Abstract
Approximately one quarter of cases of hereditary spherocytosis exhibit autosomal recessive inheritance. Patients with recessive hereditary spherocytosis (rHS) are more severely affected than patients with typical, dominant HS. Many rHS patients present in infancy with life-threatening hemolytic anemia; some are transfusion-dependent (TD). Erythrocytes from most rHS patients are spectrin deficient, leading to destabilization of the lipid bilayer. In most cases, the genetic defects in the SPTA1 gene leading to rHS are unknown. We studied 19 families with rHS suspected by clinical and laboratory data including 5 of the original rHS kindreds described by Agre et al. Probands from 9 of the 19 families were severely affected, demonstrating life-long transfusion dependence. After obtaining a genetic diagnosis, two probands were splenectomized and became transfusion independent. In the remaining 10 families, most patients had a history of receiving blood transfusions, particularly during infancy. Most underwent early splenectomy that ameliorated but did not cure their anemia. Hematocrits of affected patients postsplenectomy ranged from 21-51%. In untransfused patients, biochemical analyses included SDS-PAGE of erythrocyte membrane proteins. Quantitative analyses of spectrin content, measured by spectrin/band 3 ratios, demonstrated spectrin deficiency from 43-78% in rHS erythrocyte membranes. Genetic studies included Sanger sequencing (6 families) and/or whole exome analyses (WES, 13 families). Genomic DNA was amplified using primers flanking the 52 coding exons and promoter region of the SPTA1 gene, followed by Sanger sequencing analysis of the amplicons. Alternatively, genomic DNA was subjected to WES via targeted exon capture followed by ultrahigh throughput sequencing of captured DNA. Sequencing data were mapped and variants called by the GATK algorithm. Variants were filtered then further assessed via conservation and mutation prediction programs. Numerous unique null SPTA1 alleles were found. Two TD patients had deleterious mutations in both SPTA1 alleles; one with nonsense mutations in trans died of liver failure associated with transfusion-related iron overload, the other with nonsense and splicing mutations in trans remains transfusion dependent. Additional null alleles included 6 nonsense, 4 splicing, and 3 frameshift mutations. These null alleles were frequently in trans to novel missense mutations. Eight unique missense mutations (in spectrin repeats 2, 4 and 14) were identified in conserved residues and all were predicted to be deleterious by mutation prediction algorithms. Two missense mutations were found in multiple families (repeat 14 in 4 families, repeat 2 in 2 families). Two TD patients were homozygous for missense mutations located in critical residues of the alpha-beta spectrin self-association domain. One of these patients had a sibling homozygous for the same mutation die in the perinatal period due to complications of anemia. The repeat 2 mutation was found in the heterozygous state in 3 other rHS kindreds in trans to 3 different null alleles. Probands from two kindreds suffered from homozygous missense mutations in conserved residues that were predicted to be deleterious by mutation prediction algorithms (repeat 2 and repeat 14). In 7 alpha-spectrin linked rHS probands, only a single deleterious allele was identified. Erythrocyte membranes from 6 of these patients (the seventh was TD) exhibited severe spectrin deficiency, indicating that another mutation associated with a production-defective SPTA1 allele was present in trans. In summary, 27 novel mutations were identified on 38 SPTA1 alleles. The data demonstrate how the synthesis of complementary lines of investigation, i.e. clinical, laboratory, biochemical and genetic data, can be leveraged to define and advance our understanding of inherited hematologic disease. They also demonstrate the significant genetic heterogeneity in hereditary spherocytosis and support the use of genomic strategies for diagnosis, particularly in severe cases, allowing design of appropriate therapy.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.