Fetal hemoglobin in sickle cell anemia: Bayesian modeling of genetic associations

We genotyped single nucleotide polymorphisms (SNPs) in: (1) the beta-globin gene-like cluster, (2) quantitative trait loci (QTL) previously associated with fetal hemoglobin (HbF) concentration on chromosomes 6q, 8q, and Xp, and (3) candidate genes that could effect HbF levels, in sickle cell anemia subjects. HbF concentration was modeled as a continuous variable with values in a finite interval using a novel Bayesian approach. We first tested the associations of SNPs with HbF in a group of 1,518 adults and children (CSSCD study), and validated the results in a second independent group of 211 adults (MSH study). In subjects aged >= 24 years, 5 SNPs in TOX (8q12.1), 2 SNPs in the beta-globin gene-like cluster, 2 SNPs in the Xp QTL, and 1 SNP in chromosome 15q22 were associated with HbF in the CSSCD and also validated in the MSH. Four other SNPs in 15q22 were associated with HbF only in the larger CSSCD data. When patients aged < 24 years in the CSSCD were examined, additional genes, including 4 with roles in nitric oxide metabolism, were associated with HbF level. These studies confirm prior analyses using traditional analytical approaches showing associations of SNPs in TOX, GPM6B, and the beta-globin gene-like cluster with HbF levels. We also identified an additional candidate regulatory region in chromosome 15q22 that is associated with HbF level. By stratifying patients by age, our results also suggest that different genes might modulate the rate of decline of HbF and the final level of HbF levels in sickle cell anemia.