Identification Of BCL11A Structure-Function Domains For Fetal Hemoglobin Silencing

Reactivation of fetal hemoglobin (HbF, α_2γ2) expression in adults ameliorates the clinical symptoms in patients with the major β-hemoglobin disorders, sickle cell disease (SCD) and β-thalassemias. The zinc-finger protein BCL11A is a major modulator of hemoglobin switching and HbF silencing. BCL11A was initially identified by genome-wide association studies (GWAS) as a new HbF-associated gene. Down-regulation of BCL11A in primary human erythroid cells induces HbF expression. Knockout of BCL11A in mice impairs HbF silencing in adult erythroid cells. Most importantly, inactivation of BCL11A alone in humanized SCD mice corrects the hematologic and pathologic defects through high-level HbF induction. These studies established BCL11A as a genetically and functionally validated transcriptional regulator of HbF switching and silencing.

In human and mouse erythroid cells, BCL11A is expressed as several isoforms, yet their individual roles in globin gene expression remain unexplored. Furthermore, the functional domains within the BCL11A protein responsible for its activity in HbF repression are largely unknown. To further understand the mechanistic roles of BCL11A in globin expression, we established a functional assay based on a BCL11A-null erythroid cell line generated by transcription activator-like effector nucleases (TALENs)-mediated deletion of an obligate erythroid-specific enhancer of BCL11A in murine erythroleukemia (MEL) cells. In the BCL11A-null cells, the expression of β-like embryonic globin genes is markedly induced (>200-fold), consistent with the role of BCL11A in repression of murine embryonic globin genes. To examine the activity of known BCL11A isoforms in HbF silencing, we expressed various BCL11A isoforms in these engineered BCL11A-null cells. Ectopic expression of full-length BCL11A-XL isoform, but not the alternatively spliced, C-terminally truncated L isoform, restored the full repression of β-like embryonic globins in BCL11A-null cells. Since XL and L differ only by 91 amino acids containing three tandem C2H2-type zinc finger motifs, these results indicate that the C-terminal zinc finger motifs are indispensable for BCL11A-mediated transcriptional repression.

To systemically define BCL11A functional domains for globin gene repression, we next generated a panel of BCL11A mutant cDNAs, including deletion of the N-terminal NuRD-interacting motif and one or more C2H2-type zinc finger domains. Analysis of various BCL11A mutants in the functional rescue assay identified several functional domains, including the N-terminal NuRD-interacting motif and five out of the six C2H2 zinc fingers, that are required for BCL11A-mediated repression. These findings provide the foundation for further molecular analysis of BCL11A functional domains in globin gene repression. BCL11A is known to interact with several transcriptional co-repressor complexes including Mi-2β/NuRD/HDAC1/HDAC2, LSD1/CoREST and SWI/SNF complexes, occupy discrete regions within the human β-globin cluster, and promote long-range chromosomal interactions. Our results suggest that BCL11A functional domains may be involved in protein-protein interactions, protein homo-/heterodimerization, and/or chromatin/DNA association that are required for its activity in HbF silencing.

In summary, we demonstrate that several functional domains on BCL11A protein are indispensable for its transcriptional activity in HbF silencing. Further focused studies of BCL11A structure-function domains in HbF silencing not only will advance our understanding of the molecular mechanisms by which BCL11A controls the clinically important fetal-to-adult globin switch, but may identify novel cellular targets for therapeutic HbF induction in β-hemoglobinopathies.