KLF1 binds to and activates the −198T>C γ-globin promoter in HUDEP-2 cells. (A) mRNA expression levels of γ-globin displayed as a percentage of γ- and β-globin [γ/γ+β]. mRNA levels were determined by qRT-PCR and cycle threshold (Ct) values were normalized to rRNA levels of 18S. Primer efficiencies were confirmed to be equivalent. Shown are clonal HUDEP-2 WT cells (n = 4) and 2 clones (n = 2 for #15 and n = 3 for #17) each carrying 1 allele with the −198T>C mutation. γ-globin mRNA levels are significantly higher in cells carrying the −198T>C mutation (**P < .01). (B) Flow cytometry analysis of HbF levels in clonal WT HUDEP-2 cells (n = 3) and a clonal HUDEP-2 cell population that is heterozygous for the −198T>C mutation (n = 3). Representatively shown is the median of 3 experiments (± standard deviation [SD]). Cells were permeabilized and then stained with APC-conjugated HbF antibody. The amount of cells expressing high HbF (“F cells”) was determined by flow cytometry. The same gating strategy was applied for WT HUDEP-2 cells and −198T>C HUDEP-2 cells using FlowJo software. (C) mRNA expression levels of γ-globin displayed as a percentage of γ- and β-globin [γ/γ+β]. mRNA levels were determined by qRT-PCR and Ct values were normalized to rRNA levels of 18S. Shown are parental HUDEP-2(Δ^Gγ)WT cells and HUDEP-2(Δ^Gγ) cells homozygous for the −198T>C mutation. γ-globin mRNA levels are significantly higher in cells carrying the −198T>C mutation (***P < .0001) (D) Flow cytometry analysis of HbF levels in HUDEP-2(Δ^Gγ)WT and HUDEP-2(Δ^Gγ)−198T>C cells. Representatively shown is the median of 3 experiments (±SD). Cells were permeabilized and then stained with APC-conjugated HbF antibody. The amount of cells expressing high HbF (“F cells”) was determined by flow cytometry. (E) HPLC traces depicting hemoglobin production in HUDEP-2(Δ^Gγ)WT and 3 clonal (Δ^Gγ)−198T>C cell populations. Highlighted in red is the peak for HbF. Percentages are HbF over total hemoglobin (HbF and HbA0). (F) ChIP-qPCR analysis of the relative enrichment of KLF1 at various genomic loci in HUDEP-2(Δ^Gγ)WT (n = 2) and −198T>C cells (n = 4). The tested genomic loci were the γ-globin promoter, the β-globin promoter, and the VEGFA promoter (−ctrl). The SP1 promoter served as a positive control (+ctrl) for successful pulldown with the respective antibody and all values were normalized to enrichment at the SP1 promoter. Enrichment of the γ-globin promoter after KLF1 pulldown is significantly higher in HUDEP-2(Δ^Gγ)−198T>C cells whereas enrichment of the β-globin promoter is significantly lower (**P < .01). (G) ChIP-qPCR analysis of the relative enrichment of KLF1 at genomic loci in 5-day differentiated HUDEP-2(Δ^Gγ)WT (n = 3) and −198T>C cells (n = 3). Again, KLF1 enrichment is significantly higher (*P < .05) at the γ-globin promoter but lower (**P < .01) at the β-globin promoter. (H) 3C assay measuring locus-wide crosslinking frequencies in HUDEP-2(Δ^Gγ)WT (n = 3) and 3 clonal −198T>C cell populations. A schematic of the human β-globin locus is shown on top of the graph. The x-axis indicates distances in kilobases (kb) from the ε-globin gene. Vertical lines represent HindIII restriction sites. The dark gray bar denotes the anchor HindIII fragment containing hypersensitive site 3. Light gray bars denote analyzed HindIII fragments. Shown is mean ± standard error of the mean (s.e.m.). (I) Proposed model of molecular mechanism in British HPFH. In WT adult erythroid cells, KLF1 drives the expression of adult β-globin via the promoter (top panel). The presence of the −198T>C mutation (bottom panel) allows binding of KLF1 to the fetal γ-globin promoter and leads to upregulation of γ-globin expression through recruitment of the LCR. In these cells, binding of KLF1 to the β-globin promoter and β-globin transcript levels are reduced. n.s., not significant; SSC, side scatter.