A comprehensive approach to identification of pathogenic FANCA variants in Fanconi anemia patients and their families

Abstract

Fanconi anemia (FA) is a rare recessive DNA repair deficiency resulting from mutations in one of at least 22 genes. Two-thirds of FA families harbor mutations in FANCA. To genotype patients in the International Fanconi Anemia Registry (IFAR) we employed multiple methodologies, screening 216 families for FANCA mutations. We describe identification of 57 large deletions and 261 sequence variants, in 159 families. All but seven families harbored distinct combinations of two mutations demonstrating high heterogeneity. Pathogenicity of the 18 novel missense variants was analyzed functionally by determining the ability of the mutant cDNA to improve the survival of a FANCA-null cell line when treated with MMC. Overexpressed pathogenic missense variants were found to reside in the cytoplasm, and nonpathogenic in the nucleus. RNA analysis demonstrated that two variants (c.522G > C and c.1565A > G), predicted to encode missense variants, which were determined to be nonpathogenic by a functional assay, caused skipping of exons 5 and 16, respectively, and are most likely pathogenic. We report 48 novel FANCA sequence variants. Defining both variants in a large patient cohort is a major step toward cataloging all FANCA variants, and permitting studies of genotype-phenotype correlations.

Keywords: FANCA; Fanconi anemia; functional assay; pathogenic mutations; recessive disorder.

Figure 1

A schematic presentation of 216 FA families screened for FANCA mutations with their complementation and mutation status after completion of the study. The 40 patients with one FANCA mutation known is based on previously performed screening for a couple of very common FA mutations or deletions by MLPA.

Figure 2. Functional assays of FANCA missense variants to determine pathogenicity

A. MMC sensitivity in FANCA-null RA3087/E6E7/hTERT cell lines expressing indicated FANCA variants. The two graphs represent two separate experiments. Empty vector (EV), wild type (WT) FANCA, and FANCA mutant K522R was included in both experiments. Indicated cell lines were treated with increasing doses of MMC and total cell number was counted after 6–8 days. Values were normalized using untreated control to determine the percent survival of cells expressing the indicated variants. Cell lines for survival assays were plated in triplicate; error bars represent standard deviation. B. Immunoblot with anti-HA antibody to test expression levels of HA-tagged FANCA variants which were transduced into FANCA-null RA3087/E6E7/hTERT fibroblasts. The FANCA-null RA3087 cell line was derived from a FA patient with biallelic FANCA deletions (delEx9-43). C. Immunofluorescence to detect HA-tagged FANCA expressed in FANCA-null RA3087/E6E7/hTERT fibroblasts. Localization is categorized as being nuclear or cytoplasmic.

Figure 3. Splicing errors from analysis of RNA from patient cell lines harboring variants predicted to cause nonsynonymous and intronic indel variants

A. cDNA was prepared from the patient (FAM117) LCL cell line harboring the c.522G>C (predicted p.Q174H) mutation affecting the last base of exon 5. The cDNA region spanning exons 1–6 was amplified using primers located in exons 1 and 7 and sequenced. Image shows the RT-PCR products including a WT (686bp) and a fainter smaller product indicated with an asterisk (590bp). Sequence of the total cDNA (top right) indicates presence of the WT product and the product resulting from skipping of exon 5. A representative cloned cDNA with exon 5 skipping is shown (bottom right). B. cDNA from the patient (FAM89) fibroblast cell line harboring the c.1565A>G, (predicted p.K522R) mutation affecting the second to last base of exon 16. The cDNA region spanning exons 15–22 was amplified using primers located on the flanking exons 14 and 23 and sequenced. Image shows RT-PCR products including a WT (777bp) and a smaller product indicated with an asterisk (681bp). Sequence of the total cDNA (top right) indicates presence of the WT product and the product resulting from skipping of exon 16. A representative cloned cDNA with exon 16 skipping is shown (bottom right). C. cDNA prepared from patient (FAM8) LCL cell line carrying the c.2602-9_2602-8delCT mutation in the pyrimidine track of intron 27 was tested for splicing errors. RT-PCR products spanning exons 27–31 were generated using primers from the flanking exons 26 and 32. Sequencing the RT-PCR product revealed skipping of exons 28–30, which appears to be predominant (band indicated with an asterisk on the image and in the top sequence tracing; reverse sequence). RT-PCR and sequencing of the region spanning exons 23–29 region revealed exon 29 skipping (bottom sequence tracing; reverse sequence); the exon 28–30 skipping is not seen in this sequence due to reverse primer binding within exon 30.

Figure 4. Analysis of FANCA expression in FANCA patient cell lines

Expression of the endogenous FANCA was analyzed in patient-derived LCL (A) and fibroblasts (B). Cells were grown in the presence or absence of 1µM mitomycin C (MMC) for 24 hours. A cell line from a normal individual (WT) is a positive control. RA3087 fibroblasts derived from a patient with known biallelic null mutations in FANCA is a negative control. Cell line IDs (RA numbers), and family IDs (FAM numbers) are on top and the mutations present in the cell lines are indicated below the image. For the three patients with no FAM ID, only one FANCA variant is known and the presence of a second mutation has not been determined (ND). The variants denoted with * cause in-frame exon skipping.

Figure 5. Mapping of all the coding and splicing variants identified in this study of FANCA and its encoded protein

A. FANCA protein diagram displaying the distribution of mutations. Novel mutations extend upward and known mutations extend downward. Nuclear localization signal (NLS) at aa 18–34; FANCG binding domain (FANCG) at aa 18–29; Region of reported interaction w/BRCA1 at aa 740–1083; FAAP20 binding domain (FAAP20) at aa 1095–1200; phosphorylation sites, serine residues at aa 849, 850, 858, and 1449 (http://atlasgeneticsoncology.org/Genes/FA1ID102.html). B. FANCA diagram displaying the distribution of intronic splicing mutations. Exons are represented by vertical lines and not drawn to scale. Novel splicing mutations extend upwards and the known splicing mutations extend downwards.