3tMed Genet 1997;34:43-49

Detection of the CMTlA/HNPP recombinationhotspot in unrelated patients of Europeandescent

Vincent Timmerman, Bemd Rautenstrauss, Lawrence T Reiter, Thearith Koeuth,Ann Lofgren, Thomas Liehr, Eva Nelis, Klaus D Bathke, Peter De Jonghe,Holger Grehl, Jean-Jacques Martin, James R Lupski, Christine Van Broeckhoven

Laboratory ofNeurogenetics,FlandersInteruniversityInstitute ofBiotechnology (VIB),Born-BungeFoundation (BBS),University of Antwerp(UIA), Department ofBiochemistry,Antwerpen, BelgiumV TimmermanA LofgrenE NelisP De JongheC Van Broeckhoven

Institute of HumanGenetics andDepartment ofNeurology, Friedrich-Alexander-University,Erlangen, GermanyB RautenstraussT LiehrK D BathkeH Grehl

Department ofMolecular and HumanGenetics, BaylorCollege of Medicine,Houston, Texas, USAL T ReiterT KoeuthJ R Lupski

Division of Neurology,University HospitalAntwerp (UZA),Antwerpen, BelgiumP De Jonghe

Laboratory ofNeuropathology, Born-Bunge Foundation,University of Antwerp(UIA), Department ofMedicine, Antwerpen,BelgiumJ-J Martin

Department ofPediatrics, HumanGenome Center,Baylor College ofMedicine, Houston,USAJ R Lupski

Correspondence to:Professor Van Broeckhoven,Laboratory ofNeurogenetics, University ofAntwerp (UIA), Departmentof Biochemistry,Universiteitsplein 1, B-2610Antwerpen, Belgium.Received 23 May 1996Revised version acceptedfor publication7 August 1996

AbstractCharcot-Marie-Tooth type 1 disease(CMT1) and hereditary neuropathy withliability to pressure palsies (HNPP) arecommon inherited disorders of the peri-pheral nervous system. The majority ofCMT1 patients have a 1.5 Mb tandemduplication (CMT1A) in chromosome17pll.2 while most HNPP patients have adeletion of the same 1.5 Mb region. TheCMT1A duplication and HNPP deletionare the reciprocal products of an unequalcrossing over event between misalignedflanking CMT1A-REP elements. We ana-lysed 162 unrelated CMT1A duplicationpatients and HNPP deletion patients from11 different countries for the presence ofa recombination hotspot in the CMT1A-REP sequences. A hotspot for unequalcrossing over between the misalignedflanking CMTIA-REP elements was ob-served through the detection ofnovel junc-tion fragments in 76.9% of 130 unrelatedCMT1A patients and in 71.9% of 32unrelated HNPP patients. This re-

combination hotspot was also detected ineight out of 10 de novo CMT1A duplicationand in two de novo HNPP deletionpatients. These data indicate that the hot-spot of unequal crossing over occurs inseveral populations independently of eth-nic background and is directly involved inthe pathogenesis of CMT1A and HNPP.We conclude that the detection ofjunctionfragments from the CMTIA-REP elementon Southern blot analysis is a simple andreliable DNA diagnostic tool for the iden-tification of the CMT1A duplication andHNPP deletion in most patients.(J Med Genet 1997;34:43-49)

Keywords: CMT; HNPP; recombination hotspot.

Charcot-Marie-Tooth neuropathy type 1(CMT1), or hereditary motor and sensory

neuropathy type I (HMSN I), is the mostcommon autosomal dominant disorder of theperipheral nervous system. CMT1 is char-acterised by weakness and atrophy ofdistal limbmuscles, diminished or absent deep tendonreflexes, pes cavus deformity, and severelyslowed motor and sensory nerve conductionvelocities (NCVs). De- and remyelination withonion bulb formations are observed on peri-

pheral nerve biopsies.' Hereditary neuropathywith liability to pressure palsies (HNPP) is anautosomal dominantly inherited demyelinatingneuropathy of the peripheral nervous systemcharacterised by sausage-like formations of themyelin sheaths known as tomaculae. The symp-toms in HNPP can often be correlated directlywith minor trauma or prolonged compressionof the peripheral nerves, resulting in nervepalsies for a few days to several weeks. NCVsare slowed and sensory nerve action potentials(SNAPs) are reduced or absent in damagednerves.2

Molecular genetic studies have identified theunderlying gene defects in CMT1 and HNPP.3CMT1A, the major form of CMT1, is causedby a tandem duplication of a 1.5 Mb region inchromosome 17pll.2.7 HNPP is caused bythe reciprocal deletion ofthis 1.5 Mb region.8-"In a recent European collaborative study it wasfound that 70.7% of unrelated CMT1 patientshave the CMT1A duplication and that 84.0%of unrelated HNPP patients have the HNPPdeletion.'2 The peripheral myelin protein 22gene (PMP22) is located within the 1.5 Mbregion duplicated in CMT1A and deleted inHNPP.813'6 Point mutations in the PMP22gene have been found in both CMT1A patientswithout duplicationl-20 and in an HNPP patientwithout deletion2' implicating the PMP22 genein the pathogenesis of both disorders.Unequal crossing over during meiosis has

been proposed as the mechanism of both theCMT1A duplication and HNPP deletion.48This unequal crossing over event is probablythe result of the misalignment of two low copyrepeat elements of 30 kb (CMTIA-REP)which flank the 1.5 Mb region duplicated inCMT1A7 and deleted in HNPP.8 Three copiesof the CMTIA-REP sequence are located onthe CMT1A duplication chromosome, whileone copy is detected on the HNPP deletionchromosome.9 Currently, a cloned CMT1A-REP sequence is used in the DNA diagnosisof CMT1A and HNPP patients using pulsedfield gel electrophoresis (PFGE).9 11 22 23

Recently a recombination hotspot was iden-tified within the CMTlA-REP elements in acohort of unrelated CMT1A and HNPPpatients ofpredominantly North American des-cent.24 The region encompassing the hotspotof recombination was 98% identical in theproximal CMTIA-REP element and the distalCMT1A-REP element.24 This remarkably act-ive hotspot of homologous recombination was

43

group.bmj.com on July 15, 2011 - Published by jmg.bmj.comDownloaded from

Timmerman et al

identified as a 1.7 kb EcoRINszIl junction frag-ment in 75% of the CMT1A duplicationpatients and as a 7.8 kb EcoRIlEcoRI junctionfragment in 84% ofthe HNPP deletion patientsexamined. These junction fragments resultingfrom the unequal crossing over event betweenmisaligned CMTlA-REP elements are stablyinherited with the disease phenotypes ofCMT1and HNPP and are observed in de novo du-plication and deletion patients.24 Some patientswithout these junction fragments appeared tohave undergone a strand exchange event out-side the 1.7 kb hotspot region but within theboundaries of the 7.8 kb EcoRI fragment fromthe proximal CMT1A-REP, as indicated bydosage analysis (unpublished observation).Sequence analysis of both the proximal and thedistal CMT1A-REP clones showed a marinertransposon-like element (MITE) near the hot-spot ofunequal recombination, which was pro-posed to be involved in the initiation of strandexchange events between misaligned CMT1A-REPs through the generation of double strandbreaks by a mariner transposase.24

In this study we estimated the frequency ofthis hotspot of unequal crossing over in anindependent cohort ofpatients. A large numberof unrelated CMT1A and HNPP patients ofdifferent European origin and from two othercountries were screened for the presence orabsence of the hotspot leading to CMT1A orHNPP. Also the frequency and origin of thishotspot was examined in a set of de novoCMT1A and HNPP patients.

Materials and methodsPATIENT SAMPLES AND DUPLICATION/DELETIONSTATUSAll 162 unrelated CMT1 and HNPP patientswere screened for the presence of the CMT1Aduplication or HNPP deletion according tomethods previously described.45152526 Patients'DNA samples were used for research (CMTnumbers) or diagnostic purposes (PN num-bers). The presence and size of the CMT1Aduplication or HNPP deletion was determinedwith restriction fragment length polymorphism(RFLP) probes pVAW409R3a (D17S122) and

Table 1 Overview of the EcoRIlSacI Southern blot results in 162 unrelated CMTIAduplication (A) and HNPP deletion (B) patients of 11 different ethnic originsCountry 1 2 3

2:3:3:2 dosage 3.2 kb junction 3:2:2:3 dosage

(A) CMT1ABelgium (90) 6 66 18Germany (22) 1 20 1Israel (3) 0 2 1Lebanon (1) 0 0 1Croatia (2) 0 2 0Mali (1) 0 1 0Poland (1) 0 0 1Switzerland (1) 0 1 0Tjechia (2) 0 2 0The Netherlands (5) 0 4 1Turkey (2) 0 2 0Total 7 100 23

2:1:1:2 dosage 7.8 kb junction 1:2:2:1 dosage(B) HNPPBelgium (23) 2 16 5Germany (9) 1 7 1Total 3 23 6

1, 2, and 3 refer to the sites ofunequal crossing over as shown in fig 2. Values between parenthesesrepresent the total number of CMT1A and HNPP patients from each country.

pEW401HE (D17S61) according to Ra-eymaekers et al.45 In uninformative cases,polymorphic (CA)n repeats RM1 1-GT(D17S122), Mfd41 (D17S261), andAFM191xh12 (D17S921) were analysed withan automated DNA sequencer (Applied Bio-systems Inc, Foster City, USA) according toNavon et al25 and Timmerman et al.23 Pulsedfield gel electrophoresis (PFGE) was used inthe remaining uninformative patients in orderto detect specific CMT1A or HNPP junctionfragments with probes pVAW409R3a27 andpNEA102.23 In some patients, fluorescence insitu hybridisation (FISH) analysis was used toprove the presence of the CMT1A duplicationor HNPP deletion.2628The clinical, electrophysiological, histo-

pathological, and genetic data of the de novoCMT1A duplication families CMT-G, PN-9,PN-20, PN-36, PN-39, and PN-42, as well asde novo HNPP deletion family PN-97, havebeen described previously.4232930 In this studywe included the CMT1 families PN-187, PN-242, PN-412, and CMT-ER1, and the HNPPfamily PN-340, in which a de novo CMT1Aduplication and a de novo HNPP deletion wereidentified using previously described moleculargenetic detection methods.2326283'

DETECTION OF RECOMBINANT JUNCTIONFRAGMENTSTo detect novel junction fragments from therecombinant CMTlA-REPs, genomic DNAwas digested with EcoRI, SacI, and Nsil ac-cording to the manufacturer's instructions(New England Biolabs and Boehringer Mann-heim) in double or triple restriction digestions.The DNA probe used to detect the 3.2 kb and7.8 kb EcoRI/SacI specific junction fragmentsin Southern blot hybridisations was pLR7.8,which is a 7.8 kb EcoRI fragment from proximalCMT1A-REP cosmid c56A5. The DNA probeused to detect the 1.7 kb EcoRI/Nsil junctionfragment in EcoRI/SacI/Nszl digests waspLR6.0, which is a 6.0 kb EcoRI fragment fromdistal CMTlA-REP cosmid c112C10. BothEcoRI DNA fragments were cloned into theEcoRI site of plasmid vector pTZ19R.24 Thelabelled probes were prehybridised with humanplacental DNA and hybridised according tostandard procedures previously described.24

ResultsWe selected 130 unrelated CMT1 patients witha 1.5 Mb CMT1A duplication and 32 un-related HNPP patients with the reciprocal1.5Mb deletion in 17pll.2. The CMT1Apatients were referred to our laboratories inAntwerpen (Belgium) and Erlangen (Ger-many) for research purposes orDNA diagnosis,and are from a variety of countries includingBelgium, Germany, Israel, Croatia, Lebanon,Mali, Poland, Switzerland, The Netherlands,Tjechia, and Turkey (table 1A). The HNPPpatients are from Belgium and Germany (table1B). All patients were initially screened for thepresence of a recombination hotspot in theCMTlA-REP sequences using EcoRlISacI

44

group.bmj.com on July 15, 2011 - Published by jmg.bmj.comDownloaded from

CMTIAIHNPP recombination hotspot

A B

'e r-N 1C D CM (.0 r1 1 CD Cd CD

COCD X CD O C N CN CN m m -t- - - - - - C C N C CNCN CNCN

r <= Lo

(lt 2n A OZ0 ON rt> L. m 00 LCS r(- (- CD C 00 r- -Jd,q Q0 CY It *--7 LO r- 1- 00 CN C

>- >- -0 CY C NC CN CN -N C',4 OSc kb7.8 HNPP jct

6.0

5.0

3.2 -_CMT jct

2.8

1 .8

# 2 1 2 3 2 2 3 2 3 2 2 2 2 3 2 1 # 3 2 2 2 2 2 3 3 3 2 2 2 3 2 2 2

Figure 1 EcoRIlSacI Southern blot of unrelated CMTIA (A) and unrelated HNPP (B) patients hybridised with CMTIA-REP probe pLR7.8.Lanes YC49H7 and YC225A3 contain YAC DNA. The 5.0 kb and 2.8kb EcoRIfragments are unique to YAC YC49H7 which spans the proximalCMT1A-REP? The 6.0kb and 1.8kb EcoRIfragments are unique to YAC YC225A3 which spans the distal CMT1A-REP7 Lane 1034 is a norrnalcontrol without CMTIA duplication or HNPP deletion, and lane 1027 is a CMTI patient with a 1.5Mb CMT1A duplication.24 The numbers 1, 2,and 3 below the lanes refer to fig 2 and indicate the sites of strand exchange as determined by the detection of novel junction fragments (2) or dosage ofbands (1 or 3).

double digests and Southern blot hybridisationwith the CMT1A-REP probe pLR7.8 (fig 1).A 3.2 kb EcoRI/SacI junction fragment was

identified in 100 out of 130 CMT1A du-plication patients (76.9%) and a 7.8 kb EcoRI/EcoRI junction fragment was observed in 23out of 32 HNPP deletion patients (71.9%).These data indicated that there is a frequentrecombination site between the EcoRI site uniqueto the distal CMT1A-REP and the Sacd siteunique to the proximal CMTIA-REP. This un-equal crossing over event between the CMT1A-REPs resulting in the CMT1A duplication or theHNPP deletion is illustrated in fig 2.

In the remaining CMT1A and HNPPpatients no junction fragments were observedbut dosage differences in the 6.0 kb, 5.0 kb,2.8 kb, and 1.8 kb fragments could be detected(fig 1). These dosage differences in the hy-bridisation signals are most likely because ofunequal crossing over events just outside thehotspot region (fig 2, 1 and 3). The hy-bridisation pattern 2:3:3:2 (1) was observed inseven CMT1A duplication patients (5.3%),and the pattern 2:1:1:2 (1) was found in threeHNPP deletion patients (9.1%), indicating thatin these patients the unequal crossing overevent occurred in or proximal to region 1. Adosage pattern of 3:2:2:3 (3) is found in 23CMT1A duplication patients (17.9%) and apattern of 1:2:2:1 (3) is observed in six HNPPdeletion patients (18.2%), indicating a strandexchange event in or distal to region 3. Anoverview of the EcoRI/SacI junction fragmentsand dosages detected in the 162 unrelated

CMT1A duplication and HNPP deletionpatients from 11 different countries are shownin table 1. Some digestions and hybridisationswere performed twice in order to confirm thereproducibility ofthe results. The data obtainedby visual inspection of the autoradiographswere consistent with the results of the den-sitometric analysis (data not shown).DNA sequence analysis of the 7.8 kb region

containing the junction fragment boundariesin regions 1, 2, and 3 (fig 2) showed restrictionsites which are unique to either the proximalor distal CMTlA-REP.24 An NsiI site uniqueto the proximal CMT1A-REP was used byReiter et al4 to refine the region of strandexchange further. Southern blot analysis ofEcoRI/SacIJNsil triple digested DNA of un-related CMT1A duplication patients hy-bridised with probe pLR6.0 showed a 1.7 kbEcoRI/Nsil junction fragment in all CMT1Apatients with the 3.2 kb EcoRI/SacI junctionfragment. Also, a 7.8 kb EcoRIlEcoRI junctionfragment is observed by Southern analysis withprobe pLR6.0 hybridised to EcoRI/SacI/Nsiltriple digested DNA of the unrelated HNPPpatients (data not shown). This observationindicates that the region of strand exchange inthe majority ofCMT1A duplication andHNPPdeletion patients of non-North American des-cent is located between the EcoRI site uniqueto the distal CMTlA-REP and the NsiI siteunique to the proximal CMTlA-REP. A rel-ative risk calculation indicates that the strandexchange event in both CMT1A and HNPPpatients occurs within the 1.7 kb hotspot pref-

r <E= Lo)O CD.r± C14

45

group.bmj.com on July 15, 2011 - Published by jmg.bmj.comDownloaded from

Timmerman et al

CMTlA-REP EcoRi EcoRI

[::OoD/~I --I7.8 kb \

30kb

EcoRIso 5.0

6.0 kb >EcoRI

I *Ar _

:\ :.:\.

Nsil Sacl

-a 3.5kb -

/0

V/ -CMTlAREPEcoRI

8 kb _._ <60 kb

30 kb

GO X.

CMT1A HNPP

7.8 kb

6.0 kb - -

CMT1A HNPP

jctCMT1A HNPP

5.0 kb

3.2 kbjct

2.8 kb

1.8 kb ~m-

Figure 2 Crossover junctions in CMTIA duplication and HNPP deletion patients. Shown above are the proximal anddistal - 30 kb CMTIA-REPs illustrating the misalignment of the 7.8 kb region with the distal CMT1A-REP (top) withthe 7.8 kb region from the proximal CMTIA-REP (below). Dashed lines represent strand exchange events leading toduplication (CMT1A) and deletion (HNPP). Three potential crossover regions are illustrated (1, 2, and 3). Themajority of CMTIA and HNPP patients examined undergo a strand exchange event between the EcoPi site unique tothe distal CMT1A-REP and the NsiI site unique to the proximal CMT1A-REP (crossover 2). At the bottom of thefigure are the predicted Southern blot patterns and allele dosages for CMTIA duplication and HNPP deletion patientswhere the genomic DNA was digested with Sad and EcoRI. The 3.2 kb and 7.8 kb EcoRIlSacIjunction fragments aswell as other restriction fragments originating from the normal and recombinant CMT1A-REPs detected by probe pLR7. 8are indicated. The dosages indicated for the 6.0 kb:5. 0 kb:2.8 kb:1.8 kb alleles are (1) CMTIA (2:3:3:2) and HNPP(2:1:1:2), (2) CMTIA (2:2:3:3) and HNPP (1:1:1:1), and (3) CMTIA (3:2:2:3) and HNPP (1:2:2:1).

erentially 52:1 over the surrounding -28 kbof sequence homology between the repeats.Furthermore, this hotspot of homologous re-

U U U

combination was detected in unrelatedCMT1A patients from 11 different countries(data not shown).

U

~~~~rn*~~~~~:

2.

Figure 3 EcoRIlSacI Southern blot offive de novo CMTIA (A) and two HNPP (B) families hybridised with CMTlA-REP probe pLR7.8. Onlythe unaffected parents and their affected child of each family are shown. Filled symbol== affected child. The hybridisation data of CMTIA families PN-

20, PN-36, PN-42, PN-412, and CMTER1 are not shown. Note the absence of a 3.2 kb junction fragment in CMTIA de novo patient PN-9. This

patient has a dosage pattern of 3:2:2:3 (6.0 kb:5. 0 kb:2.8 kb:1.8 kb) indicative of a strand exchange event at or proximal to 3 in fig 2.

i..! ,

60

1. t -4

46

.M

group.bmj.com on July 15, 2011 - Published by jmg.bmj.comDownloaded from

CMTIAIHNPP recombination hotspot

BA

r <IL LCF) CNIt Cl

sCMTG PN20 PN39 PN187 PN242 PN412cl- --fir

_LD-- > - - .- - . >

kb

4.0

30 *

20 0

CMT ct .116 t I I .- ..4

*

PN97

UkbD

7.8-4 HNPP jct

, -,!.!xe * 6 1

4.0

3.0

2.0

1.6

Figure 4 EcoRI/SacI/NsiI Southern blot of six de novo CMTIA (A) and one HNPP (B) families hybridised withCMTIA-REP probe pLR6. 0. Lanes YC49H7 and YC225A3 contain YAC DNA. The 5. 0 kb, 3.8 kb, 2.8 kb, and1. 6 kb fragments are unique to YAC YC49H7 which spans the proximal CMTJA-REP The 6. 0 kb EcoRIfragment isunique to YAC YC225A3 which spans the distal CMTJA-REPJ7 The hybridisation data of CMTIA families PN-9,PN-36, PN-42, and CMTER1, and ofHNPPfamily PN-340 are not shown. Note the absence of a 1. 7kb junctionfragment in CMT1A de novo patient PN-412.

In this study we also analysed 10 de novoCMT1A and two de novo HNPP families forthe presence of the recombination hotspot. Inall cases the unequal crossing over event wasof paternal origin, as indicated by RFLP and(CA)n repeat analysis ofDNA markers locatedwithin the CMTIA/HNPP region (data notshown). Using the CMTlA-REP probepLR7.8 we observed the 3.2 kb EcoRI/SacIjunction fragment on EcoRI/SacI Southernblots in 8/10 de novo CMT1A duplicationpatients and the 7.8 kb EcoRJJEcoRI junctionfragment in two de novo HNPP deletionpatients. The EcoRI/SacI hybridisation pat-terns of five CMT1A and two HNPP de novofamilies are shown in fig 3. The CMT1A 3:2:2:3 (3) dosage pattern was observed in theaffected daughter of family PN-9 (fig 3A). The1.7 kb EcoRIINsil junction fragment was foundin 8/9 de novo CMT1A duplication patientsand the 7.8 kb EcoRI/EcoRI junction fragmentin two de novo HNPP deletion patients. TheEcoRVISacIJNsil hybridisation patterns of sixCMT1A and one HNPP de novo families areshown in fig 4. These data support the hy-pothesis that a 1.7 kb EcoRIINsil recombinationhotspot for unequal crossing over results in theCMT1A duplication and HNPP deletion.

DiscussionPrevious studies suggested that unequal cross-ing over between homologous chromosomeswas a mechanism for the generation of theCMT1A duplication and HNPP deletion.48Physical mapping studies showed the presenceof two low copy repeat elements of at least30 kb (CMT1A-REP) which flank the 1.5 Mbregion duplicated in CMT1A patients7 anddeleted in HNPP patients.9 The CMTIA-REPrepeats are present in the CMT1A duplication

and HNPP deletion specific junction fragmentsdetected by Southern blot and PFGEanalysis."23 The analysis of dosage patternsfor CMTIA-REP specific junction fragmentssuggested a preferential region for crossing overevents between CMT1A-REPs.9-1 l 23 32-34

Reiter et at24 showed the presence of a re-combination hotspot associated with the re-ciprocal recombination which leads to DNArearrangements resulting in both CMT1A andHNPP. The recombination hotspot was definedby identifying rearrangement specific junctionfragments which reflect crossing over eventsoccurring between restriction sites that areunique to either the proximal or distal CMT1A-REP sequence. The hotspot was determinedby scoring the number ofpeople with crossoverevents in the small 1.7 kb region flanked byEcoRI and Nsil in the recombinant CMT1A-REP and comparing these to the number ofcrossovers in the remaining -28 kb of theCMTlA-REP region. These experiments in-dicated that there is a 53:1 preference for strandexchange in the small region bounded by theunique EcoRI and NszI sites.24 We observedthat 76.9% of CMT1A duplication patientshad the 1.7 kb EcoRI/Nsil junction fragment.CMT1A duplication patients without thisnovel junction fragment had dosage differencessimilar to those predicted if the crossing overevent had occurred proximal to the EcoRI sitein the distal CMTlA-REP or distal to the NszIsite in the proximal CMTlA-REP. In HNPPdeletion patients, 71.9% had the 7.8 kb EcoRI/EcoRI junction fragment indicative of a strandexchange event within this 1.7 kb region. Inthis study, the relative risk for strand exchangein the 1.7 kb EcoRJINsil hotspot region was 52:1, confirming the findings of Reiter et a124 andindicating that in most cases the site of strandexchange is between the EcoRI site unique to

47

group.bmj.com on July 15, 2011 - Published by jmg.bmj.comDownloaded from

Timmerman et al

the distal CMTlA-REP and the Nsil siteunique to the proximal CMTlA-REP. The1.7 kb hotspot of recombination in theCMTIA-REP elements leading to CMT1Aduplication and HNPP deletion was found ina large set of unrelated CMT1A and HNPPpatients of different ethnic origin. Thus, thehotspot responsible for the unequal crossingover event in the 17pll.2 region occurs inseveral populations independent of ethnicbackground and is an intrinsic structural prop-erty of the genome and not a polymorphismspecific to a particular genetic background.The de novo CMT1A duplications and

HNPP deletions occurred in most caseson the paternally inherited chromosome,indicating that the unequal crossing overevent occurs preferentially during malegametogenesis. 48293035-38 De novo cases ofCMT1A and HNPP of maternal origin arerare.323841 In this study, we analysed the DNAsamples from several sporadic CMT1A andHNPP cases with paternal origin of the du-plication or deletion. The 1.7 kb EcoRI/Nsiljunction fragment was found in 8/9 de novoCMT1A patients and the 7.8 kb EcoRIlEcoRIfragment was observed in two de novo HNPPdeletion patients. These junction fragmentswere not detected in the parents of the affectedsubjects indicating that there is a correlationbetween the CMTIA/HNPP phenotypes andthe appearance of the junction fragments. Thedetection of these novel junction fragments inde novo CMT1A cases rules out the possiblilitythat they are merely polymorphisms which se-gregate with the disease phenotype in selectedpopulations. In fact, the de novo appearanceof these junction fragments in affected patientsof both North American and European decentsuggests that the homologous recombinationevent between misaligned CMTIA-REPs iscaused by an intrinsic genomic charateristicwhich is not susceptible to the influence ofvarious genetic backgrounds.Dosage patterns were previously observed

with CMT1A-REP probe pNEA102 on EcoRISouthern blots of unrelated CMT1A duplica-tion and HNPP deletion patients.9-" 23 32-34 Thedetection of junction fragments from the re-combinant CMT1A-REP in CMT1A andHNPP patients on EcoRIlSacI or EcoRIlSacI/Nsil Southern blot analysis is more readilyinterpreted and gives more consistent resultsthan dosage analysis of hybridisation signalsdetected with pNEA102.42 Restriction frag-ment patterns and allele dosages for CMT1Aand HNPP patients who undergo a strandexchange outside the 1.7 kb hotspot regioncould be used to predict whether the recom-bination junctions in these patients are locatedproximal (crossover 1 or proximal to the left-most EcoRI site) or distal (crossover 3 or distalto the rightmost EcoRI site) to the hotspot (fig2). The development of a PCR method for thedetection of CMT1A and HNPP recombinantjunctions would be quite useful for fast andaccurate mutation screening of most CMT1and HNPP patients. However, some CMT1Aand HNPP patients apparently undergo strandexchange events just outside the hotspot region

and other rare patients have been reported withlarger or smaller duplications or deletions.29314These patients will most likely have otherbreakpoints located within or outside the1.5 Mb CMT1A/HNPP region and will requirealternate detection methods.

The first two authors contributed equally to the preparation ofthis manuscript. We are grateful to the clinicians who providedus with DNA material as well as the patients and their relativesfor their kind cooperation in our research project. We wouldlike to thank Els De Vriendt and Arif Ekici for their excellenttechnical assistance. Part of this work was performed at BaylorCollege of Medicine by VT and BR. BR wishes to thank ProfDr B Neuendorfer and Prof Dr R-A Pfeiffer for their continuedsupport. This work was supported in part by grants from theBelgian National Fund for Scientific Research (NFRS) and aspecial research fund of the University of Antwerp to CVB,from the "Deutsche Forschungsgemeinschaft" (DFG) and the"Deutsche Gesellschaft Muskelkranke" (DGM) to BR, theBoehringer Ingelheim Funds and the Herbert-Quandt-Founda-tion of the VARTA AG Inc to TL, and the United StatesNational Institute of Health to JRL. VT and EN are researchassistants of the NFSR, Belgium.

1 Dyck PJ, Chance PE, Lebo RV, Carney JA. Hereditarymotor and sensory neuropathies. In: Dyck PJ, ThomasPK, Griffin JW, Low PA, Poduslo JF, eds. Peripheralneuropathy. Philadelphia: Saunders, 1993:1094-136.

2 Windebank AJ. Inherited recurrent focal neuropathies. In:Dyck PJ, Thomas PK, Griffin JW, Low PA, Poduslo JF,eds. Peripheral neuropathy. Philadelphia: Saunders, 1993:1137-148.

3 Harding AE. From the syndrome of Charcot-Marie-Toothto disorders of peripheral myelin proteins. Brain 1995;118:809-18.

4 Raeymaekers P, Timmerman V, Nelis E, et al and theHMSN Collaborative Research Group. Duplication inchromosome 17pll.2 in Charcot-Marie-Tooth neur-opathytype la (CMT la). NeuromuscDisord 1991;1:93-7.

5 Raeymaekers P, Timmerman V, Nelis E, et al and the HMSNCollaborative Research Group. Estimation of the size ofthe chromosome 17pll.2 duplication in Charcot-Marie-Tooth neuropathy type la (CMT I a). _7 Med Genet 1992;29:5-11.

6 Lupski JR, Montes de Oca-Luna R, Slaugenhaupt S, et al.DNA duplication associated with Charcot-Marie-Toothdisease type la. Cell 1991;66:219-32.

7 Pentao L, Wise CA, Chinault AC, Patel PI, Lupski JR.Charcot-Marie-Tooth type IA duplication appears to arisefrom recombination at repeat sequences flanking the1.5 Mb monomere unit. Nat Genet 1992;2:292-300.

8 Chance PF, Alderson MK, Leppig KA, et al. DNA deletionassociated with hereditary neuropathy with liability topressure palsies. Cell 1993;72: 143-51.

9 Chance PF, Abbas N, Lensch NW, et al. Two autosomaldominant neuropathies result from reciprocal DNA du-plication/deletion of a region on chromosome 17. HumMol Genet 1994;3:223-8.

10 LeGuern E, Sturz F, Gugenheim M, et al. Detection ofdeletion within 17pll.2 in 7 French families with her-editary neuropathy with liability to pressure palsies(HNPP). Cytogenet Cell Genet 1994;65:261-4.

11 Lorenzetti D, Pareyson D, Sghirlanzoni A, et al. A 15 Mbdeletion in I7pll.2-pl2 is frequently observed in Italianfamilies with hereditary neuropathy with liability to pres-sure palsies. AmJtHum Genet 1995;56:91-8.

12 Nelis E, Van Broeckhoven C. Estimation of the mutationfrequencies in Charcot-Marie-Tooth disease type 1 andhereditary neuropathy with liability to pressure palsies: aEuropean collaborative study. Eur7JHum Genet 1996;4:25-33.

13 Matsunami N, Smith B, Ballard L, et al. Peripheral myelinprotein-22 gene maps in the duplication in chromosome17pll.2 associated with Charcot-Marie-Tooth IA. NatGenet 1992;1:176-9.

14 Patel PI, Roa BB, Welcher AA, et al. The gene for theperipheral myelin protein PMP-22 is a candidate for Ch-arcot-Marie-Tooth disease type IA. Nat Genet 1992;1:159-65.

15 Timmerman V, Nelis E, Van Hul W, et al. The peripheralmyelin protein gene PMP-22 is contained within theCharcot-Marie-Tooth disease type 1A duplication. NatGenet 1992;1:171-5.

16 Valentijn LI, Bolhuis PA, Zorn I, et al. The peripheral myelingene PMP-22/GAS-3 is duplicated in Charcot-Marie-Tooth disease type IA. Nat Genet 1992;1:166-70.

17 Valentijn LJ, Baas F, Wolterman RA, et al. Identical pointmutations of PMP-22 in Trembler-J mouse and Charcot-Marie-Tooth disease type IA. Nat Genet 1992;2:288-91.

18 Roa BB, Garcia CA, Suter U, et al. Charcot-Marie-Toothtype IA: association with a spontaneous point mutationin the PMP22 gene. NEngl3rMed 1993;329:96-101.

19 Roa, BB, Garcia CA, Pentao L, et al. Evidence for a recessivePMP-22 point mutation in Charcot-Marie-Tooth diseasetype 1A. Nat Genet 1993;5:189-94.

20 Nelis E, Timmerman V, De Jonghe P, Van Broeckhoven C.Identification of a 5' splice site mutation in the PMP-22gene in autosomal dominant Charcot-Marie-Tooth diseasetype 1. Hum Mol Genet 1994;3:515-16.

48

group.bmj.com on July 15, 2011 - Published by jmg.bmj.comDownloaded from

CMTIAIHNPP recombination hotspot

21 Nicholson GA, Valentijn LJ, Cherryson AK, et al. Aframe shift mutation in the PMP22 gene in hereditaryneuropathy with liability to pressure palsies. Nrat Genet1994;6:263-6.

22 Roa BB, Ananth U, Garcia CA, Lupski JR. Moleculardiagnosis of CMT1A and HNPP. Labmed Int 1995;12:22-4.

23 Timmerman V, Lofgren A, LeGuem E, et al. Moleculargenetic analysis of the 17pl1.2 region in patients withhereditary neuropathy with liability to pressure palsies(HNPP). Huni Genet 1996;97:26-34.

24 Reiter LT, Murakami T, Koeuth T, et al. A recombinationhotspot responsible for two inherited peripheral neuro-pathies is located near a mariner transposon-like element.Nat Genet 1996;12:288-97.

25 Navon R, Timmerman V, Lofgren A, et al. Prenatal diagnosisof Charcot-Marie-Tooth disease type 1A (CMT1A) usingmolecular genetic techniques. Prenat Diagn 1995;15:633-40.

26 Liehr T, Thoma K, Kammler K, et al. Direct preparationof uncultured EDTA-treated or heparinized blood forinterphase FISH analysis. Appl Cytogenet 1995;6: 185-8.

27 den Dunnen JT, Liang P, Van Ommen GJB, Van Broeck-hoven C. Mutation detection and diagnosis using PFGE.In: Monaco AP, ed. Pulsedfield gel electrophoresis: a practicalapproach. Oxford: Oxford University Press, 1995.

28 Liehr T, Rautenstrauss B, Grehl H, et al. Mosaicism for theCharcot-Marie-Tooth disease type IA duplication sug-gests somatic reversion Hum Genet 1996,98:22-8.

29 Palau F, Lofgren A, De Jonghe P, et al. Origin of the denovo duplication in Charcot-Marie-Tooth disease type1A: unequal nonsister chromatid exchange during sper-matogenesis. Hum Mol Genet 1993;2:2031-5.

30 Verhalle D, Lofgren A, Nelis E, et al. Deletion in theCMT1A locus on chromosome 17pl 1.2 in hereditaryneuropathy with liability to pressure palsies. Ann Neurol1994;35:704-8.

31 Ionasescu VV, Ionasescu R, Searby C, Barker DF. Charcot-Marie-Tooth neuropathy type IA with both duplicationand non-duplication. Hum Mol Genet 1993;2:405-10.

32 LeGuern E, Gouider R, Ravise N, et al. A de novo case ofhereditary neuropathy with liability to pressure palsies

(HNPP) of matemal origin: a new mechanism for deletionin 17p ll .2? Hunm Mol Genet 1996;5:103-6.

33 Tyson J, Malcolm S, Thomas PK, Harding AE. Deletionsof chromosome 17p II.2 in multifocal neuropathies. AnnNeurol 1996;39: 180-6.

34 Kiyosawa H, Lensch MW, Chance PF. Analysis of theCMTlA-REP repeat: mapping crossover breakpoints inCMT1A and HNPP. Hunm Mol Genet 1996;4:2327-34.

35 Wise CA, Garcia CA, Davis SN, et al. Molecular analyses ofunrelated Charcot-Marie-Tooth (CMT) disease patientssuggest a high frequency of the CMT1A duplication. AnmJ Hum Genet 1993;53:853-63.

36 Mariman ECM, Gabreels-Festen AAWM, et al. Prevalenceof the 15 Mb 17p deletion in families with hereditaryneuropathy with liability to pressure palsies. Ann Neurol1994;36:650-5.

37 Hertz JM, Borglum AD, Brandt CA, Flint T, Bisgaard C.Charcot-Marie-Tooth disease type IA: the parental originof a de novo l7pl I.2-p12 duplication. Clin Genet 1994;46:291-4.

38 Blair IP, Nash J, Gordon MJ, Nicholson GA. Prevalenceand origin of de novo duplications in Charcot-Marie-Tooth disease type IA: first report of a de novo duplicationwith a maternal origin. AmJHum Genet 1996;58:472-6.

39 Reisecker R, Leblhuber F, Lexner R, Radner G, RosenkranzW, Wagner K. A sporadic form of hereditary neuropathywith liability to pressure palsies: clinical electrodiagnostic,and molecular genetic findings. Neurology 1994;44:753-5.

40 Mancardi GL, Uccelli A, Bellone E, et al. 17pll.2 du-plication is a common finding in sporadic cases of Charcot-Marie-Tooth type 1. Eur Neurol 1994;34:135-9.

41 Silander K, Meretoja P, Nelis E, Timmerman V, Van Broeck-hoven C, Savontaus ML. A de novo duplication in 17p 11.2and a novel mutation in the P0 gene in two Dejerine-Sottas syndrome patients. Hum Mutat (in press).

42 Lupski JR. DNA diagnosis for Charcot-Marie-Tooth diseaseand related inherited neuropathies. Clin Chen, 1996;42:995-8.

43 Valentijn LJ, Baas F, Zom I, Hensels GW, de Visser M,Bolhuis PA. Altematively sized duplication in Charcot-Marie-Tooth disease type 1A. Hurn Mol Genet 1993;2:2143-6.

49

group.bmj.com on July 15, 2011 - Published by jmg.bmj.comDownloaded from

doi: 10.1136/jmg.34.1.43 1997 34: 43-49J Med Genet

 V Timmerman, B Rautenstrauss, L T Reiter, et al. patients of European descent.recombination hotspot in unrelated Detection of the CMT1A/HNPP

http://jmg.bmj.com/content/34/1/43Updated information and services can be found at:

These include:

References http://jmg.bmj.com/content/34/1/43#related-urls

Article cited in:

serviceEmail alerting

the box at the top right corner of the online article.Receive free email alerts when new articles cite this article. Sign up in

Notes

http://group.bmj.com/group/rights-licensing/permissionsTo request permissions go to:

http://journals.bmj.com/cgi/reprintformTo order reprints go to:

http://group.bmj.com/subscribe/To subscribe to BMJ go to:

group.bmj.com on July 15, 2011 - Published by jmg.bmj.comDownloaded from