GATA-1 and BCL11A Collaboratively Regulate Erythroid Gene Expression through HDAC1 Dependent Manner

Transcription factor GATA-1 plays important roles in hematopoiesis through interaction with other key transcription factors and epigenetic modulators. Previously, we established that GATA-1 can physically interact with HDAC1 at two arginine residues of the linker region between two zinc fingers. Arginine to alanine mutation at these sites blocks the HDAC1 interaction and results in constitutive GATA-1 acetylation, and reduced GATA-1 chromatin binding at the gene regulatory regions. In addition, the loss of HDAC1 interaction causes the blockage of erythroid and megakaryocytic differentiation in mice. Having established the importance of GATA-1 and HDAC1 interaction in GATA-1 function, next we aim to further investigate whether HDAC1-associated proteins are involved in regulating the function of GATA-1. We performed LC-MS/MS analysis using GATA-1 wild type and mutant expressing cells and identified that BCL11A was one of the major proteins that associates with GATA-1 through HDAC1.

BCL11A is a master regulator of γ-globin gene silencing through association with the NuRD complex and blocking of NFYA recruitment to γ-globin gene promoters. The role of BCL11A in regulating gene expression during erythroid differentiation is not well studied. HDAC1 mediated GATA-1 and BCL11A association was validated by pull down and immunoprecipitation assay. Next, the BCL11A and HDAC1 interacting site was determined. The N-terminal region of BCL11A is found to be important for HDAC1 interaction by pull down assay. Through analysis of the secondary structure of BCL11A N-terminal, L12/K14 are predicted to be important for protein-protein interaction. Indeed, L12A/K14A mutation significantly reduced HDAC1 interaction as well as HDAC1 mediated GATA-1 and NuRD complex association. To assess the effects of the L12A/K14A mutation of BCL11A on gene expression, we performed RNA-seq analysis using BCL11A knock down (KD) or L12A/K14A mutant overexpressed in HUDEP-2 cells. The result shows that the L12A/K14A mutation de-repressed γ-globin expression and positively affects genes for erythrocyte development that is similar to the expression profile of BCL11A KD. Therefore, L12/K14 residues of BCL11A might have an important role in γ-globin silencing and erythropoiesis by mediating HDAC1 associated GATA-1 and NuRD complex. The Cut&Run assay was performed to study the recruitment of BCL11A, GATA-1, HDAC1, NuRD complex, and NFYA in WT, BCL11A KD, and Flag-BCL11A L12A/K14A mutant cells. There are over 7000 GATA-1 and BCL11A co-binding peaks detected in wild type cells. Within these peaks, about 2000 peaks are at BCL11A binding motif site. GATA-1, HDAC1 and NuRD binding were abolished at these sites in BCL11A KD or L12A/K14A mutant cells, while BCL11A binding was not affected. There are about 3000 peaks at GATA-1 binding motif sites. While GATA-1 binding is not affected at these sites in BCL11A KD or L12A/K14A mutant cells, BCL11A, HDAC1 and NuRD complex binding were significantly reduced. This study demonstrates that there are two modes of BCL11A and GATA-1 co-recruitment: BCL11A can recruit GATA-1 and GATA-1 can recruit BCL11A to their specific binding motifs, respectively. Interestingly, these two modes regulate differential gene targets. BCL11A recruits GATA-1 at genes mainly involved in chromatin organization and cell cycle, while GATA-1 recruits BCL11A to genes mainly involved in transcription, erythrocyte and megakaryocyte differentiation. At the β-globin locus, BCL11A KD or L12A/K14A mutation reduced GATA-1 and NuRD binding and enhanced NFYA binding at the γ-globin promoter, suggesting that BCL11A cooperates with GATA-1, HDAC1, and NuRD complex and prevents NFYA binding at the distal binding region to suppress γ-globin expression. At LCR hypersensitive site 2 (HS2) region, we also observed a reduction of GATA-1 and NuRD recruitment with BCL11A KD or L12A/K14A mutation while GATA-1 recruitment at other promoters or LCR regions were not changed, suggesting that the HS2 region may be involved in reactivating γ-globin in BCL11A KD or L12A/K14A mutation cells. Future study will continue to analyze how BCL11A KD or mutation affects BCL11A and GATA-1 co-binding and target gene expression, and NFYA or other factors recruitment. In summary, our study shows that BCL11A interacts with GATA-1 in a HDAC1 dependent manner and this interaction is important in regulating fetal globin expression and erythroid development.