Research Article: Diagnosis of Fanconi Anemia: Mutation Analysis by Next-Generation Sequencing

Date Published: June 3, 2012

Publisher: Hindawi Publishing Corporation

Author(s): Najim Ameziane, Daoud Sie, Stefan Dentro, Yavuz Ariyurek, Lianne Kerkhoven, Hans Joenje, Josephine C. Dorsman, Bauke Ylstra, Johan J. P. Gille, Erik A. Sistermans, Johan P. de Winter.


Fanconi anemia (FA) is a rare genetic instability syndrome characterized by developmental defects, bone marrow failure, and a high cancer risk. Fifteen genetic subtypes have been distinguished. The majority of patients (≈85%) belong to the subtypes A (≈60%), C (≈15%) or G (≈10%), while a minority (≈15%) is distributed over the remaining 12 subtypes. All subtypes seem to fit within the “classical” FA phenotype, except for D1 and N patients, who have more severe clinical symptoms. Since FA patients need special clinical management, the diagnosis should be firmly established, to exclude conditions with overlapping phenotypes. A valid FA diagnosis requires the detection of pathogenic mutations in a FA gene and/or a positive result from a chromosomal breakage test. Identification of the pathogenic mutations is also important for adequate genetic counselling and to facilitate prenatal or preimplantation genetic diagnosis. Here we describe and validate a comprehensive protocol for the molecular diagnosis of FA, based on massively parallel sequencing. We used this approach to identify BRCA2, FANCD2, FANCI and FANCL mutations in novel unclassified FA patients.

Partial Text

Fanconi anemia (FA) is a recessive chromosomal instability syndrome with diverse clinical symptoms and a high risk for acute myeloid leukemia and squamous cell carcinoma of the head and neck region [1]. Clinical suspicion of FA is mostly based on growth retardation and congenital defects in combination with life-threatening bone marrow failure (thrombocytopenia and later pancytopenia), which usually starts between 5 and 10 years of age. However, the clinical manifestations of FA patients are highly variable, and therefore the FA diagnosis should be confirmed by a positive chromosomal breakage test and/or pathogenic mutations in one of the FA genes. Currently, mutations in 15 different genes are known to cause FA, and their gene products act in a pathway that takes care of specific problems that may arise during the process of DNA replication [2].

We designed an in-solution oligonucleotide hybridization capture kit (SureSelect, Agilent) targeting the open reading frames of all FA genes, except for regions that contain repetitive and low complexity DNA sequences as assessed by RepeatMasker ( All exons-, 3′-, and 5′-UTR-regions, and exon-intron boundaries were targeted by this approach (the oligonucleotide locations, in a  .BED format, are available upon request). In addition, a number of other genes involved in cancer predisposition and routinely screened in our diagnostic lab were included in the enrichment kit. We used the Illumina GAIIx platform for sequencing.

Sequence data from DNA libraries of eleven carriers of mutations in the FA genes: FANCA (4), FANCB (1), FANCC (1), FANCD1 (1), FANCE (1), FANCG (1), FANCI (1), FANCN (1), and one individual carrying a mutation in BRCA1, were generated from an Illumina GAIIx sequencer. An average of 2.8 million unique reads were obtained per library resulting in a median sequence depth of about 100 fold, with an average enrichment efficiency of >75% (Figure 1). Several types of disease-causing genetic aberrations were present in the assayed DNA samples including single nucleotide substitutions, small deletions (1–8 nucleotides), and large deletions (multiple exons). We developed a variation detection pipeline detecting all these types of aberrations.

The mutation detection strategy described here proved to be efficient for the molecular diagnosis of FA although patients with mutations in FANCF, -J, -M, -O, and -P were not included in our study. All the FA mutations identified by Sanger sequencing were also detected by next-generation sequencing. Moreover, we discovered a novel large deletion in the FA-I patient, for whom only one truncating mutation was previously identified [4]. The exact breakpoint at nucleotide level could be distinguished as also the intronic regions of the FA genes were enriched and sequenced. We confirmed the deletion by PCR and Sanger sequencing, using a SNP in the last unaffected exon and two sets of primers amplifying the regions up- and downstream of the breakpoint (Figure 3). Besides this novel large deletion we identified large genomic deletions in FANCA (2 samples) and PALB2 (1 sample). The sensitivity of the method was demonstrated by mixing two DNA samples prior to library preparation and enrichment. A deletion of one allele in a background of four alleles can be detected, suggesting that the method is even applicable for the classification of mosaic FA patients. However, a thorough assessment of the method using serial dilutions with samples harboring large deletions is required to determine the detection limit of the assay. The identification of large deletions is essential for FA molecular diagnostics since about 40% of pathogenic mutations in the major FA complementation group, FA-A, are caused by large deletions in FANCA. As large deletions have also been demonstrated for FANCI (this study) and FANCN [6], it is plausible that these types of aberrations are present in other FA samples, which were previously unclassified by conventional molecular screening methods. Therefore, these types of mutations should be examined in the standard molecular diagnostics of FA.