Systematic Analysis of Known Candidate Genes in 58 Probands with Previously Uncharacterized Congenital Sideroblastic Anemia: Evidence for Genetic Heterogeneity and Identification of Novel Mutations in ALAS2 and PUS1

Sideroblastic anemias are a heterogeneous group of congenital and acquired bone marrow disorders characterized by pathologic iron accumulation in the mitochondria of erythroid precursors. Whereas the molecular ontology of acquired sideroblastic anemia is largely obscure, the genetic etiology is known for many patients with congenital sideroblastic anemia (CSA). Mutations in the heme biosynthetic enzyme 5-aminolevulinate synthase 2 gene (ALAS2)—associated with the classic, X-linked form of the disease—or mitochondrial DNA mutations/deletions (as in Pearson syndrome) have been reported in many cases. Rare cases are associated with mutations in the genes encoding a thiamine transporter (SLC11A2), pseudouridine synthase 1 (PUS1), a mitochondrial ATP-binding cassette transporter (ABCB7), and glutaredoxin 5 (GLRX5). Nonetheless, the molecular defect in a majority of cases of CSA remains unknown and there appears to be a significant bias in underreporting those cases that go genetically uncharacterized. Given this, we undertook a systematic genetic investigation of a large series of previously unreported CSA patients.

METHODS AND RESULTS: 58 probands (26 female, 32 male) with CSA were studied. This cohort was comprised of 49 singleton patients and 10 families, including 8 families with at least 2 affected full siblings (1 set of monozygotic male twins, 1 each male or female sibling pairs, 2 male/female sibling pairs, 1 set each of 3 female or 3 male siblings, and one set of 3 siblings of unknown sex), and 2 families in which there was a history of CSA in a prior generation. In the great majority, CSA was the only clinical phenotype. We employed a sequential sequencing strategy to identify variants in ALAS2, PUS1, and GLRX5; ABCB7, SLC19A2, and mitochondrial DNA variants were not included in our analysis due to the absence of associated syndromic phenotypes in the cohort. To begin, ALAS2 was characterized by sequencing the coding region, intron-exon boundaries, proximal promoter, and intron 8, which contains an erythroid-specific enhancer, in all subjects. We detected 7 different missense mutations in 8 of the 58 probands, each of which was a singleton patient. No ALAS2 mutations were found in the promoter or intron 8. Two of the ALAS2 mutations were novel: V301A in a male patient and R517G in a female patient with a skewed X chromosome inactivation pattern in peripheral blood leukocytes as assessed by the human androgen receptor assay (HUMARA). The coding regions and intron-exon boundaries of PUS1 and GLRX5 were sequenced in all probands lacking ALAS2 mutations. Among these, a novel homozygous null (Q154X) PUS1 mutation was found in one familial case; only known polymorphisms in GLRX5 were detected. In 48 probands, including the 9 remaining familial cases, we did not identify a disease-causing mutation. However linkage studies in three families with affected males only were consistent with linkage to the ALAS2 locus in each case. CONCLUSIONS: These results demonstrate that:

1. ALAS2 mutations are commonly associated with CSA,

2. ALAS2 mutations and non-random X inactivation in affected females underscore the importance of ALAS2 analysis in women as well as in men with CSA,

3. many potentially X-linked cases of CSA may have either ALAS2 mutations not detected by conventional sequencing approaches or mutations in other X-linked genes,

4. PUS1 and GLRX5 variants are unusual causes of CSA, and

5. there is strong genetic evidence that other, autosomal recessive forms of CSA exist.