Work

Recurrent GATA 1 mutations in D iamond‐ B lackfan anaemia

Year: 2014

Type: article

Abstract: Diamond-Blackfan anaemia (DBA) is a congenital bone marrow failure syndrome characterized by normochromic macrocytic anaemia and reticulocytopenia with selective hypoplasia of erythroid precursors in the bone marrow (Vlachos et al, 2008). The disease usually presents in infancy and is often associated with physical anomalies. Heterozygous mutations in the RPS19 gene, encoding ribosomal protein S19, accounts for approximately 25% of cases (Draptchinskaia et al, 1999) and 10 additional ribosomal protein genes represent another 25% of cases (Raiser et al, 2014). Thus, the cause of DBA remains unresolved in a substantial proportion of patients despite a probable genetic cause of the disease. We investigated a Swedish family with three brothers affected by anaemia recognized within the first 3 months of life. The three siblings presented with similar haematological features: macrocytic anaemia (Hb 50–90 g/l; mean cell volume 104–109 fl), haematocrit of 19–31%, low reticulocyte count (0–0·5%), erythroid hypoplasia in the bone marrow with otherwise normal cellularity, and a modest elevation of fetal haemogobin. All three brothers had neutrophil counts within the normal lower range (3–4 × 109/l) as well as normal platelet counts (275–440 × 109/l). The two elder brothers had increased erythrocyte adenosine deaminase levels and they initially responded to steroid treatments, however, both of them became refractory at age 20 and 28 years and they now require regular transfusions every 8–16 weeks. The youngest brother, now 30 years of age, has a persistent anaemia (80–95 g/l) without any treatment. None of the three brothers have physical abnormalities and their heights are within the normal range (−1 standard deviation). Based on repeated analysis of haematological parameters in the siblings DBA was considered the most likely diagnosis. Genetic analysis was initially performed on samples from one affected brother who was negative for mutations in the RPS19, RPL5, RPL11 and RPS26 genes. However, pedigree analysis revealed that the youngest brother has a different father suggesting the possibility of X-linked recessive inheritance through the healthy mother. Mutations in the GATA1 gene have previously been associated with X-linked familial dyserythropoietic anaemia and/or thrombocytopenia (Nichols et al, 2000). Furthermore, two recent reports presented three families with haematological characteristics of DBA associated with different mutations in GATA1 (Sankaran et al, 2012; Parrella et al, 2014). These findings, in combination with a pattern consistent with X-linked inheritance in our family, prompted us to analyse the coding regions of the GATA1 gene in available family members. Sequencing revealed a c.220G>C transversion in the exon 2 donor splice site. Interestingly, the mutation was identical to that reported previously in a family with impaired erythropoiesis (Hollanda et al, 2006) as well as in two brothers with DBA (Sankaran et al, 2012). The c.220G>C transversion causes skipping of exon 2 and results in loss of the long isoform of GATA1 (Nichols et al, 2000). Similarly, the other and previously reported GATA1 mutations predict loss of the long GATA1 isoform. The importance of GATA1 in haematopoiesis is well documented (Bresnick et al, 2010; Kaneko et al, 2010), particularly as a driver of erythropoiesis at the proerythroblast stage. Consequently, it may be hypothesized that the loss of the long GATA1 isoform and haploinsufficiency for ribosomal subunits interfere with similar pathways or mechanisms during erythropoiesis. To clarify if the mutation c.220G>C was recurrent or shared by descent in our family and the family reported by Sankaran et al (2012), we performed haplotype analysis around GATA1 on probands from each of the two families. We sequenced a 2 kb stretch of genomic DNA immediately flanking the 5′ and the 3′ regions of GATA1 respectively, in search for non-coding sequence differences in the two individuals. Sequence analysis revealed two single nucleotide differences [rs113966884 (c.-183G>A) and a novel variant (c.*391C>T)] on each side of the GATA1 gene when comparing the probands (Fig 1A). The nucleotide positions are located at positions g.48,644,890 and g.48,652,960 (GRCh37.p10), immediately flanking the GATA1 gene. Thus, the two nucleotide positions spanning the GATA1 gene indicate different haplotypes in the two probands (Fig 1B). This led us to conclude that the c.220G>C mutation had occurred independently in these two families. This is the fourth family reported with clinical findings consistent with DBA and a GATA1 gene mutation. Intriguingly, two of these families have an identical splice site mutation and all three mutations result in the loss of the long GATA1 isoform. The DBA diagnosis is based on clinical findings and may be confirmed by a ribosomal protein gene mutation in approximately 50% of cases (Farrar & Dahl, 2011). When no mutation is identified, DBA is diagnosed based on haematological and clinical findings consistent with established criteria (Vlachos et al, 2008). The observation of GATA1 mutations in a few cases consistent with the current diagnostic criteria for DBA, suggest that the primary cause of the disease is not necessarily confined to a ribosomal protein gene or ribosome biogenesis. In conclusion, our observation, together with two previous reports (Sankaran et al, 2012; Parrella et al, 2014) show that specific GATA1 mutations may be associated with DBA or DBA-like features, albeit in a small subgroup of affected males, that can be considered in familial cases consistent with X-linked inheritance. The authors would like to thank the participating family members. This work was supported by grants from the Swedish Research Council (K2013-66X-10829-20-3) and Uppsala University Hospital. J.K is supported by the Swedish Society for Medical Research. JK performed the research and analysed the data together with AK and PSA. HTG contributed essential reagents. ND designed the study and wrote the paper.

Source: British Journal of Haematology

DOI: 10.1111/bjh.12919

Authors: Joakim Klar, Ayda Khalfallah, Pakeeza Shaiq Arzoo, Hanna T. Gazda, Niklas Dahl