient 910 may havebeen caused by the GJB1 mutation N250I, althoughthe absence of self-reported hearing loss among the af-

fected children of this patient suggests that this muta-tion is not sufficient to cause deafness, or that there isan age-dependent penetrance to the deafness pheno-type, or that patient 910 has another cause for herdeafness.

MPZ MutationsPatients with mutations of MPZ, which encodes P0,usually have clinical and electrophysiologic features of aperipheral demyelinating neuropathy, although severalpatients with axonopathic features have been reportedalso. Our patients 238 (Y82C), 1333 (S78L), 1370(G123C), and 1567 (V136E) all had a clinical diagno-sis of CMT1 or DSS, whereas patient 598(I89N$V92M$I162M) had a diagnosis of CMT2.

P0 is expressed exclusively in compact myelin and isrequired for formation and development of compactmyelin. Based on the crystal structure of the extracel-lular domain of P0 and functional studies, P0 appar-ently mediates extra- and intracellular juxtaposition ofthe Schwann cell plasma membrane by homophilic in-teractions of interdigitating homotetramers.58 Muta-tions disrupting the interactions within the homotet-ramer or between homotetramers hypotheticallyproduce disease by a dominant negative effect, whereasmutations retarding production of P0 produce disease

Table 4. Patients with Myelinopathy Secondary to a MPZ Mutation

Family (Patient)

HOU63(238)

HOU176(598)

HOU518(1333)

HOU536(1370)

HOU573(1567)

Mutation 245A#G,Y82C

266T#A, I89N $274G#A,V92M $ 486C#G, I162M

233C#T,S78L

367G#T, G123C 407T#A,V136E

Clinical diagnosis CMT1 CMT2 CMT1 DSS DSSSex F F M F FInheritance pattern AD AD Sporadic Sporadic ADAge at onset (yrs) 5 N/A 9 1 &10Current age (yrs) 25 50 11 52 61Motor involvement Distal

mildDistal mild Distal

mildDistal severe Distal severe

Age of walking (yrs) N/A N/A 1 Never independently 1.8Cranial nerve involve-

mentNo N/A No No No

Sensory loss Yes Yes Yes Yes YesUnsteady gait Yes N/A No Yes YesDeep tendon reflexes Absent Decreased Decreased Absent AbsentFoot deformity Pes cavus N/A Pes cavus Pes cavus, hammer

toesPes cavus

Motor nerve conductionvelocity (median orulnar)

13.7m/s 48–52m/s 29.6m/s 4m/s UD

Peripheral nerve histo-pathology

N/A N/A N/A N/A N/A

CMT % Charcot-Marie-Tooth disease; DSS % Dejerine-Sottas syndrome; F % female; M % male; AD % autosomal dominant; N/A % notavailable; UD % undetectable.

Boerkoel et al: CMT-Causing Mutations and Their Distribution 195

by a loss-of-function effect.59,60 Mutations Y82C,S78L, G123C, V136E, I89N, and V92M all lie withinthe extracellular domain of P0. Mutations (Y82C andG123C) introducing a thiol group into the extracellu-lar domain likely have dominant negative effects causedby aberrant crosslinking between P0 monomers or ho-motetramers. Based on interactions defined by the ex-tracellular P0 crystal structure, V136 lies within a do-main mediating interaction of P0 molecules fromapposing membranes and thus mutation of this residuewould be expected to disrupt those interactions and theformation and maintenance of compact myelin.58 Res-idues I89 and V92 lie outside defined P0 adhesion do-mains and I162 is in the transmembrane domain; al-though no experimental evidence is reported for theeffects of mutation of I89, V92, and I162, these mu-tations might alter the secondary structure, trafficking,and/or stability of P0.

The mechanism underlying expression of a predom-inantly axonopathic versus a predominantly demyeli-nating phenotype for a given MPZ mutation remainsunclear, although demyelination is likely the primarycause of both phenotypes. Of the six mutations (S44F,D61G, D75V, A76V, Y119C, T124M) previously re-ported among patients with CMT261–67 and of thethree mutations (I89N$V92M$ I162M) in our pa-tient 598, all mutant amino acids are in the extracel-lular domain of P0 except I162M, which is in thetransmembrane domain. However, only D61 and D75are in defined adhesion domains. With the exception

of T124M, which has been identified in patients withCMT1 and CMT2,61,62,65–67 each mutation has onlybeen reported in patients with CMT2. These observa-tions might suggest that in many instances, the CMT2phenotype is determined by the amino acid mutatedwithin P0.

Complex alleles such as that observed in our patient598 have been observed in 2 patients with GJB1 mu-tations and 3 with MPZ mutations.34,68–71 However,only 1 patient has been reported previously with threemutations, and that patient also had three MPZ muta-tions.70 Only one of the three mutations (486C#G) inour patient occurred at a CpG dinucleotide, and theother two mutations (266T#A, 274G#A) do not re-side in a region of DNA with any obvious secondarystructure that would predispose this region to mutationthrough pseudopalindrome formation.72 Moreover, theobservation that none of these three mutations havebeen previously reported argues against these bases be-ing mutation hotspots. Adjacent mutations such as266T#A and 274G#A have been observed withchemical mutagenesis, ionizing radiation, and imper-fect DNA patch repair; the operation of one of thesemechanisms in the gametes or gametic precursors of anancestor of patient 598 could account for the occur-rence of three mutations on one MPZ allele.73,74 DNAsamples were not available from the affected father ofpatient 598 to determine if he also carried all threemutations or whether patient 598 had accumulated ad-ditional mutations.

Fig 2. MPZ alterations identified in five families. The genotypes are shown below each individual tested. All patients were het-erozygous for their respective mutations. Black symbols indicate affected status. Parental DNA was not available to confirm thepaternity of patients 598, 1333, and 1370. Analyses of cloned PCR fragments from patient 598 were consistent with all three mu-tations residing on one allele.

196 Annals of Neurology Vol 51 No 2 February 2002

PMP22 MutationsPatients with mutations of PMP22 usually have clinicaland electrophysiologic features of a peripheral demyeli-nating neuropathy such as HNPP, CMT1, DSS, orCHN. Our patients 666 and 667 (W28R) had CMT1with severe to profound sensorineural deafness; patients1110 (S72L) and 1390 (S72L) had CHN, and patient1398 (L71P) had CMT1, whereas her daughters hadsevere CMT1 suggestive of DSS.

Within the peripheral nervous system, PMP22 ispredominately expressed in compact myelin.75 The in-sertion of PMP22 into the membranes composingcompact myelin is promoted by contact between theSchwann cell and axon.76 Genetic studies of PMP22have shown that PMP22 is necessary for both myelinformation and maintenance. Some studies have shownthat PMP22 and P0 complex together within Schwanncells; this association suggests that some mutations ofPMP22 might cause disease by destabilizing the P0 in-teractions necessary for the formation and maintenanceof compact myelin.77 In addition, in vitro studies ofthe PMP22 mutant S72L have shown that the mutantprotein is not transported to the plasma membrane,and similar studies of mutants G150D and L16P haveshown that these two mutant proteins also inhibit the

trafficking of wild type PMP22 to the plasma mem-brane.78–80 The retained wild-type and mutantPMP22 accumulate within the cell and appear to havea toxic effect on the Schwann cell.81–83 By generaliza-tion, therefore, most PMP22 missense mutations arethought to cause disease by a combination of a domi-nant negative effect that inhibits the transport of mu-tant and wild-type PMP22 to the plasma membraneand a toxic gain of function effect resulting from theaccumulation of PMP22 within the cell.

PMP22 is expressed in cranial nerves but not in themature central nervous system; however, during devel-opment it is expressed initially in all three germ layersand subsequently in migratory neural crest cells.84,85

These observations suggest that mutations of PMP22might cause sensorineural deafness by demyelination ofthe eighth cranial nerve or by maldevelopment of theinner ear, which is a neural crest derivative, or by acombination of the two. These two mechanisms can-not be distinguished by the available experimental data;nevertheless, the rarity of severe deafness among fami-lies with PMP22 mutations suggests that most PMP22mutations have minimal effects on inner ear develop-ment or cranial nerve myelination. Nevertheless, thePMP22 mutation W28R segregates with profound

Table 5. Patients with Myelinopathy Secondary to a PMP22 Mutation

Family (Patient)

HOU216(666)

HOU406(1110)

HOU524(1348)

HOU548(1390)

HOU551(1397)

HOU551(1398)

Mutation 82T#CW28R

215C#T S72L 281delG,G94fsX110

215C#T S72L 212T#CL71P

212T#CL71P

Clinical diagnosis CMT1 $deafness

CHN DSS CHN DSS CMT1

Sex F F F F F FInheritance pattern AD Sporadic Sporadic Sporadic AD ADAge at onset (yrs) 13–15 Birth 1 0.5 0.9 25Current age (yrs) 46 13 27 8 10 35Motor involvement Distal, severe Distal, severe Distal, se-

vereDiffuse, severe Proximal, moderate Diffuse,

moderateAge of walking (yrs) 1 3–4 Nonambulatory 5 N/ACranial nerve involvement Yes No Yes No No NoSensory loss Yes No Yes No Yes YesUnsteady gait Yes Yes Yes Yes Yes YesDeep tendon reflexes Absent Absent Absent Absent Absent AbsentFoot deformity Pes cavus N/A Pes cavus,

hammertoes

No Pes planus Pes cavus

Motor nerve conductionvelocity (median orulnar)

13m/s 3.7m/s 4.4m/s N/A 2–3m/s 2–3m/s

Peripheral nerve histopa-thology

N/A Absent or thinmyelin sheath,rare OBF,rare Walleriandegeneration

N/A Absent or thinmyelin sheath,early OBF,loss of MF

Severe loss of MF,OBF

N/A

CMT % Charcot-Marie-Tooth disease; CHN % congenital hypomyelinating neuropathy; DSS % Dejerine-Sottas syndrome; F % female;AD % autosomal dominant; N/A % not available; OBF % onion bulb formation.

Boerkoel et al: CMT-Causing Mutations and Their Distribution 197

early onset deafness and CMT1 in family HOU216and the mutation A67P with deafness and CMT1 inthe family reported by Kovach and colleagues.86 Basedon molecular modeling of the PMP22 protein, bothW28R and A67P are located at the base of the firstextracellular loop; thus these mutations might be adja-cent in the protein and effect hearing loss through acommon mechanism.

ConclusionsIn summary, CMT is a genetically and clinically het-erogeneous disease for which genotype-phenotype cor-relations are difficult to define. Both the phenotypicfeatures and disease severity can either be consistent orvary widely both within and among families. Themarked difference in disease severity between themother and daughters of family HOU551 (PMP22,L71P) exemplifies intrafamilial variation in phenotypicseverity. The inconsistent finding of hearing lossamong affected members in family HOU337 (GJB1,N205I) illustrates intrafamilial variation in phenotypicfeatures. On the other hand, illustrating intrafamilialconsistency, all affected members of family HOU216(PMP22, W28R) develop adolescent-onset CMT1 andprofound deafness. Also, patient 1348 (PMP22,G94fsX110) has features almost identical to those ob-served in the patient previously reported by Ionasescuand colleagues42 and thus demonstrates the interfamil-ial consistency observed with some mutations. The am-

biguity in genotype-phenotype correlation probably re-flects limitations in our knowledge and understandingof the effects of genetic background, environment, andthe biologic function of the affected proteins. Never-theless, continued reporting of the clinical features ofmolecularly defined CMT patients and correlation ofthose features with an increasingly better biochemicaland biological understanding of the proteins withCMT-causing mutations is necessary to develop betterprognostic capabilities.

This study was supported in part by grants from the National In-stitute of Diabetes, Digestive, and Kidney Diseases, NIH (K08DK02738) to CFB, and from the National Institute of NeurologicalDisorders and Stroke, NIH (R01 NS27042), and the MuscularDystrophy Association to JRL.

We thank the families described for cooperation. We thank J.Badano and N. Katsanis for defining and confirming the 17p12deletion and duplication status of the samples in our patient cohort.We thank V. Timmerman and A. Jordanova for screening our pa-tient cohort for NEFL mutations and A. Bolino for screening it forMTMR2 mutations. HT is a recipient of a postdoctoral fellowshipfrom the Charcot-Marie-Tooth Association.

References

1. Dyck PJ, Oviatt KF, Lambert EH. Intensive evaluation of re-ferred unclassified neuropathies yields improved diagnosis. AnnNeurol 1981;10:222–226.

Fig 3. PMP22 alterations identified in 5 families. The genotypes are shown below each individual tested. All patients were het-erozygous for their respective mutations. Black symbols indicate affected status. We confirmed the paternity of patients 1110 and1348 by analysis of polymorphic repeats; paternal DNA was not available to confirm the paternity of patient 1390.

198 Annals of Neurology Vol 51 No 2 February 2002

2. Ouvrier RA, Nicholson GA. Advances in the genetics of hered-itary hypertrophic neuropathy in childhood. Brain Dev 1995;17(suppl):31–38.

3. Skre H. Genetic and clinical aspects of Charcot-Marie-Tooth’sdisease. Clin Genet 1974;6:98–118.

4. Dyck PJ, Lambert EH. Lower motor and primary sensory neu-ron diseases with peroneal muscular atrophy. I. Neurologic, ge-netic, and electrophysiologic findings in hereditary polyneurop-athies. Arch Neurol 1968;18:603–618.

5. Dyck PJ, Lambert EH. Lower motor and primary sensory neu-ron diseases with peroneal muscular atrophy. II. Neurologic,genetic, and electrophysiologic findings in various neuronal de-generations. Arch Neurol 1968;18:619–625.

6. Boerkoel CF, Lupski JR. Hereditary motor and sensory neurop-athies. In: Rimoin DL, Pyeritzt R, Korf B, Connor M, eds.Emery and Rimoin’s principles and practice of medical genetics.4th ed. London: Harcourt, 2002. In press.

7. Lupski JR, Garcia CA. Charcot-Marie-Tooth peripheral neu-ropathies and related disorders. In: Scriver CR, Beaudet AL,Sly WS, et al, eds. The metabolic and molecular bases of in-herited disease. 8th ed. New York: McGraw-Hill, 2001:5759 –5788.

8. Nelis E, Haites N, Van Broeckhoven C. Mutations in the pe-ripheral myelin genes and associated genes in inherited periph-eral neuropathies. Hum Mutat 1999;13:11–28.

9. Zhao C, Takita J, Tanaka Y, et al. Charcot-Marie-Tooth dis-ease type 2A caused by mutation in a microtubule motorKIF1B!. Cell 2001;105:587–597.

10. Bomont P, Cavalier L, Blondeau F, et al. The gene encodinggigaxonin, a new member of the cytoskeletal BTB/kelch repeatfamily, is mutated in giant axonal neuropathy. Nat Genet 2000;26:370–374.

11. Lewis RA. The challenge of CMTX and connexin 32 muta-tions. Muscle Nerve 2000;23:147–149.

12. Scherer SS, Xu YT, Nelles E, et al. Connexin32-null mice de-velop demyelinating peripheral neuropathy. Glia 1998;24:8–20.

13. Anzini P, Neuberg DH, Schachner M, et al. Structural abnor-malities and deficient maintenance of peripheral nerve myelinin mice lacking the gap junction protein connexin 32. J Neu-rosci 1997;17:4545–4551.

14. Boerkoel CF, Takashima H, Stankiewicz P, et al. Periaxin mu-tations cause recessive Dejerine-Sottas neuropathy. Am J HumGenet 2001;68:325–333.

15. Bolino A, Lonie LJ, Zimmer M, et al. DHPLC analysis of themyotubularin-related 2 gene (MTMR2) in unrelated CMT pa-tients suggests a low frequency of mutation in inherited neu-ropathy. Neurogenetics 2001;3:107–109.

16. den Dunnen JT, Antonarakis SE. Mutation nomenclature ex-tensions and suggestions to describe complex mutations: a dis-cussion. Hum Mutat 2000;15:7–12.

17. Janssen EA, Kemp S, Hensels GW, et al. Connexin32 genemutations in X-linked dominant Charcot-Marie-Tooth disease(CMTX1). Hum Genet 1997;99:501–505.

18. Nelis E, Van Broeckhoven C, De Jonghe P, et al. Estimation ofthe mutation frequencies in Charcot-Marie-Tooth disease type1 and hereditary neuropathy with liability to pressure palsies: aEuropean collaborative study. Eur J Hum Genet 1996;4:25–33.

19. Panas M, Karadimas C, Avramopoulos D, et al. Central ner-vous system involvement in four patients with Charcot-Marie-Tooth disease with connexin 32 extracellular mutations. J Neu-rol Neurosurg Psychiatry 1998;65:947–948.

20. Senderek J, Bergmann C, Quasthoff S, et al. X-linked domi-nant Charcot-Marie-Tooth disease: nerve biopsies allow mor-phological evaluation and detection of connexin32 mutations(Arg15Trp, Arg22Gln). Acta Neuropathol (Berl) 1998;95:443–449.

21. Sorour E, Upadhyaya M. Mutation analysis in Charcot-Marie-Tooth disease type 1 (CMT1). Hum Mutat 1998;(suppl1):S242–247.

22. Wicklein EM, Orth U, Gal A, et al. Missense mutation(R15W) of the connexin32 gene in a family with X chromo-somal Charcot-Marie-Tooth neuropathy with only female fam-ily members affected. J Neurol Neurosurg Psychiatry 1997;63:379–381.

23. Ionasescu V, Ionasescu R, Searby C. Correlation between con-nexin 32 gene mutations and clinical phenotype in X-linkeddominant Charcot-Marie-Tooth neuropathy. Am J Med Genet1996;63:486–491.

24. Bort S, Nelis E, Timmerman V, et al. Mutational analysis ofthe MPZ, PMP22 and Cx32 genes in patients of Spanish an-cestry with Charcot-Marie-Tooth disease and hereditary neu-ropathy with liability to pressure palsies. Hum Genet 1997;99:746–754.

25. Fairweather N, Bell C, Cochrane S, et al. Mutations in theconnexin 32 gene in X-linked dominant Charcot-Marie-Toothdisease (CMTX1). Hum Mol Genet 1994;3:29–34.

26. Ressot C, Latour P, Blanquet-Grossard F, et al. X-linked dom-inant Charcot-Marie-Tooth neuropathy (CMTX): new muta-tions in the connexin32 gene. Hum Genet 1996;98:172–175.

27. Bone LJ, Deschenes SM, Balice-Gordon RJ, et al. Connexin32and X-linked Charcot-Marie-Tooth disease. Neurobiol Dis1997;4:221–230.

28. Haites NE, Nelis E, Broeckhoven CV. European CMTconsortium: 54th ENMC international workshop on genotype/phenotype correlations in Charcot-Marie-Tooth type 1 and he-reditary neuropathy with liability to pressure palsies, November28–30, 1997, Naarden, The Netherlands. Neuromuscul Disord1998;8:591–603.

29. Silander K, Meretoja P, Pihko H, et al. Screening for connexin32 mutations in Charcot-Marie-Tooth disease families withpossible X-linked inheritance. Hum Genet 1997;100:391–397.

30. Williams MM, Tyfield LA, Jardine P, et al. HMSN andHNPP. Laboratory service provision in the south west of En-gland—two years’ experience. Ann N Y Acad Sci 1999;883:500–503.

31. Fabrizi GM, Taioli F, Cavallaro T, et al. Focally folded myelinin Charcot-Marie-Tooth neuropathy type 1B with Ser49Leu inthe myelin protein zero. Acta Neuropathol (Berl) 2000;100:299–304.

32. Latour P, Blanquet F, Nelis E, et al. Mutations in the myelinprotein zero gene associated with Charcot-Marie-Tooth diseasetype 1B. Hum Mutat 1995;6:50–54.

33. Nelis E, Timmerman V, De Jonghe P, et al. Rapid screening ofmyelin genes in CMT1 patients by SSCP analysis: identifica-tion of new mutations and polymorphisms in the P0 gene.Hum Genet 1994;94:653–657.

34. Silander K, Meretoja P, Juvonen V, et al. Spectrum of muta-tions in Finnish patients with Charcot-Marie-Tooth disease andrelated neuropathies. Hum Mutat 1998;12:59–68.

35. Himoro M, Yoshikawa H, Matsui T, et al. New mutation ofthe myelin P0 gene in a pedigree of Charcot-Marie-Tooth neu-ropathy 1. Biochem Mol Biol Int 1993;31:169–173.

36. Mitsui Y, Matsui T, Nakamura Y, et al. A familial Charcot-Marie-Tooth disease type 1B (CMTD1B) manifesting a newmutation of myelin P0 gene. Rinsho Shinkeigaku 1994;34:1162–1167.

37. Ionasescu VV, Ionasescu R, Searby C, et al. Dejerine-Sottas dis-ease with de novo dominant point mutation of the PMP22gene. Neurology 1995;45:1766–1767.

38. Ionasescu VV, Searby C, Greenberg SA. Dejerine-Sottas diseasewith sensorineural hearing loss, nystagmus, and peripheral facialnerve weakness: de novo dominant point mutation of thePMP22 gene. J Med Genet 1996;33:1048–1049.

Boerkoel et al: CMT-Causing Mutations and Their Distribution 199

39. Marques W, Jr., Thomas PK, Sweeney MG, et al. Dejerine-Sottas neuropathy and PMP22 point mutations: a new basepair substitution and a possible “hot spot” on Ser72. Ann Neu-rol 1998;43:680–683.

40. Roa BB, Dyck PJ, Marks HG, et al. Dejerine-Sottas syndromeassociated with point mutation in the peripheral myelin protein22 (PMP22) gene. Nat Genet 1993;5:269–273.

41. Simonati A, Fabrizi GM, Pasquinelli A, et al. Congenital hy-pomyelination neuropathy with Ser72Leu substitution inPMP22. Neuromuscul Disord 1999;9:257–261.

42. Ionasescu VV, Searby CC, Ionasescu R, et al. Severe Charcot-Marie-Tooth neuropathy type 1A with 1-base pair deletion andframeshift mutation in the peripheral myelin protein 22 gene.Muscle Nerve 1997;20:1308–1310.

43. Ionasescu VV, Ionasescu R, Searby C. Screening of dominantlyinherited Charcot-Marie-Tooth neuropathies. Muscle Nerve1993;16:1232–1238.

44. Wise CA, Garcia CA, Davis SN, et al. Molecular analyses ofunrelated Charcot-Marie-Tooth (CMT) disease patients suggesta high frequency of the CMT1A duplication. Am J Hum Genet1993;53:856–863.

45. Antonarakis SE, Krawczak M, Cooper DN. The nature andmechanisms of human gene mutation. In: Scriver CR, BeaudetAL, Sly WS, et al, eds. The metabolic and molecular bases ofinherited disease. 8th ed. New York: McGraw-Hill, 2001:343–377.

46. Hoogendijk JE, Hensels GW, Gabreels-Festen AA, et al. De-novo mutation in hereditary motor and sensory neuropathytype I. Lancet 1992;339:1081–1082.

47. Scherer SS, Bone LJ, Deschenes SM, et al. The role of the gapjunction protein connexin32 in the pathogenesis of X-linkedCharcot-Marie-Tooth disease. Novartis Found Symp 1999;219:175–185.

48. Abrams CK, Oh S, Ri Y, et al. Mutations in connexin 32: themolecular and biophysical bases for the X- linked form ofCharcot-Marie-Tooth disease. Brain Res Brain Res Rev 2000;32:203–214.

49. Ressot C, Gomes D, Dautigny A, et al. Connexin32 mutationsassociated with X-linked Charcot-Marie-Tooth disease showtwo distinct behaviors: loss of function and altered gating prop-erties. J Neurosci 1998;18:4063–4075.

50. Abrams CK, Freidin MM, Verselis VK, et al. Functional alter-ations in gap junction channels formed by mutant forms ofconnexin 32: evidence for loss of function as a pathogenicmechanism in the X-linked form of Charcot-Marie-Tooth dis-ease. Brain Res 2001;900:9–25.

51. Omori Y, Mesnil M, Yamasaki H. Connexin 32 mutationsfrom X-linked Charcot-Marie-Tooth disease patients: func-tional defects and dominant negative effects. Mol Biol Cell1996;7:907–916.

52. Castro C, Gomez-Hernandez JM, Silander K, et al. Altered for-mation of hemichannels and gap junction channels caused byC- terminal connexin-32 mutations. J Neurosci 1999;19:3752–3760.

53. Oh S, Ri Y, Bennett MV, et al. Changes in permeability causedby connexin 32 mutations underlie X- linked Charcot-Marie-Tooth disease. Neuron 1997;19:927–938.

54. Dermietzel R, Hwang TK, Spray DS. The gap junction family:structure, function and chemistry. Anat Embryol 1990;182:517–528.

55. Bahr M, Andres F, Timmerman V, et al. Central visual, acous-tic, and motor pathway involvement in a Charcot-Marie-Toothfamily with an Asn205Ser mutation in the connexin 32 gene.J Neurol Neurosurg Psychiatry 1999;66:202–206.

56. Senderek J, Hermanns B, Bergmann C, et al. X-linked domi-nant Charcot-Marie-Tooth neuropathy: clinical, electrophysio-logical, and morphological phenotype in four families with dif-ferent connexin32 mutations(1). J Neurol Sci 1999;167:90–101.

57. Stojkovic T, Latour P, Vandenberghe A, et al. Sensorineu-ral deafness in X-linked Charcot-Marie-Tooth disease withconnexin 32 mutation (R142Q). Neurology 1999;52:1010 –1014.

58. Shapiro L, Doyle JP, Hensley P, et al. Crystal structure of theextracellular domain from P0, the major structural protein ofperipheral nerve myelin. Neuron 1996;17:435–449.

59. Ekici AB, Fuchs C, Nelis E, et al. An adhesion test systembased on Schneider cells to determine genotype- phenotype cor-relations for mutated P0 proteins. Genet Anal 1998;14:117–119.

60. Warner LE, Hiz MJ, Appel SH, et al. Clinical phenotypes ofdifferent MPZ (P0) mutations may include Charcot-Marie-Tooth type 1B, Dejerine-Sottas, and congenital hypomyelina-tion. Neuron 1996;17:451–460.

61. Chapon F, Latour P, Diraison P, et al. Axonal phenotype ofCharcot-Marie-Tooth disease associated with a mutation in themyelin protein zero gene. J Neurol Neurosurg Psychiatry 1999;66:779–782.

62. De Jonghe P, Timmerman V, Ceuterick C, et al. TheThr124Met mutation in the peripheral myelin protein zero(MPZ) gene is associated with a clinically distinct Charcot-Marie-Tooth phenotype. Brain 1999;122:281–290.

63. Lee MJ, Houlden H, Sweeney M, et al. Fequency of P0 genemutations in HMSN type II families and identification of aBritish family with a novel mutation. J Periph Nerv Syst 1999;4:295.

64. Marrosu MG, Vaccargiu S, Marrosu G, et al. Charcot-Marie-Tooth disease type 2 associated with mutation of the myelinprotein zero gene. Neurology 1998;50:1397–1401.

65. Misu K, Yoshihara T, Shikama Y, et al. An axonal form ofCharcot-Marie-Tooth disease showing distinctive features in as-sociation with mutations in the peripheral myelin protein zerogene (Thr124Met or Asp75Val). J Neurol Neurosurg Psychiatry2000;69:806–811.

66. Schiavon F, Rampazzo A, Merlini L, et al. Mutations of thesame sequence of the myelin P0 gene causing two different phe-notypes. Hum Mutat 1998;(suppl1):S217–219.

67. Senderek J, Hermanns B, Lehmann U, et al. Charcot-Marie-Tooth neuropathy type 2 and P0 point mutations: two novelamino acid substitutions (Asp61Gly; Tyr119Cys) and a pos-sible “hotspot” on Thr124Met. Brain Pathol 2000;10:235–248.

68. Mersiyanova IV, Ismailov SM, Polyakov AV, et al. Screeningfor mutations in the peripheral myelin genes PMP22, MPZ andCx32 (GJB1) in Russian Charcot-Marie-Tooth neuropathy pa-tients. Hum Mutat 2000;15:340–347.

69. Tyson J, Ellis D, Fairbrother U, et al. Hereditary demyelinatingneuropathy of infancy. A genetically complex syndrome. Brain1997;120:47–63.

70. Warner LE, Shohat M, Shorer Z, et al. Multiple de novo MPZ(P0) point mutations in a sporadic Dejerine-Sottas case. HumMutat 1997;10:21–24.

71. Baas F, Bienfait H, Vreijling J, et al. Two amino acid substi-tutions in the P0 gene of a HMSN1 patient. J Periph Nerv Syst1999;4:284.

72. Sinden R. DNA structure and function. San Diego: Academic,1994.

73. Hutchinson F. Chemical changes induced in DNA by ionizingradiation. Prog Nucleic Acid Res Mol Biol 1985;32:115–154.

200 Annals of Neurology Vol 51 No 2 February 2002

74. Ward JF. DNA damage produced by ionizing radiation inmammalian cells: identities, mechanisms of formation, andreparability. Prog Nucleic Acid Res Mol Biol 1988;35:95–125.

75. Snipes GJ, Suter U. Molecular anatomy and genetics of myelinproteins in the peripheral nervous system. J Anat 1995;186:483–494.

76. Pareek S, Notterpek L, Snipes GJ, et al. Neurons promote thetranslocation of peripheral myelin protein 22 into myelin.J Neurosci 1997;17:7754–7762.

77. D’Urso D, Ehrhardt P, Muller HW. Peripheral myelin protein22 and protein zero: a novel association in peripheral nervoussystem myelin. J Neurosci 1999;19:3396–3403.

78. Colby J, Nicholson R, Dickson KM, et al. PMP22 carrying thetrembler or trembler-J mutation is intracellularly retained inmyelinating Schwann cells. Neurobiol Dis 2000;7:561–573.

79. Naef R, Suter U. Impaired intracellular trafficking is a commondisease mechanism of PMP22 point mutations in peripheralneuropathies. Neurobiol Dis 1999;6:1–14.

80. Tobler AR, Notterpek L, Naef R, et al. Transport ofTrembler-J mutant peripheral myelin protein 22 is blocked inthe intermediate compartment and affects the transport of thewild-type protein by direct interaction. J Neurosci 1999;19:2027–2036.

81. Niemann S, Sereda MW, Suter U, et al. Uncoupling of myelinassembly and schwann cell differentiation by transgenic overex-pression of peripheral myelin protein 22. J Neurosci 2000;20:4120–4128.

82. Notterpek L, Ryan MC, Tobler AR, et al. PMP22 accumula-tion in aggresomes: implications for CMT1A pathology. Neu-robiol Dis 1999;6:450–460.

83. Notterpek L, Shooter EM, Snipes GJ. Upregulation of theendosomal-lysosomal pathway in the trembler-J neuropathy.J Neurosci 1997;17:4190–4200.

84. Hagedorn L, Suter U, Sommer L. P0 and PMP22 mark a mul-tipotent neural crest-derived cell type that displays communityeffects in response to TGF-beta family factors. Development1999;126:3781–3794.

85. Wulf P, Bernhardt RR, Suter U. Characterization of peripheralmyelin protein 22 in zebrafish (zPMP22) suggests an early rolein the development of the peripheral nervous system. J Neuro-sci Res 1999;57:467–478.

86. Kovach MJ, Lin JP, Boyadjiev S, et al. A unique point muta-tion in the PMP22 gene is associated with Charcot-Marie-Tooth disease and deafness. Am J Hum Genet 1999;64:1580–1593.

Boerkoel et al: CMT-Causing Mutations and Their Distribution 201