DNA-Based Prenatal Diagnosis of Generalized Recessive Dystrophic Epidermolysis Bullosa in Six Pregnancies at Risk for Recurrence

REPORTS

DNA-Based Prenatal Diagnosis of Generalized Recessive Dystrophic Epidermolysis Bullosa in Six Pregnancies at Risk for Recurrence Alain Hovnanian,*t Latifa Hilal,* Claudine Blanchet-Bardon,$ Christine Bodemer,§ Yves de Prost,§ Caroline A. Stark,t Angela M. Christiano,1I Marc Dommergues,** Joseph D. Terwilliger,§ Luis Izquierdo,tt Patricia Conteville,* Yves Dumez,** Jouni Uitto,H and Michel Goossens* •Laboratory of Biochemistry, INSEl^JVl U.91, Hopital Henri Mondor, Creteil, France; tThe Wellcome Tru.st Centre for Human Genetics, Oxford, United Kingdom; ^Department of Dermatology, Hopital Saint-Louis; §Departnicnt of Dermatology, Hopital Necker-Enfants Malades, Paris, France; UDepartments of Dermatology, Biochemistry and Molecular Biology, Jefferson Medical College, Thomas Jefferson University, Philadelphia, Pennsylvania, U.S.A.; "Department of Obstetrics, Hopital Port-Royal, Paris, France; and ttCentre for Genetic Investigation, Madrid, Spain

Linkage analyses in generalized recessive dystrophic epidermolysis bullosa (RDEB) have implicated the type VII collagen gene (COL7A1), which encodes the major component of anchoring fihrils, and recent identification of COL7A1 mutations has provided direct evidence for COL7A1 defects underlying RDEB. In this study, COL7A1 gene analysis was used to successfully perform first-trimester prenatal diagnosis in six fannilies at risk for recurrence of the disease. In four families, three affected with the most severe variant of RDEB (the Hallopeau-Siemens form, HS-RDEB) and one with generalized nonmutilating RDEB, prenatal diagnosis was established by linkage analysis using polymerase chain reactionbased detection of PvuU and Alul intragenic restriction fragment length polymorphism. In two other HS-RDEB families, prenatal diagnosis was carried out by direct detection of mutations in COL7A1,

using denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genomic fragments. Analysis of fetal DNA from chorionic villus biopsy or from amniotic fluid cells showed that the fetus had inherited at least one normal COL7A1 allele in all cases. Therefore, the fetus was predicted to be unaffected in the six pregnancies, and this has been confirmed in the newborn infants. Genotype analysis with COL7A1 polymorphic markers, or direct COL7A1 mutation detection in families at risk for the disease, represent early and rapid diagnostic alternatives to second-trimester evaluation of fetal skin samples, and thus offer a major advance in prenatal diagnosis of this life-threatening form of epidermolysis bullosa. Key words: type VII coUagenl linkage analysis I mutation detectionlHallopeau-Siemens. J Invest Dermatol 104:456-461, 1995

G

eneralized reces.sive dystrophic epidermolysis bullosa (RDEB) is cbaracterised by widespread scarring and blistering of tbe skin and mucous membranes from birtb due to loss of dermal-epidermal adhesion [1-3]. In tbis form of epidermolysis bullosa, cleavage occurs beneatb the lamina densa and is associated witb absence or marked reduction in the number of ancboring fibrils (AF) [4,5]. Indirect immunofluorescence labeling using antibodies against type VII collagen (COL7A1), the major component of AF [6], is markedly reduced or negative in tbe skin of a majority of patients [7-10]. Tbe mutilating variant of generalized RDEB (i.e., the Hallopeau-Siemens form HS-RDEB) (Mendelian Inheritance in Man, Manuscript received October 7, 1994; revised December 14, 1994; accepted for publication January 10, 199.S. Reprint requests to: Dr. Alain Hovnanian, The Wellcome Trust Centre for Human Genetics, Windmill Road, Headington, Oxford OX3 7BN, United Kingdom. Abbreviations: AF, anchoring fihrils; COL7A1, type VII collagen; DGGE, denaturing gradient gel electrophoresis.

access number 226600),tt is tbe most severe form of RDEB, and is manifested by extensive mucocutaneous blistering leading to mutilation and mitten-like deformities of the hands and feet, esophageal strictures, and severe disabling complications sticb as malnutrition, sepsis, and squamous cell carcinomas [11]. Close genetic linkage between generalized RDEB and the COL7A1 gene, tbe candidate gene for tbe disease, lias previously been shown using an intragenic 7'i'iiII restriction fragment length polymorphism (RFLP) [12,13]. Recently, several mutations in the COL7A1 gene bave been reported in patients with generalized RDEB, tbus providing evidence for COL7A1 defects underlying RDEB [14-16]. Specifically, tbese mutations include a Met-to-Lys substitution [17], an insertion-deletion in tbe COL7A1 gene leading to a null allele [18], tbree premature termination codons resulting from one or two base insertions or deletion.s [19], as well as four nonsense mutations (TGA) resulting from a C-to-T substi-

%X McKusick VA: Mendeliau Inheritance in Man. McKusick VA (ed.). 10th edition. Johns Hopkins University Press. Baltimore and London, 1992

()022-2()2X/9.S/S()9.50 • SSDI0022-202X(94)()(l()29-K • Copyright C3 1995 by The Society for Investigative Dermatology, Inc. 456

VOL. 104, N O . 4

DNA-BASED PIUENATAL DIAGNOSIS OF BX)EB

APRIL 1995

tutioii in CGA arginine codons [20,21], and one termination codon due to a G to A transition in a tryptophan codon 119]. Until recently, carrier detection was not possible, and prenatal diagnosis of HS-RDEB relied on electron microscopy [22,23] and/or indirect inimunofluorescent examination of fetal skin samples obtained at 18-20 weeks gestation [24-26], Although reliable, this method of diagnosis neces.sitates late termination of the pregnancy, if elected, in tbe case of an affected fetus. The recent cloning of" the COL7A1 gene tbat has been mapped to 3p21 [27-311 anci the availability of intragenic lVFLPs [32,33] now offers the pt>ssibility of DNA-based diagnosis in RDEB families informative for these markers [34], Alternatively, although relatively few mutations have been reported thus far, the identification of mutations in COL7A1 in patients affected with RDEB makes prenatal diagnosis by direct mutation analysis feasible in a number of families. We report first-trimester prenatal e.vclusion of severe generalized RDEB by linkage analysis using the intragenic Pi'(/ll and Aiul RFLPs in four families, and by direct COL7A1 mutation analysis in two other families at risk of recurrence ot the disease, MATERIALS AND METHODS Families Charaetei isties of the .six families studied are shown in Table I and in Fig 1, All alFceted individtials ftiltilled tlie elinieal and eleetionmieroscopie eriteria for generalized RDEli as defined by Fine et al \2\. Patients from families I and ^ to 0 had fti.sion ofthe fingers and/or toes, with mierostoniia and aiikyloglossia, growtli retardation and aiieniia, and were cla.ssified as HS-lU)Eli. The patient from family 2 had widespread skin blistering without fusion of the digits and was diagnosed as having generalized noii-nuitilatiiig RDEB. Eleetroii iiiieroscopy eNamination of skin from affeeted patients in the six families showed eleavage beneath the lamina densa, and AFs were ludinientary or absent. Iiidireet inimunoHuoreseeiiee staining of skin biopsies using the moiioelonal antibody LH7:2 against type VII eollagen revealed no staining in all the patients tested (fiimilies 3, 4, and 5). Fetal ONA Sampling Transabdoininal elu>riouie villus biopsy was performed under ultrasound guidance between 10 and 12 weeks ol gestation, in tlie ease of pregnancies in families 1 and 3 to 6, Aniniotie fluid eells wore obtained by aninioeentesis performed at 13 weeks gestation in laniily 2, Extraction of Genomic D N A Genomie DNA from parents and sibs was extracted by standard pioeetinies frtini peripheral blood leukoeytes |35]. Fetal DNA was extracted from elioriouie villus samples (families I and 3-6) or from cultured aniniotie fluid eells (family 2), by incubating the biopsy or the cell pellet in 300 /J-l of lysis buffer (8 M urea, 0.3 M NaCl, 10 niM Tris-HCl, pH 7.4, 10 niM ethylenediaminetetraaeetie aeid |EDTA], pH 8.0, 2% sodiunidodeeylsulfate) at room temperature for S h. After phenolchlorofoiin extraction and etiianol proeipitation, the DNA was suspended in 50 fJ-1 of 10 mM Tris-HCl (pH 7.4), 1 niM EDTA (pH 8.0) solution, and clectropboresed on an agarose gel to assess DNA concentration and qtiality. Amplification of Genomic DNA by Polymerase Chain Reaction (PCR) DNA samples were ainplilied using primers specific for the RFLl's for Pnill, Ahi\, and ,S()'I at the C:C")L7AI locus |32,33], and for the Mspl RFLP at locus D3S2 [36| described in Table II. Seven pairs of primers suitable for denaturing gradient gel eleetroplioiosis (DGGE) analysis, one of each pair with a GC-rich clamp added |37], were also used to amplify exons

457

5, 6, 12-15, and 19 to screen for mutations witliin the COL7A1 gene. The primers used for detection of the premature stop codon at arginine codon 109 in exon 6 (FN-IA) and the insertion-deletion in exon 12 (FN-4A), a.s well as the conditions used for DGGE, are indicated in Table III. The sequence of the other primers and the corresponding DGGE condiriotis have been described previously [20], Polymerase ehain reaction (PCR) was performed in a UNA thermal cycler (Pcrkin Elmer Cetus) at 94°C for 4 min followed by 3.S cycles of 94°C for 1 min, the relevant annealing temperature (Tables II and III) for 1 min, and 72°C for 1 min 30 seconds, with a final extension of 7 min at 72°C. DGGE Analysis and Nucleotide Sequencing DGGE eonditiotis were determined using the MELT87 and SQHTX programs written and provided by L. Lerman and colleagues [38,39]. Those eoniptiter algorithms were used to select tbe position of PCR primers best suited for DGGE analysis, and to determine tbe electropborctic conditions resulting in optimal resolution and hetorodtiplex formation. Twenty tiiicroliters of each amplified DNA sample was subjected to elcctropboresis at 160 V (10 V/cni) at 6()''C in a 6.5% polyacrylamide gel containing a linearly increasing denaturant coneentration (100% denaturant = 7 M urea and 40% formamide). The D N A fragniLMits sliovving a shift in electTophoretic mobility were subjected to seqtience determination after asymmetric amplification, as previously described [40]. Restriction Endonuclease Analysis of Amplified Genomic DNA Twenty-microliter aliqnots of the amplified genomie DNA was elcaved vvith tbe appropriate restriction enzyme according to the manufacturer's instructions (New England Biolabs), analyzed by clectropboresis on n 2% agarose gel and vistialized by staining witb ethidium bromide. Tbe PCR products were putified on QlAquick-spin (Qiagon) columns before cleavage witb tbe Alu\ restriction enzyme. Genetic Linkage Analysis Comptitcr programs MLINK and LINKMAP of tbe Linkage Package Version S.I [41] were used to perform litikage analyses in our RDE13 families. Lod score calculations were perfomiod by designating RDEB as a recessive disease witb complete penetrance., and taking into aceotmt tbo observed consanguinity. Tbe fi-eqticney of tbe disease allele was estimated to be 1/500, assnming an incidence of tbe disease of 1/250,000, Heterogeneity analysis was perfomiod witb tbo H O M O G program [42], RESULTS The six pregnancies at risk for RDEB were studied by genotyping tbe families and the fetus for the Pi'idl and Alul RFLPs within the COL7A1 gene, and by investigating the affeeted members of these families for mutations within the COL7A1 gene. Additionally, these families were tested for the Mspl RFLP at the nearby locus D3S2 at 3p21,2-p21,3 [36], and retrospectively for a rare Styl RFLP within tbe COL7A1 gene [33],

Genotyping for COL7A1 Markers and Prenatal Exclusion of RDEB Family 1: RFLP analysis of the Pi'iill polymorphism proved partially informative in this family, whereas the Alul and the D3S2 markers were not informative (Fig 1), Analysis of the fetal D N A using tbe Pi'iill RFLP showed that the fetus had inherited the normal paternal allele, thus predicting an unaffected fetus (Fig 1),

Table I. Characteristics of the 6 Families Studied, with Results of the Genotype and Mutation Analyses RFLP analysis'

Pedigree

Origin

Consanguinity"

Family 1 Family 2 Family 3 Family 4 Family 5 Family 6

French Spanisb Ttmisian Moroccan Spanisb Froneb

No No

1st degroo 2nd dogreo No No

Ntimbor of Offspring witb RDE15/ Number of Offspring 2/4 1/1 2/5 1/1 1/1 1/1

" Observed consanguinity. '' At the time prenatiil liiiigntisis was requested, '' Fl, fully iniornialive; PI. partially ini'onnativc; NI, not int'onnative

COL7A1 Pi'iiII

Alu\

Styl

D3S2 Mspl

PI

NI Nl PI Nl Nl Nl

NI Nl

NI NI

Fl Nl NI Nl

PI Fl Nl Nl

Fl PI

PI Fl Nl

COL7A1 Mutation Analysis''

Insertion-doletion in oxon 12 [18] Arg-109-Stop in oxon 6 [20]

458

HOVNANIAN ET AL

THE JOURNAL OF INVESTIGATIVE

Family 1

Family 2

Family 3

1

'CHpO

LJ—\—KJ I 1 2 1

coL7Air2!!r Lay, O3S2

uvl

2* I 2 1

12 1 1 22 11

2 2 1 1

2 2 1 I

/U// Slyl AV

D3S2

'^ '6 1 2 1 1

[

12 11 22 1 1

Pwl

1 ^ or norm.

1 t 2 2 11

t 1 22 1 1

2 1 2 1

1 1 1 1

2 1 2 1

2 2 1 1

CCH.7A1

Atlt Slyl

D3S2

2 1 1 2

2 2

'6

2 1 2 1 .

Family 4

D3S2

1 2* 2 t 2 1

PwU

1 2* 1 1 2 2 I 1

1 2'

DERMATOLOGY

1 2 2 1 2 1 22

2 1 2 t

2* t 1 2

•2 2*

1 2 2 1 2 2 21

1 1 1 1

2 2

< 2 2 2

2 1 t 2

Family 6

Atil Sly/ 1*^

1t 2 2 1 2

•1 2 1 1 22 2 I

C0L7A1 ay/ 2 2 I- in*, dal. + -

"t 2 1 I 2 2

LJ—1

j-ftu*

' 1 1 1 1

CCK.7A1

D3S2

D3S2

Slyl LRIOSX *M

2 2

4- 1 1

' ^ 1 i" 2 2 2 2 _ 1 1

3

1 1 1 1 22 22

1 1 1 1 2 2 t 2

•11* 1 2 2 2

22

1 rofnorm 1 2 22

Figure 1. The pedigrees ofthe six families demonstrating DNA-hased prenatal diagnosis using amplified COL7A1 markers, or direct COL7A1 mutation detection. The haplotypes ofthe COL7A1 intragenic RFLI's (Pi'»II, Alnl, and Sty\) and ofthe extragenic RFLP (Mspl) at locus D3S2, are presented. The Styi marker has been tested retrospectively. The results of the mutation analysis in families 5 and 6 are also indicated (ins. del., insertion-deletion; R109X, nonsense mutation changing arginine 109 to a stop codon). Open symhol, healthy; shaded symhol, affected; crossed throngh symhol, deceased; diamond, fetus; *, mutated COL7A1 allele; (*), one ofthe two COL7A1 alleles is mutated, but these alleles are indistinguishable due to lack of informativeness. * or nomi., mutated or normal COL7A1 allele. 1 and 2 indicate the ahsence and the presence ofthe corresponding restriction enzyme site, respectively.

Family 2: This family was fully informative for the PvuW polymorphism, and not informative for tbe Alu\ and D3S2 markers (Fig 1). Tbe affected child was homozygous for the presence of tbe Pi'//II site and linkage analysis using tbis marker sbowed that tbe fetus had inherited botb normal parental COL7A1 alleles (Fig 1).

allele (Fig 1), thus predicting the fetus to be unaffected. However, because ofthe potential risk of recombination due to tbe extragenic location ofthe Mspl marker, ultrastructural analysis of skin biopsies was performed at 20 weeks gestational age, whicb confirmed tbat the fetus was not affected.

Family 3: In this consanguineous family, the 7^I'HII, Alu\, and Mspl markers were found to be partially informative, but the combined u.se of these markers led to full infonnativeness. Tbe affected living offspring was bomozygous for tbe COL7A1 markers tested, and analysis ofthe Pvull and tbe Ahtl markers showed tbat the fetus had inherited the normal COL7A1 allele from bis father. Tbe fetus was therefore predicted to be unaffected.

Family 5: This family was fully informative for the 7^i'/(Il RFLP, and not informative for tbe Aiul and Mspl markers. Tbe affected offspring was homozygous for the absence of the Pvull site, and analysis of the fetal DNA using the PI'MII RFLP sbowed that tbe fetus was heterozygous for the COL7A1 allele linked to the disease. This result was confirmed by direct detection of tbe mutation underlying the disease in this family (see below).

Family 4: This consanguineous family was partially informative for the PI'MII RFLP, not informative for the Aiul site, and fully informative for the extragenic Mspl marker (Fig 1), witb an affected offspring homozygous for tliese markers. Analysis of tbe fetal DNA using the PI^KII RFLP showed that the fetus had inberited the maternal allele associated witb tbe disease, and testing of tbe Mspl RFLP suggested that tbe fetus had inherited the normal paternal

Family 6: This family was uninforinative for tbe Pi'ull, the Aiul, and the D3S2 RFLPs, and therefore DNA-based prenatal diagnosis relied on the identification of COL7A1 mutatioii(s) in this family.

Table II.

DNA Polymorphisms Used in This Study, and Corresponding PCR Amplification Conditions

Locus

RFLP

Primer Sequence

COL7A1 3p21

PmiU

5 ' A C A G T C C C G G G T C C T G G G 3' (F) 5 ' A C A G G C G T C A G G G A G A A G A T 3' (R) 5 ' A G G A G G C T T G T G G A G G A C A G 3' (F) 5 ' C A C T G C C T G T G C T C C A G C T A 3' (R) 5 ' G G C C A G A A G A G A T C C T G A G T 3' (F) 5 ' C T G A C C T G T C A C T C C T G C T C 3' (R) 5 ' C C A A G T G G C A G A G C T A C T T A 3' (F) 5 ' A C T A T C T T G C C C T A G A G C A A 3 ' (R)

Aiul Slyl D3S2 3p21.2-21.3

Prenatal Exclusion of RDEB Using Detection of COL7A1 Mutations At tbe time prenatal diagnosis was performed, tbe search for mutations has proved po.sitive in families 5 [18] and 6

Mspl

" These annealing temperatures are slightly different from those published in [32,33,35]. The sum of these fragments is less than 195 because the primers contain Aiul sites.

Annealing Temperature" 55°C 55-0 65°C 55°C

Size (hp) 698 (593/10.=;) 19.'i (101/59)'' 422 (211/211) 473 (237/236)

Allelic Frequency

Reference

0.40/0.60

[32]

0.25/0.75

[32]

0.14/0.86

|.33)

0.30/0.70

|36]

VOL. 104. N O . 4

Table III.

DNA-BASED PRENATAL DIAGNOSIS OF RX>EB

APRIL 1995

459

PCR and DGGE Conditions Used for Direct Mutation Analysis of Exon 6 and 12 in the COL7A1 Gene DGGE Conditions Primer Sequence

Mutation

Exon

Arg-109-Stop 12

55°C

50-90

[18]

55°C

50-90

[20]

Family 5: DGGE analysis was used to detect a framesliift insertion-deletion in exou 12 of the COL7A1 gene present in the homozygous state in the affected offspring from family 5 [18]. The DGGE migration pattern showed that the fetus had inherited the mutated COL7A] allele from his father, and the non-mutated COL7A1 allele from his lnother (Fig 2a), thus predicting the fetus to be an unaffected carrier. Family 6: We recently identified in the patient from family 6 a premature stop codon in exon 6 inherited from the father, whereas the maternal defect remains to be characterized [20]. I'GGE analysis showed that the fetus had not inherited the mutated paternal allele, thus predicting the fetus to be unaffected (Fig 2b). Because this mutation creates a NlalU restriction site, the absence of this mutation in the fetal DNA was also verified by endonuclease digestion of PCR-amplified DNA. The pregnancies were carried to term, aud the six newborn infants were clinically unaffected. Retrospectively, testing of a rare intrageiiic S()'I RFLP showed that family 3 was fully informative for this marker, whereas none of the other families studied were informative for this marker (Fig 1). DISCUSSION W e present prenatal exclusion of HS-RDEB in six pregnancies at risk for the disease, using linkage analysis with COL7A1 polymor-

2

3

4

C

1

2

3

4

C

- mutant - normal polymorphic mutant normal

Reference

phic markers or direct COL7A1 mutation detection. Prenatal diagnosis relied on segregation analysis of the iutragenic Pviill RFLP in four families in which no mutations had been identified after screening part of the COL7A1 coding sequence [20]. The PI'HII marker was fully informative in two families (2 and 5), and partially informative iu three other families (1, 3, and 4). Segregation of this marker showed that the fetus was hoinozygous for the uormal COL7A1 alleles in family 2, was heterozygous for the mutated allele in family 5, and had inherited at least one uomial COL7A1 allele iu families 1 and 3, thus predicting an unaffected fetus in these pregnancies. In the case of fainily 4, analysis of the Pvull marker showed the fetus had inherited the maternal COL7A1 allele associated with the disease and analysis of the Mspl RFLP at the D3S2 locus suggested that the fetus had inherited the normal paternal COL7A1 allele. However, the risk of reconibiiiatioii between this extragenic marker aud the disease locus led us to obtain confirmation of the unaffected status of the fetus by electron microscopic examination of fetal skin biopsies. The identification of causative COL7A1 mutations in the affected offspring of families 5 and 6, pennitted us to use direct mutation detection for preuatal diagnosis of RIOEB iu these families. Although the maternal COL7A1 mutation remained unidentified in family 6, the possibility of testing the fetal DNA for the paternal mutation was particularly important iu this family, which was uuinformative for all the COL7A1 markers tested. Direct detection analysis of PCR-amplified products using DGGE showed that in family 6 the fetus had not inherited the mutated paternal allele, and

[20], thus allowing direct lnutation detection for prenatal diagnosis in these families.

1

Dcnaturant Range (%)

5'GGCAGTGCTGATTCCATCCT 3' 5'(6{)GC)ACCCCGCAGCCGCACACT 3' 5'GTAACAGACCTGCAAGCCAC 3' 5'(60GC)GAGAGGGCTGGAGGTACAC 3'

25 bp insertion 11 deletion

Running Time (h) at 160 V

Annealing Temperature

Figure 2. Prenatal diagnosis of HS-RDEB by analysis of COL7A1 mutations in families 5 and 6 using DGGE. a) In family 5, the affected offspring is homozygous for a COL7A1 allele of reduced electrophoretic mobility correspoiidiug to an insertiondeletion in exon 12 [18]. The fetus inherited the mutated allele from his father, and the maternal COL7A1 allele not associated with the disea.se, as indicated by the presence of two honioduplexes and lieteroduplexes identi^'^^ ^^ ''•'* mother. The reduced electrophoretic mobility of the nonniutated maternal allele is due to the presence of a rare Fspl RFLP 118]. h) In family 6, the affected offspring has inherited the COL7A1 paternal allele of reduced mobility', corresponding to a C-to-T transition at argiiiine codon 109 in exon 6 [20]. The fetus did not inherit the mutated paternal allele, as indicated by the absence of a band of decreased mobility, as well as the absence of lieteroduplexes. H, honioduplexes; h, heteroduplexes.

460

HOVNANIAN ET AL

confirmed that in family 5 the fetus wa.s heterozygous for the insertion-deletion. The reliability of DNA-hased prenatal diagnosis by indirect genotype analysis is dependent on the absence of locus heterogeneity for the disease, and on the frequency of recombination between the markers used and the mutation causing the disease. The PI/KII RFLP within the COL7A1 gene has previously been shown in three independent studies to be closely linked to the BJ^EB locus in 19 families the majority of French origin (lod score of 3.97 at B = 0) [12], in 16 fiimilies mostly of British and Italian origin (lod score of 3..S1 at 0 = ()) [13], and in 14 families of diverse ethnic backgrounds (lod score of 2.31 at 0 = 0) [16] (Christiano et al, unpublished). We have recently investigated a total of 42 families from the French EB registry, affected with generalized RDEB, for linkage of the disease to the intragenic Pt'itU and Alii\ RFLPs. Two-point linkage analysis using the MLINK program gave a maximum lod score of 4.66 at 6 = () with the Pi'iiU marker, and a lod score of 0.76 at fl = 0 with the Aln\ site, with no obligate recombinations (unpublished). Multipoint analysis using the LINKMAP program gave a maximum lod score of 5.74 with the PtniU and the Alii] markers at the COL7A1 locus with no evidence for locus heterogeneity. Therefore, the Pi>uU and the Ahil RFLPs provide reliable markers for carrier detection and prenatal diagnosis in these families, with a very low risk of recombination between these markers and the COL7A1 mutation due to their intragenic location and the relatively compact size (32 kh) of the gene [31]. We also tested for linkage of the disease to the Mspl RFLP at the nearby locus D3S2 [36]. A lod score of 3.04 at fl = 0.04 was found with this marker (unpublished), which had an obligate recombination with the disease locus in one out of 42 HS-RDEB families studied, indicating that this marker presents a potential risk of misdiagnosis due to random recombination with the RDEB locus. At present, the Pmill RFLP is the most informative intragenic marker available for linkage studies in RDEB [33]. Of the 42 1U:)EB families from the French epidermolysis bullosa registry studied, six families (14.3%) proved fully informative for this marker alone, and five families (11.9%) became fully informative when both the Pi'iiII and Alul markers were used (unpublished). Recently, four new COL7A1 RFLPs have been characterized: a Styl polymorphism that wo retrospectively tested in our six families, a M.9pl, an Eco0109I, and a Mull site. However, the allelic frequencies of these markers indicate that they are likely to be of limited informativeness for genotype analyses [33]. Therefore, novel intragonic markers or highly informative multiallelic polymorphisms such as (CA),, repeats near the COL7A1 locus are needed to perfonn carrier detection and prenatal diagnosis by linkage analysis in a greater proportion of families. An alternative to linkage analysis in prenatal diagnosis of RDEB is direct analysis of the COL7A1 gene defect within the families, as illustrated in families 5 and 6. The intron-exon organization of the COL7A1 gene lias recently been established |31]. It is a complex gene consisting of 118 exons encoding a 9.2-kb niRNA [27-311. At present, only a few COL7A] mutations have been published in patients with BJDEB, most of which have been family specific [17-21]. The identification of causative COL7A1 mutations in an increasing number of patients affected with RDEB is anticipated to have direct diagnostic applications, and may become the method of choice for prenatal diagnosis of RDEB.

THE JOURNAL OF INVESTTGATIVIi DERMATOLOGY

Wellcoiue 'Trust fellowship, C. Stark is supported hy a DEBRA UK studentship, and A. M. Christiauo is the recipient of a Dermatology foinidntioii Career Dereloptuent Award.

REFEl^NCES 1.

2.

3.

4.

5.

6. 7.

8.

9.

10.

11.

12.

13.

14. 15. 1 6.

1 7.

18.

19. 21).

21. IVe are grateful to Dr. M. Adda, Dr. M. Brenier, Dr.J. Chassevent, Dr. B. Michel, and Dr. P. Morierfor referring the patients to our laboratory. We wish to thank P. Plassa, Dr. S. Lyonnel, and Dr. P. Saugier for their participation in this work. We thank the Epidermotyses Bulleuses Association Entraide (BBAE), the Dystrophic Epidermoly.ns Bullosa Reseach Associalion (DEBU.A) UK and America, and the patients and their families who kindly gave samples. This work was supported hy the Association Fraiifaise contre les Myopathies (AFM), the INSllRM, The Wellcome Trust, the DEBRA UK, the EBAE, the March of Dimes Birth Defects Foundation, and the USPHS, NIH grant l>()l-AR3li92.1. A. Hovuaniau is supported hy a

22.

23. 24.

25.

Gedde-Dahl TJ, Anton-Lamprcclit I: Epidi;rniolysis bullosa. In: Principles and Practice of Medical Ccnetics. Rinioiii DL (cds.). Clnircliill Livingstone. New York. New York. 1990. pp 8S5-H7(i Tine JI3. Baner F.A. Brigganian RA. C:arter DM. Eady RA. nsterly Nli. Hollirook KA. Hiirwitz S. Jolinson L. Lin A. Pearson R. Sybert VP: Revised clinical and laboratory criteria for subtypes of inlierited epidermolysis biiUosa: :i consensus report by the subcommittee on diagnosis and classification of the national epidermolysis bnllosa registry. / Am Acad Dennntol 24:1 19-135. 1991 Uitto J. Christiano A: Molecular genetics of tbe cutaneotis basement membrane zone. Perspectives on Epidermolysis Bnllosa and otber blistering skin diseases. / Clin Invest 90:687-692. 1992 Tidman MJ. Eady RAJ: Evaluation cif anchorhig fibrils and otber components of the derm.il-epidernial Junction in dystropbic epidermolysis bullosa by a quantitative ultrastructural tecbniqtie. J Invest Deniiatol 84:374—."^77. 1985 McGralh JA. Isbida-Yamamoto A. O'Grady A. Leigb IM. Eady RAJ: .Structural variations in ancboring fibrils in Dystropbic Epidermolysis Bullosa: correlation witb type VII collagen expression. / Invest Dermitlol IO(l:3(i6-372. 1993 .Sakai LY. Keene DR. Morris NP. Burgeson RE: Type Vll collagen is a major structural component of ancboring fibrils. / Cell Hiot 103:1 577-1 .S86. 1986 Heagerty AHM. Kennedy AR. Leigh IM. Purkis P. Eady RAJ: Identification of an epidermal basement membrane defect iu recessive forms of dystrophic epidermolysis bullosa by LH7:2 monoclonal antibody: use in diagnosis. lir J Damahil 1 I5:12.S-13I. l9Kd Leigb IM. Eady RAJ. Heagerty AHM. Purkis PE. Wbitebead PA. Burgeson RE: Type VII collagen is a normal component of epidermal basement membrane, wbicb sbows altered e.xpression in recessive dystrophic epidennolysis bullosa. J hwest Dnmalol 90:639-642. 1988 Bruckner-Tudermaii L. Riiegger S. Odermatt B. Milsuhasbi Y. Schnyder UW: I.ack of type VII collagen in unaffected skin of palieiUs with severe recessive dystrophic epidennolysis bullosa. Demiatoloi^ica 176:57-64. 1988 Iiruckner-Tudennan L. Mitsubasbi Y, Scbnyder UW. Bruckner P: Anclioring fibrils and type VII collagen are absent from skin in severe recessive dystropliic epidermolysis bullosa. J Invest Dermatol 93:.3-9. 1989 Briggaman I^A: Recessive dystrophic epidermolysis bullosa: a clinical overview. In: lijtiilvmwlysis Bullosa. Basic ami Clinical Aspects. Lin AN. Carter DM (eds.). Springler-Verlag. New York. 1992. pp 13.S-I51 Hovnanian A. Duquesnoy P. Blanchet-Bardon C. Knowltoii Rti. Amselem .S, Lathrop M. Dubertret L, Uitto J, Goossens M: Genetic linkage of recessive dystrophic epidermolysis bnllosa to tbe type VII collagen gene. / C.ViM Invest 90:1032-1036. 1992 Dunnill MGS. Richards AJ. Milana C. Mollica T. Atberton D. Winsbip I. Farrell M. Al-hnani L. Eady RAJ. Pope PM: Genetic linkage to tbe type VII collagen gene in 26 families with generalised recessive dystropbic epidermolysis btiilosa and ancboring fibril abnormalities. / Mftl Getift 31:745—748. 1994 Burgeson RE: Type VII collagen, anchoring fibrils, and epidermolysis bullosa. J Invest Di-nnatol 101:252-255. 1993 Hovnanian A. Christiano AM. Uitto J: The molecular genetics of dystropbic epidermolysis bullosa. Arch Dermatol 129:1566-1570. 1993 Uitto J. Cbristiano AM: Molectihir basis of the dystrophic forms of epidermolysis bullosa: mutations in tbe type VII collagen gene. Arcli Dertnatol Res 287:16-22, 1994 Cliristiano AM. Greenspan DS. I lollhian GG. Zbang X. Tamai Y. Lin AN. Dietz HC. Movnanian A. Uitto J: A missense mutation hi type Vll collagen in two affected siblings witb recessive dystrophic epidermolysis bullosa. Natnie Genet 4:62-66. 1993 Hilal II. Rochat A. Duquesnoy P. lilanchet-Bardon C. WechslerJ. Martin N. C;llristiano AM. Bariandoll Y. UittoJ. Goossens M. Hovnaniau A: A honuKy.. gous insertion-deletion in tbe type VII collagen gene (COL7A1) iu Hallopeau.Siemens dystropbic epidermolysis bullosa. Nature Genet 3:287-293. 1993 Christiano AM. Anhalt G. Gibbons S, Bauer EA. UittoJ: Premature termination codons in the type VII collagen gene (COL7A1) underlie severe, mutilating recessive dystrophic epidennolysis bullosa. Gemmiics 21:160-168. 1994 Hovnanian A, Hilal L. Blancliet..l).irdon C, de Prost Y. Cbristiano AM. UittoJ. Cloossens M: Recurrent nonsense mutations witbin the (ype VII collagen gene in patients witb severe recessive dystrophic epidermolysis bullosa. Am J Hum Gem-I 55:289-296. 1994 Dunnill MGS. Ricbards AJ. Milana G, Mollica F. Eady RA. Pope FM: A novel homozygous point mutation in tbe collagen Vll gene (CX)L7A1) in two cousins witb recessive dystropbic epidermolysis bullosa. Hum Mot Genet 3:1693-1694. 1994 Anton-Lamyirecbt 1. Jovanovic V. Arnold M-L. Ranskolb R. Kern B. Scbenck W: Prenatal diagnosis of epidennolysis bullosa dystrt)iiliica Mallopeati-Sieniens witb electron microscopy of fetal skin. Lancet V:I077-1079. 1981 Anton-Lainprecht I: Prenatal diagnosis of epidermolysis bullosa bereditaria: a review. Semin Dermatol i:22'>-24i). 1984 Heagerty Al IM. Kennedy AR. Guuner DB. l-ady RAJ: Rapid prenatal diagnosis and exchisi()n of epidernioiysis bullosa using novel antibody probes. / htvest Dermatol 86:603-605. 1986 Sybert VP. Flolbrook K: Prenatal diagnosis and genetic screening for epideiinol-

DNA-BASED PRENATAL DIAGNOSIS OF RDEB

VOL. 104, N O . 4 APRIL 1995

26. 27.

28

29.

30

31.

32

ysis bullosa. In: Epidennolysis Bnllosa: Basic and Clinical Aspects. Lin AN, Carter DM (eds.). Spriiigler-Verlag, New York, 1992, pp 235-251 Holbrook KA, Smith LT, Elias S: Prenatal diagnosis of genetic skin disease using fetal skin biopsy samples. Arch Dermatol 129:1437-1454, 1993 Parente MG, Chung LC, Ryyniinen J, Woodley DT, Wynn KC, Bauer EA, Mattei M-G, Chu M-L, Uitto J: Human type VII collagen: cDNA cloning ,ind chromosomal mapping ofthe gene. Proc Nail ACMI Sci USA 88:6931-6935, 1991 Christiano AM, Rosenbaum LM, Chung-Honet LC, Parente MG, Woodley DT, Pan T-C, Zhang RZ, Chu M-L, Burgeson RE, Uitto J: The large noncollagenous domain (NC-1) of type VII collagen is amino-terniinal and chimeric. Homology to cartilage matrix protein, the type 111 domains of fibronectin and the A domains of von Willebrand factor. Hnin Mol Cenel 1:475-481, 1992 Gammon W R , Abemethy ML, P,idilla KM, Prisayanh PS, C:ook ME, Wright J, Briggaman RA, Hunt SWl: Noncollagenous (NCI) domain of collagen VII resenihles multidomain adhesion proteins involved in tissue-specific organization of extracellular matrix. J Invest Dermatol 99:691-696, 1992 Greenspan DS: The carhoxy-terniinal half of type VII collagen, including the non-collagenous NC~2 domain and iiitron/exon organization of the corresponding region ofthe COL7AI gene. Hum Mol Ceiiet 2:273-278, 1993 Christiano AM, Hoffman GG, Chuug-Honet LC, Lee S, Wen C, Uitto J, Greenspan DS: Structural organization of the human type VII collagen gene (COL7A1), composed of more exons than any previou.sly characterized gene. Ceiiomics 21:169-179, 1994 Christiano AM, Chung-Honet LC, Hovnanian A, Uitto J: PCR-based detection of two exonic polymorphisms in the human type VII collagen gene (COL7A1) at 3p21,l. Ceiiomics 14:827-828, 1992

33.

34. 35. 36. 37.

38.

39.

40.

41. 42.

Christiano AM, Greenspan DS, Lee S, Uitto J: Cloning of human type VII collagen. Complete primary sequence ofthe a l (VII) chain and identification ofintr,igenicpol>TOorphisms.J'Bi()i Chem 269:20256-20262, 1994 Christiano AM, Uitto J: DNA-based prenatal diagnosis of heritable skin diseases. Areh Dernwiol 129:1455-1459, t993 John SWM, Weitzner G, Rozen R , Scriver CR: A rapid procedure for extracting genomic DNA from leukocytes. Nucteic Acids Res 19:408, 1991 Ganly PS, Rahbitts PH: Polymerase chain reaction (PCR) for detection of Mspl polymorphism at the D3S2 locus. Nucleic Acid Res 19:3761, 1991 Sheffield VC, Cox DR, Lernian LS, Myers KM: Attachment of a 40-base-pair G + C-rich sequence (GC-clamp) to genomic DNA fragments by the polymerase chain reaction results in improved detection of single-base changes. Proc Natl Acad Sci USA 86:232-236, 1989 Lerman LS, Silverstoin K: Computational simulation of DNA melting and its application to denaturing gradient gel electrophore.iiis. In: Methods in Eiizyimloxy. Wu R (ed.). Academic Press, N e w York, 1987, pp 482-501 Myers RM, Maniati.s T, Leniian LS: Detection and localization of single base changes hy denaturing gradient gel electrophoresis. In: Methods iu Eu::yinoloiiy. Wu R (eds.). Academic Press, N e w York, 1987, pp 501-527 Gyllensten UB, Erlich HA: Generation of single-strand DNA by the polymerase chain reaction and its application to direct sequencing ofthe HLA-DQA locus. Proc Natl Acad Sci USA 85:7652-7656, 1988 Lathrop MG, Lalouel JM: Easy calctihitions of lod scores ,ind genetic risks on small computers. .4m / Hntn Genet 3r>:460—465, 1984 Terwilliger JD, Ott J: Haiidlionk of Human Genetic Liukiige. Johns Hopkins University Press, Baltimore, 1994, pp 307

SKIN DISEASES MUTATION DATABASE The Joimtctl of hu'cstificttivc Dcrmatolofiy has taken the initiative to develop a computerized database for skin diseases mutations. Tentatively, this service will be available through Internet during the early part of 1995, The individuals listed below have agreed to serve as coordinators ofthe respective area of heritable skin diseases. We would encourage researchers working in the areas indicated to contact coordinators to obtain forms for data submission. The information required brief description of the clinical features, methods used to verify the diagnosis, description of the mutation (tbe gene, nucleotide changes, consequence of the nuitation), and possibly a reference to publication describing the mutation. These forms should be submitted to the coordinating investigators for verification of the completeness and accuracy of the data. The first version of the database will be published in one of the forthcoming issues of the Journal. W e hope that tbis service will facilitate elucidation of heritable disorders affecting the skin. Organizing Committee: Ervin H. Epstein, Jr,, Lowell A, Goldsmith, and Jouni Uitto

Data Coordinators on Clinical Areas Covered by the Skin Diseases Mutation Database Diseases/Mutated Gene-Protein Systems

461

Data Coordinator (Fax)

1, Dystropliie EU/COL7A1

Angela M, Christinno (215-955-5 788)

2, EB Simplex, EHK, Ielithyoses, Keratinization Disorders/Keratins, Transglutaniinases

John Compton (301-402-2886)

3, Jtinetional EB/Laminin 5, a6/34 Integrin, B1'AG2

Guerrino Menegiizzi (33-93-81-14-04)

4, Heritable Conneetive Tisstie Disorders (EDS, CL, 1'XE)/COL1A1, ELN, FBN-1 & 2, ete,

Heather YeowcU (919-684-3002)

5, Pigmentation Disordcts/Tytosinase, eKit, ete.

Riehard Spritz (608-262-2976)