Abstract

Abstract 1002

Fetal hemoglobin (HbF) levels significantly modulate the severity of the 2 major β-hemoglobin disorders - sickle cell anemia (SCA) and β-thalassemia. Three major quantitative trait loci (QTLs; Xmn1-HBG2, the HBS1L-MYB [HMIP] intergenic region on chromosome 6q23, and BCL11A on chromosome 2p16) account for 20–50% of the common variation in HbF levels in SCA and β - thalassemia patients, and in healthy adults (Thein et al., Hum Mol Genet (2009) 18:R216). Lettre et al. (PNAS (2008) 105:11869) confirmed the influence of SNPs at the BCL11A and HBB loci in an African American cohort and a Brazilian cohort of SCA patients; as well as a significant influence of the HMIP region SNPs (rs7776054, rs9389268 and rs4895441) on HbF expression in the Brazilian SCA cohort. A strong association between HMIP polymorphisms that have a high frequency in the European population and modulation of F cell numbers has been reported (Creary et al., PLoS One (2009) 4:e4218). Given the unusually high admixture of the Brazilian population, the current study aimed to look at the influence of such HMIP markers on HbF production in SCA patients from this population (two regions, in the Northeast and Southeast of Brazil). We studied the influence and frequencies of the HMIP allele marker rs9376090 (that specifically tracks European chromosomes), as well as the rs9399137 marker (that has a much higher frequency in European descendents than in African descendents), as well as two HMIP markers (rs9389269 and rs9402686) that are also common in African descendents. Patients (220 HbSS, aged 12–68 years) were recruited at the Hematology Center, UNICAMP and at Fundação Hemope. The study was approved by the local Ethics Committees and informed consent was provided by all participants. Patients presenting the XmnI Gγ polymorphism (N = 2) were identified and excluded from further analysis, as this polymorphism has a known influence on HbF. The HMIP markers were genotyped by Taqman assays. Percentage HbF levels were determined by HPLC, using the Variant™ Bio-Rad kit, and were log transformed to normalize distribution for regression analyses. For those patients on hydroxyurea (HU) therapy, pre-HU HbF levels were used for analyses. Tests for associations between SNPs and HbF levels were conducted using linear regression models (SPSS v.15), including age and sex as covariates. High minor allele frequencies (MAF) for all four HMIP markers were observed in the population of patients studied (MAF; 0.09, 0.10, 0.12 and 0.12 for the rs9376090, rs9399137, rs9389269 and rs9402686 markers, respectively). For all four SNPs studied, higher levels of HbF were observed for the SCA individuals that were homozygotes for the minor allele, with strikingly higher levels of HbF presented by those individuals that were homozygotes for the rs9376090 and rs9399137 polymorphisms (see Table). The clinical courses of these patients were consistent with the higher levels of HbF observed (data not shown). Significantly higher HbF was also found in heterozygotes for the HMIP SNPs, compared to the major allele homozygotes. The variance in HbF levels due to rs9376090 was 7.1% (β= 0.270; p = 6.36 10−5), due to rs9399137 was 7.1% (β= 0.270; p = 9.59 10−5), due to rs9389269 was 8.3% (β= 0.287; p = 2.31 x10−5) and to rs9402686 was 8.3% (β=0.291; p = 2.18 x10−5). Our results confirm the HBS1L-MYB intergenic region as a key determinant of HbF levels in Brazilian SCA patients. The admixture of the Brazilian population has apparently led to a much higher incidence of European haplotypes at chromosome 6 in this population studied, when compared to the British and Tanzanian SCA populations. Importantly, the presence of these SNPs at the HMIP appears to have a very significant effect on HbF levels in the Brazilian SCA population, with probable clinical benefits.

Table.

HbF Levels in SCA Individuals, according to HMIP genotypes

SNP (HBS1L-MYB locus)% HbF (Median ± S.D.)
TTTCCC
rs9376090 6.35 ± 4.24 8.45 ± 3.63 17.25 ± 1.63 
 N = 182 N = 32 N=2 
  p<0.01 p<0.05 
rs9399137 6.47 ± 4.38 8.68 ± 3.63 17.25 ± 1.63 
 N = 167 N=37 N = 2 
  p<0.05 p< 0.05 
rs9389269 6.40 ±4.29 8.59 ± 3.70 12.68 ± 2.45 
 N = 162 N = 42 N = 5 
  p< 0.01 p< 0.05 
rs9402686 GG AG AA 
 6.26 ± 4.28 8.36 ± 4.67 12.68 ± 2.45 
 N =166 N = 41 N = 5 
  p<0.01 p<0.05 
SNP (HBS1L-MYB locus)% HbF (Median ± S.D.)
TTTCCC
rs9376090 6.35 ± 4.24 8.45 ± 3.63 17.25 ± 1.63 
 N = 182 N = 32 N=2 
  p<0.01 p<0.05 
rs9399137 6.47 ± 4.38 8.68 ± 3.63 17.25 ± 1.63 
 N = 167 N=37 N = 2 
  p<0.05 p< 0.05 
rs9389269 6.40 ±4.29 8.59 ± 3.70 12.68 ± 2.45 
 N = 162 N = 42 N = 5 
  p< 0.01 p< 0.05 
rs9402686 GG AG AA 
 6.26 ± 4.28 8.36 ± 4.67 12.68 ± 2.45 
 N =166 N = 41 N = 5 
  p<0.01 p<0.05 

Significant differences for heterozygotes and homozygotes for the minor allele, compared to the homozygote major allele group are indicated by P values. Kruskal-Wallis test, Dunn's multiple comparison post test.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.