Dyserythropoietic anaemia with an intronic GATA1 splicing mutation in patients suspected to have Diamond‐Blackfan anaemia

Abstract Diamond‐Blackfan anaemia (DBA) shares clinical features with two recently reported sporadic cases of dyserythropoietic anaemia with a cryptic GATA1 splicing mutation (c.871‐24 C>T). We hypothesized that some patients clinically diagnosed with DBA but whose causative genes were unknown may carry the intronic GATA1 mutation. Here, we examined 79 patients in our DBA cohort, who had no detectable causative genes. The intronic GATA1 mutation was identified in two male patients sharing the same pedigree that included multiple cases with anaemia. Cosegregation of this mutation and disease in multiple family members provide evidence to support the pathogenicity of the intronic GATA1 mutation.


INTRODUCTION
Diamond-Blackfan anaemia (DBA) is one of the inherited bone marrow failure syndromes (IBMFSs), characterized by macrocytic pure red cell aplasia, variable malformations and a predisposition to malignancies [1]. It generally presents in the first year of life. DBA patients generally exhibit increased levels of foetal haemoglobin (HbF). The activity of erythrocyte adenosine deaminase (eADA) and the concentration of erythrocyte reduced glutathione (GSH) are elevated in 75%-90% and 59.1% of DBA patients, respectively [1][2][3]. Nearly 60% of DBA patients will respond to steroids over the long term [1]. To date, 22 ribosomal protein (RP) genes have been reported as causative genes of DBA in close to 70% of patients [1,[4][5][6][7]. All known DBA-causative mutations involve RP genes, except for rare germline GATA1 and TSR2 mutations [8,9]. Missense mutations in GATA1 cause DBA, congenital dyserythropoietic anaemia, thalassemia and a variety of thrombopoietic defects [10]. Recently, Abdulhay et al. reported two sporadic dyserythropoietic anaemia cases with an intronic mutation in GATA1 (c.871-24 C>T).
This mutation results in reduced canonical splicing and increased use of an alternative splice acceptor site that causes a partial intron retention event, leading to a five amino acids insertion at the C-terminal zinc finger domain of GATA1 [11]. The resultant altered GATA1 protein has no observable activity. These cases and DBA patients share common clinical features including the early onset of severe macrocytic anaemia with an inadequate reticulocyte response, elevated HbF and high eADA levels. Therefore, we hypothesized that some patients with a clinical diagnosis of DBA whose causative genes were unknown may harbour the intronic mutated GATA1.

RESULTS AND DISCUSSIONS
Our DBA cohort consisted of 215 individuals, including 109 males.
Among them, we examined the intronic GATA1 mutation by direct sequencing analysis in 79 individuals (including 47 males), for whom the causative gene could not be identified by target and whole-exome sequencing (Table S1) (Table S2). In both of our cases, their siblings died of hydrops foetalis, and one of the previously reported cases also received intrauterine transfusion for hydrops foetalis [11], suggesting that male individuals with this mutation have a higher incidence of hydrops foetalis and should be given more attention in perinatal management.
However, there are some differences between our cases and the individuals from the previous report. For example, in our cases, anaemia improved in response to corticosteroids, whereas the previous cases recovered spontaneously from severe anaemia without medication.
Although the mechanism of steroid effect is unknown, steroid therapy Notably, the pedigree included three older members (I-2, II-1, 3) who had been diagnosed with aplastic anaemia ( Figure 1A). We analysed the genotype of the mutant allele of four family members whose samples were available (mother of individual 35: III-10, and parents and sister of individual 36: III-6, 7, IV-7). Three female members had the GATA1 mutation heterozygously, and the remaining male member (father of individual 36: III-6) had a wild-type allele hemizygously (Figure 2A).
To confirm the effect of the mutation, we performed semiquantitative reverse transcription PCR (RT-PCR) analyses using total ribonucleic acid (RNA) purified from the PB of individual 35 and his mother as well as a healthy control, using previously reported primers ( Figure 2B) [11].
We found a longer GATA1 transcript that retained the 15 nucleotides It is difficult to predict which intronic base substitutions will cause splicing abnormalities. We investigated the possibility that the muta-tions around the substitution sites found in this study could cause similar splicing abnormalities using splice site prediction software, Net-Gene2 [12,13]. We found that c.871-24 C>T increased the acceptor prediction score from 0.054 to 0.067. However, there was no change in either c.871-24 C>A or C>G. When various mutations were entered, the expected score was shown to increase to 0.1 for c.871-23 A>T and c.871-25 A>T (Table S3). Because this change was larger than for c.871-24 C>T, RT-PCR analysis was performed with these mutations in the minigene construct as described previously [14]. As expected, c.871-24 C>A and C>G did not cause splicing abnormalities, and c.871-23 A>T showed splicing abnormalities with slightly lower efficiency than c.871-24 C>T. However, c.871-25 A>T showed no abnormal transcripts, contrary to our expectation. These results indicate that similar splicing abnormalities may be caused by c.871-23 A>T ( Figure 2C).
Interestingly, there are three older members with aplastic anaemia in this pedigree. Although we could not obtain samples for further study, they may have heterozygous or hemizygous intronic GATA1 mutation ( Figure 1A). It is known that IBMFSs patients develop aplastic anaemia and have a predisposition for malignancies; including myelodysplastic syndrome (MDS) or acute myeloid leukaemia [15]. This study indicates that even if an individual carries this GATA1 mutation heterozygously, the possibility of developing aplastic anaemia or MDS at an older age might not be excluded.
In conclusion, we examined the intronic GATA1 mutation by direct sequencing analysis of 79 patients without known DBA causative genes in our DBA cohort. We detected the intronic GATA1 mutation in two male patients in a family with multiple individuals affected by anaemia, sharing many clinical characteristics with two individuals from a previous report [11]. Cosegregation of this mutation and disease in multiple family members provided supporting evidence of the pathogenicity. These results further support the concept that dyserythropoietic anaemia with the intronic GATA1 mutation is a distinct clinical entity and demonstrate the usefulness of screening for the intronic GATA1 mutation in the diagnostic evaluation of congenital anaemia.