1. A 3-bp deletion in the HBS1L-MYB intergenic region on chromosome 6q23 is associated with HbF expression
by John J. Farrell, Richard M. Sherva, Zhi-yi Chen, Hong-yuan Luo, Benjamin F. Chu, Shau Yin Ha, Chi Kong Li, Anselm C. W. Lee, Rever C. H. Li, Chi Keung Li, Hui Leung Yuen, Jason C. C. So, Edmond S. K. Ma, Li Chong Chan, Vivian Chan, Paola Sebastiani, Lindsay A. Farrer, Clinton T. Baldwin, Martin H. Steinberg, and David H. K. Chui
Blood
Volume 117(18):
May 5, 2011
©2011 by American Society of Hematology

2. Manhattan plot showing the GWAS results of 619 unrelated adult β-thalassemia heterozygotes in Hong Kong, using the additive model.
Manhattan plot showing the GWAS results of 619 unrelated adult β-thalassemia heterozygotes in Hong Kong, using the additive model. The association was with normalized HbF expressed as percentage determined by HPLC.
John J. Farrell et al. Blood 2011;117:
©2011 by American Society of Hematology

3. Effect of BCL11A and HMIP QTL upon HbF and F cells.
Effect of BCL11A and HMIP QTL upon HbF and F cells. (A) Box plots showing the complete distribution of HbF in 619 Chinese β-thalassemia heterozygotes, expressed as percentage in log scale on the y-axis. Each rectangle represents the data between the 25th and 75th quartiles, and the bar in each rectangle is the median value for the HbF in log scale. rs766432: AA, AC, and CC represent the SNP genotypes for rs in BCL11A on chromosome 2p16. rs : TT, CT, and CC represent the SNP genotypes for rs in HMIP on chromosome 6q23. (B) Effect of the 3 major HbF QTLs on HbF and F-cell levels in adult β-thalassemia heterozygotes who had HbF and F cells determined. Data are mean ± SD. Control indicates 80 subjects (age, 48.3 ± 9.0 years) who are homozygous for the major alleles for rs in HBG2 promoter, rs in BCL11A, and rs in HMIP. rs766432: 8 subjects (age 46.9 ± 9.6 years) with identical HbF QTL genotypes as those in control, except that they are homozygous for the minor allele for rs in BCL11A. rs : 6 subjects (age 40.3 ± 13.8 years) with identical HbF QTL genotypes as those in control, except that they are homozygous for the minor allele for rs in HMIP.
John J. Farrell et al. Blood 2011;117:
©2011 by American Society of Hematology

4. The 3-bp deletion in the HMP on chromosome 6q23.
The 3-bp deletion in the HMP on chromosome 6q23. (A) Upper figure: Chromosome 6q23 HMIP is adapted from Thein et al.2 The 3 SNPs shown represent those with the strongest association signal in different studies: rs ,5,10,19 rs ,6,10 and rs Below are shown nucleotide sequences between chromosome 6: to bp, and transcription factor binding motifs: a, sequences with the rs major allele C in red; b, sequences with the rs minor allele T in red; and c, sequences with the 3-bp deletion. Yellow represents sequence that is the TAL1/SCL and E47 binding motif; gray, the GATA binding motif; and blue, the RUNX1 binding motif. The locations of the 4 SNPs on chromosome 6 are shown on the lower left hand corner. (B) Upper figure: Nucleotide sequences surrounding rs in a Chinese β-thalassemia heterozygote who is homozygous for the rs major allele, T/T. This person is also homozygous for having intact TAC without deletion. Lower figure: Nucleotide sequences surrounding rs in a Chinese β-thalassemia heterozygote who is homozygous for the rs minor allele, C/C. This person is also homozygous for the TAC 3-bp deletion.
John J. Farrell et al. Blood 2011;117:
©2011 by American Society of Hematology

5. Association of genotyped and imputed SNPs throughout the HMIP region on chromosome 6q23 with HbF levels among 619 adult Chinese β-thalassemia heterozygotes.
Association of genotyped and imputed SNPs throughout the HMIP region on chromosome 6q23 with HbF levels among 619 adult Chinese β-thalassemia heterozygotes.
John J. Farrell et al. Blood 2011;117:
©2011 by American Society of Hematology

6. ChIP assay.
ChIP assay. Relative real-time PCR signal intensities of the ChIP assay, showing significant binding of TAL1, E47, GATA2, and RUNX1 near the 3-bp deletion polymorphism. represents transcription factor binding signals in the immediate vicinity of the 3-bp deletion polymorphism; binding signals at approximately 1.5 kb centromeric to the 3-bp deletion polymorphism; and binding signals at approximately 2.5 kb telomeric to the 3-bp deletion polymorphism. IgG indicates negative control using IgG for immunoprecipitation; H3, positive control using antibody against histone H3; and RT-PCR, using exon 3 primer set.
John J. Farrell et al. Blood 2011;117:
©2011 by American Society of Hematology

7. ENCODE data display beginning from the top row down of ChIP-seq signals in K562 cells on occupancy of GATA-1, GATA-2, BRG1, Ini1, RNA polymerase 2, and histones H3K4Me1 and H3K4Me3.
ENCODE data display beginning from the top row down of ChIP-seq signals in K562 cells on occupancy of GATA-1, GATA-2, BRG1, Ini1, RNA polymerase 2, and histones H3K4Me1 and H3K4Me3. Also shown is genomic footprinting, RNA transcript22,23 both also in K562 cells and ESFERR regulatory potential,24 among different species. The vertical red lines represent the site of the 3-bp deletion polymorphism (chromosome 6: ), which is in complete LD with rs The green lines represent the site of rs (chromosome 6: ), the SNP that was found in multiple GWASs to have the most significant P value in HbF association. These ENCODE data were generated by the ENCODE consortium and are available on the Genome Browser at UCSC.21 The GATA-1, GATA-2, BRG1, Ini1, and RNA Polymerase II ChIP-Seq data were generated by the laboratories of M. Snyder, M. Gerstein, and S. Weissman at Yale; P.J. Farnham at the University of California-Davis; and K. Struhl at Harvard. The H3K4Me1 and H3K4Me3 data were generated by the Bernstein Laboratory at the Broad Institute. The Digital Genomic Footprinting data were generated by the University of Washington ENCODE group. The GIS RNA-seq data were generated by the Genome Institute of Singapore.
John J. Farrell et al. Blood 2011;117:
©2011 by American Society of Hematology

8. Enhancer-like activity of the DNA fragment encompassing the 3-bp deletion.
Enhancer-like activity of the DNA fragment encompassing the 3-bp deletion. (A) Enhancer-like activity of the 61-bp DNA fragment surrounding the 3-bp deletion polymorphism on HBG2 proximal promoter linked to a luciferase expression vector. The luciferase activity of the plasmid without the 61-bp DNA fragment was normalized to be 1.0. The 61-bp DNA fragment enhances the HBG2 promoter activity to 3.4-fold of the control. The 58-bp DNA fragment with the 3-bp deletion enhances the HBG2 promoter activity to 5.4-fold of the control. (B) Effect of mutation of transcription factor binding sites in the 61-bp DNA fragment on its enhancer-like activity. The luciferase activity of the plasmid with the 61-bp DNA fragment was normalized to be 1.0. Mutation of either RUNX1 or 3′ GATA binding site down-regulates the enhancer-like activity to 50%. Mutation of the E-box or 5′ GATA binding site does not decrease the enhancer-like activity of the DNA fragment. (C) Effect of mutation of transcription factor binding sites in the 58-bp DNA fragment with the 3-bp deletion on its enhancer-like activity. The luciferase activity of the plasmid with the 58-bp DNA fragment was normalized to be 1.0. Mutation of the E-box, RUNX1, or 3′ GATA binding site down-regulates the enhancer-like activity to approximately 50%. Mutation of the 5′ GATA binding site does not decrease the enhancer-like activity of the DNA fragment.
John J. Farrell et al. Blood 2011;117:
©2011 by American Society of Hematology