H. Tamary, Institute of Haematology–Oncology, Schneider Children's Medical Centre of Israel, 14 Kaplan Street, Petah Tiqva 49 202, Israel. E-mail: email@example.com
Fanconi anaemia (FA) is a genetically heterogeneous disease with at least eight complementation groups (A–H). In the present study, we investigated the molecular basis of the disease in 13 unrelated Israeli Jewish (non-Ashkenazi) patients with FA. All 43 exons of the Fanconi anaemia A (FANCA) gene were amplified from genomic DNA and screened for mutations by single-strand conformation polymorphism and DNA sequencing. We identified four ethnic-specific mutations: (1) 2172–2173insG (exon 24), the first ‘Moroccan mutation’; (2) 4275delT (exon 43), the second ‘Moroccan mutation’; (3) 890–893del (exon 10), the ‘Tunisian mutation’; and (4) 2574C > G (S858R), the ‘Indian mutation’. The tetranucleotide CCTG motif, previously identified as a mutation hotspot in FANCA and other human genes, was found in the vicinity of 2172–2173insG and 890–893del. According to our study, the four mutations account for the majority (88%) of the FANCA alleles in the Israeli Jewish (non-Ashkenazi) FA population. A screening of 300 Moroccan Jews identified three carriers of the first ‘Moroccan mutation’, but we did not find any carrier of the second ‘Moroccan mutation’ among 140 Moroccan Jews, nor any carrier of the ‘Tunisian mutation’ among 50 Tunisian Jews. Two ‘Indian mutation’ carriers were identified among 53 Indian Jews. All carriers within each ethnic group had the same haplotype, suggesting a common founder for each mutation.
Fanconi anaemia (FA) is a rare autosomal recessive disease characterized by multiple congenital abnormalities, bone-marrow failure and susceptibility to cancer (Auerbach et al, 1997; Alter & Young, 1998). The diagnosis of FA exploits the sensitivity of FA cells to bifunctional alkylating agents, resulting in highly elevated chromosomal breakage induced by cross-linking agents such as diepoxybutane (DEB) (Auerbach, 1993). FA is genetically heterogeneous, with at least eight complementation groups (A–H), each presumably corresponding to a separate disease gene (Joenje et al, 1997). The Fanconi anaemia group C (FANCC) gene, which maps to chromosome 9q22.3, was the first FA gene to be cloned (Strathdee et al, 1992). Recently, the FANCA, FANCG and FANCF genes were cloned (The Fanconi Anemia/Breast Cancer Consortium, 1996; Lo Ten Foe et al, 1996; de Winter et al, 1998,2000). FANCA[Mendelian inheritance in man (MIM) no. 227650] has an open reading frame of 4365 bp encoding a protein of 1455 amino acids and is localized on chromosome 16q24.3. The gene is composed of 43 exons spanning about 80 kb of genomic DNA (Ianzano et al, 1997). The prevalence of the FA-A subtype is estimated to be 60–66%.
In the present study, we investigated the molecular basis of FA in 13 unrelated Israeli non-Ashkenazi Jewish FA patients (including one compound heterozygote of Ashkenazi and non-Ashkenazi extraction). We found that 88% of the FA alleles bear four ethnic-specific mutations.
PATIENTS and METHODS
Patients and controls The study group consisted of 13 unrelated Israeli Jewish (non-Ashkenazi) FA patients. Diagnosis in all homozygous patients was confirmed by the DEB test. In previous studies, patients of Ashkenazi Jewish extraction (seven alleles) were found to be carriers of the IVS4+ 4A > T mutation in the FANCC gene (Verlander et al, 1995). The ethnic origin of our 13 patients, determined according to parents' country of birth, was as follows: seven Moroccan-Jewish, two Indian-Jewish, one Tunisian Jewish, one Iraqi-Jewish and two compound origin (Tunisian/Moroccan and Ashkenazi/Egyptian) (Table I). Ashkenazi Jews were defined as originating in European countries. To screen for FANCA mutations (see below) in several Jewish populations living in Israel, blood samples of patients consecutively hospitalized at the Rabin Medical Centre were obtained. The study was approved by the centre's Human Subject Committee.
Table I. FANCA mutations by ethnic origin in unrelated non-Ashkenazi Jewish patients.
Number of FA alleles
% mutations identified
Other (Iraqi, Egyptian)
Clinical data The clinical data were obtained by reviewing the medical records from the different co-operating medical centres in Israel.
DNA extraction After obtaining informed consent, blood samples were drawn from the patients and family members, and genomic DNA was isolated from white blood cells by salting out (Miller et al, 1988).
Amplification of FANCA gene exons We used previously described primers for amplification of the coding regions and exon-intron junctions of the FANCA gene (Levran et al, 1997). Amplification reactions were carried out in a 50 μl sample containing 150–300 ng of genomic DNA, 10 mmol/l Tris HCl (pH 8·3), 50 mmol/l KCl, 1·5 mmol/l MgCl2, 0·25 mmol/l spermidine, 0·01% (w/v) gelatin, 0·2 mmol/l of each deoxyribonucleotide (dATP, dCTP, dGTP, and dTTP), 0·2 μmol/l of each primer and 0·75 U Taq polymerase. After an initial incubation of 3 min at 96°C, the samples were subjected to 30 cycles of denaturation at 94°C for 30 s, annealing at 56°C− 64°C for 1 min and primer extension at 72°C for 1 min in a DNA thermal cycler (Perkin-Elmer Cetus, Emeryville, CA, USA).
Single-strand conformation polymorphism (SSCP) All FANCA exons were screened by SSCP for unknown mutations (Orita et al, 1989). Samples of polymerase chain reaction (PCR) products were diluted in 0·05% sodium dodecyl sulphate (SDS), 5 mmol/l EDTA and a loading dye containing formamide, denaturated at 95°C for 180 s, cooled on ice and loaded onto 0·6 X mutation detection enhancement (MDE) gel (FMC Bioproducts, Rockland, ME, USA). Electrophoresis was performed in 1× Tris-borate-EDTA (TBE) buffer at room temperature or 4°C for 16–22 h, followed by silver staining of the gels.
DNA sequencing Amplified exons showing an aberrant pattern on SSCP gels were subjected to direct sequencing. Amplified fragments were purified using the Wizard PCR purification system (Promega, Madison, WI, USA) and then sequenced using the Sequenase [33P]-terminator cycle sequencing (Amersham Life Science, Buckinghamshire, UK). After the addition of stop solution, the products were heated at 95°C for 2·5 min, cooled on ice and loaded onto a 5% long-ranger polyacrylamide gel (FMC). Electrophoresis was performed in 1× TBE buffer at 57 W.
Restriction digests To confirm the results obtained by sequencing and to screen for mutations, the relevant PCR fragments were digested with AatII or PvuII (New England Biolabs, Beverly, MA, USA) according to the manufacturer's instructions. Digested DNA was electrophoresed on 3% NuSieve 1% agarose gel.
SSCP studies showed aberrant migration patterns for four FANCA exons. Mutations revealed by sequence analysis were: 890–893del (exon 10); 2172–2173insG (exon 24); 2574C→G (exon 27); and 4275delT (exon 43). The ethnic origins of the mutations are summarized in Table I.
2172–2173insG (exon 24) – the first ‘Moroccan mutation’
As 2172–2173insG creates a frameshift at codon 725, with the insertion of 68 new amino acids before the stop codon, a truncated protein of 792 amino acids is expected. This mutation can be detected by AatII restriction digest. PCR amplification of a 238 bp fragment containing exon 24, followed by digestion with AatII, yielded two fragments in the wild-type sequence of 91 bp and 147 bp. The G insertion abolished the restriction site between these two fragments, giving rise to a fragment of 238 bp instead. Analysis of DNA samples of additional homozygotes revealed only the 238 bp fragment, whereas heterozygotes displayed the 91 bp, 147 bp and 238 bp fragments (Fig 1A). The more frequent ‘Moroccan mutation’ was found in ten of the 13 Moroccan-Jewish FA alleles (76·9%) examined. Three carriers were identified among 300 healthy Moroccan–Jewish individuals (carrier frequency 1:200). Thirteen different haplotypes were identified among 38 healthy Moroccan Jews using five polymorphic sites (IVS7-12 A/G, 1501 A/G, IVS18 + 82 T/C, 2426 G/A and 3654 A/G). All patients were homozygous for the same haplotype (AGTGA), which was found in 40·8% of the healthy alleles. The other 12 haplotypes had frequencies of 1·3–11·8% (data not shown).
4275delT (exon 43) – the second ‘Moroccan mutation’
The 4275delT creates a frameshift in exon 43, with the addition of six amino acids after codon 1425 and a premature stop codon, predicting a truncated protein of 1431 amino acids. As there is no restriction enzyme that recognizes this mutation, it was detected by SSCP analysis (Fig 1B). The mutation was found in two patients: one homozygous and the other a compound heterozygote (4275delT/2172–2173insG). The mutation was not present among 140 Moroccan Jews. The mutation was found in the same haplotype (AGTGA) as the first ‘Moroccan mutation’, which is the most frequent haplotype among Moroccan Jews.
890–893del (exon 10) – the ‘Tunisian mutation’
The 890–893del (–GCTG, in exon 10), which is located at the 3' end of the exon, may be a splice-site mutation associated with a truncated protein. As no restriction enzyme recognizes this mutation, we used SSCP analysis (Fig 1C). The deletion was identified in two unrelated families of Tunisian origin. The mutation was detected in all four Tunisian–Jewish FANCA alleles and in none of 100 healthy Tunisian–Jewish alleles. Seven different haplotypes were found among 33 healthy Tunisian Jews. All the patients displayed the GATGA haplotype, which was found among 33% of the controls; the frequency of the other six haplotypes ranged between 4·5% and 28·8% (data not shown).
2574C→G (S858R, exon 27) – the ‘Indian mutation’
The missense mutation 2574C→G (S858R) in exon 27 is probably a disease-causing mutation as it segregated in our affected families and has also been previously described as a FANCA-causing mutation (Wijker et al, 1999). It can be detected by PuvII restriction digest. PCR amplification of a 284 bp fragment containing exon 27, followed by digestion with PuvII, yielded two fragments of 184 bp and 100 bp in the wild-type sequence. The presence of the mutation 2574C→G abolished the restriction site between the two fragments, giving rise to a fragment of 284 bp instead. Analysis of DNA samples of additional homozygotes revealed the 284 bp fragment, whereas heterozygotes displayed all three fragments (Fig 1D). The S858R mutation was found in all four Indian–Jewish FA alleles and in two of 106 healthy Indian–Jewish alleles (carrier frequency 1:53). Using the five polymorphic sites (see above), nine haplotypes were found among 41 healthy Indian–Jewish individuals. The patients were homozygous for the GATGA haplotype, which was present in 18·1% of the healthy Indian–Jewish alleles; the frequency of the other eight haplotypes ranged between 12% and 30·1% (data not shown).
Distribution of mutations among FANCA patients
FANCA mutations were identified in 88% of the Israeli–Jewish (non-Ashkenazi) FA alleles (Table I). The more frequent ‘Moroccan mutation’ was present among 10 (76·9%) of the 13 Moroccan FA alleles examined, while the three (23·1%) Moroccan alleles bore the less frequent mutation. All the FA Jewish–Indian alleles and all of the Tunisian–Jewish alleles bore the respective ethnic-related mutation. The mutations in two Iraqi–Jewish alleles and one Egyptian–Jewish allele were not identified.
Eighteen patients with FA, 12 Moroccan–Jewish patients, four Indian–Jewish patients, one Tunisian–Jewish patient and one compound Moroccan–Tunisian patient (Table II), were evaluated. Seven of the 11 Moroccan patients homozygous for the 2172–2173insG had congenital anomalies. The median age at onset of bone-marrow failure was 7 years (range, 1·5–25 years). Two of the homozygotes for the frequent ‘Moroccan mutation’ developed acute myeloblastic leukaemia (AML) at the ages of 18 years and 19 years, and one patient developed squamous cell carcinoma of the tongue at the age of 24 years. One patient homozygous for the second ‘Moroccan mutation’ (4275delT) had congenital anomalies, developed bone-marrow failure at the age of 11 years and is currently asymptomatic at age 28 years. The four Indian patients appeared to have a more severe form of the disease: all had congenital malformations and bone-marrow failure was diagnosed in three patients at 1–7 years of age. Two patients developed AML at age 7 years and 10 years.
Table II. Mutations and clinical data for non-Ashkenazi–Jewish patients.
In the present study, we investigated the molecular basis of FA in 13 unrelated Israeli–Jewish (non-Ashkenazi) patients. Using SSCP screening followed by DNA sequencing, we were able to identify four mutations: three of them were novel mutations (2172–2173insG in exon 24, 890–893del in exon 10, and 4275delT in exon 43) and the fourth (2574C→G; S858R in exon 27) was previously described in a German patient (Wijker et al, 1999).
The CCTG repeat sequence was observed in the vicinity of the two new mutations: 5 bp 5′ to the 2172–2173insG mutation and 7 bp 3′ to the 890–893del. In a survey of human gene deletions, direct repeats (2–6 bp) were found in the immediate vicinity of the majority of short deletions (Krawczak & Cooper, 1991). The CCTG sequence has been identified as a deletion hotspot in a variety of human genes, including WT1 (Huff et al, 1995) and globin genes (Krawczak & Cooper, 1991), and also in ∼50% of deletions/insertions in FANCA (Levran et al, 1997). Slipped-strand mispairing of short direct repeats during DNA replication has been suggested to be a major mechanism in the generation of gene deletions (Krawczak & Cooper, 1991).
In the present study, we found four mutations that account for the majority (88%) of the FANCA alleles in the Israeli–Jewish (non-Ashkenazi) FA population. The mutations were ethnic-specific: 2172–2173insG, the frequent ‘Moroccan mutation’, was detected in 10 of 13 Moroccan FA alleles; the less frequent ‘Moroccan mutation’ (4275delT) was detected in three Moroccan alleles; the ‘Tunisian mutation’, 890–893del, was found in all the Tunisian–Jewish FA alleles; S858R, the ‘Indian mutation’, was present in all the Indian–Jewish FA alleles. The frequency of the first ‘Moroccan mutation’ was relatively high (1:200). The ‘Indian mutation’ also appeared to be frequent (1:53), although only a relatively small number of unaffected Indian Jews were examined. The occurrence of each ethnic-specific mutation in an identical haplotype suggests a common founder for each mutation.
A review of the clinical picture of the homozygous patients suggests that the ‘Indian mutation’ is severe, similar to that described for the FANCC Ashkenazi–Jewish mutation (Gillio et al, 1997), whereas the more frequent ‘Moroccan mutation’ is associated with a milder disease with a lower frequency of malformations, and development of bone marrow failure and leukaemia at an older age. These findings should aid clinicians in making therapeutic decisions regarding FA patients with these mutations. However, as the number of patients was small (four Indian–Jewish patients, nine Moroccan–Jewish patients homozygous for the 2171–2173insG mutation), there is a need to study larger groups of patients to confirm this correlation. The basis for this phenotypic variability and the variability even within the same genetic defect, which has been previously described by Koc et al (1999), is unclear.
According to the results of our study, molecular diagnosis appears to be feasible for the majority (88%) of Israeli–Jewish (non-Ashkenazi) patients with FA. These findings enable carrier detection and prenatal diagnosis. Our data may also be helpful in designing a rational molecular diagnostic strategy for FA.
We thank Professor M. Shohat and Dr E. Feler for providing control DNA samples, and Dr A. Shabad for supplying clinical data on her patients. We are indebted to Dr A. Mahler for her skilful assistance in the preparation of this manuscript.
This work was supported in part by the National 1nstitutes of Health Grant HL32987 (A.D.A).