Mammalian erythropoiesis is coordinated by a series of sequential events and precise regulation of a gene expression program. The Ldb1 complex, consisting of transcription factors Tal1 and Gata1, the Lim domain protein Lmo2, and bridging protein Ldb1, regulates expression of key genes during erythropoiesis. The transcriptional co-repressor Eto2 has been shown to associate with the Ldb1 complex to inhibit transcription of target genes. Mice lacking Eto2 exhibit decreased proliferation of hematopoietic stem cells and defective murine T-cell development, however, the role of Eto2 in erythropoiesis is unclear. We find an expansion of immature erythroid cells from the murine fetal liver at the expense of mature erythrocytes in Eto2 null animals. Similarly, human erythroid progenitor cells lacking Eto2 exhibit impaired erythroid differentiation. We also sought to characterize the role of Eto2 in human hemoglobin switching, the transition from expression of fetal gamma-globin to expression of the adult beta-globin, which is of clinical importance for developing therapies for sickle cell disease and beta-Thalassemia hemoglobinopathies. We established a model of human hemoglobin switching using differentiating umbilical cord blood cells. These cells initially express gamma-globin and switch to predominately express beta-globin. Using this model combined with shRNA depletion of Eto2, we demonstrated that Eto2 is a negative regulator of both gamma globin expression and hemoglobin switching. Using ChIP-seq, we also found that Eto2 target genes have increased H3 acetylation in the absence of Eto2, suggesting that Eto2 regulates target gene expression through recruitment of histone deacetylases. Together, our data demonstrate that Eto2 is required for normal erythropoiesis in murine and human cells. Recent genome-wide studies suggest that long non-coding RNAs (nc-RNAs) have potential as regulators of gene expression. BGL3 is a 1 kb nc-RNA transcript located 2 kb downstream of the A-gamma-globin gene. BGL3 harbors an Ldb1 transcription-factor binding site that is involved in locus control region (LCR) looping. The BGL3 gene interacts with the gamma-globin A gene promoter and the LCR region located upstream the beta globin genes. We hypothesized that the BGL3 locus could work as an enhancer. However, although ChIP studies showed that BGL3 is marked by histone modifications usually associated with this kind of regulatory sequence, BGL3 sequences did not demonstrate enhancer activity, at least in a transient luciferase reporter assay. The expression of BGL3 parallels that of the gamma-globin genes in various erythroid cell models, although at much lower levels. We mapped the 5' end of the BGL3 gene by both 5' RACE and RT-PCR which confirmed that BGL3 is not a read-through of the A-gamma-globin gene. To assess whether the expression of gamma-globin depends on BGL3 transcripts, we knocked down and over-expressed BGL3. We found that the gamma-globin expression increased 50% with respect to the control when BGL3 is transiently over expressed 25-fold (p > 0.05). Similarly, a 2-fold reduction in the expression of BGL3 is associated with a 40% reduction of the gamma-globin mRNA levels with respect to the control (p =0.02). These analyses suggest gamma-globin transcription requires the expression of BGL3. To extend these results, we are knocking out different regions of the BGL3 gene by CRISPR/Cas9 technology. Once we generate the mutants, we will perform RNA-seq experiments to investigate how BGL3 influences the expression of gamma-globin and potentially other genes in erythroid cells. The mechanistic connection between overall genome architecture, non-genic transcription and cell specific transcriptomes remains to be tested. We investigated the potential of a CTCF-binding insulator element or a transcription terminator to alter chromatin conformation and transcriptional activity of the beta-globin locus. Using homologous recombination and cassette exchange, we placed human beta-globin HS5 or the beta-globin 3 transcription terminator (Ter) between the beta-globin locus control region (LCR) enhancer and the proximal downstream gene, embryonic epsilon-y-globin, on one allele (D) of mouse V6.5 ES cells. WT ES cells and clones with insertions were differentiated along erythroid lines to embryoid bodies. In Ter and HS5 ES cells, epsilon-y-globin gene expression was dramatically reduced on day 9 of differentiation compared to WT ES cells. In contrast, adult beta-globin transcripts that appear later in differentiation (day 15) were not affected. To extend these results in vivo, we generated mice bearing the insertions by blastocyst injection of targeted ES cells. During primitive hematopoiesis, repression of epsilon-y on the D-allele was observed in animals with the insertions, mirroring the observations in ES cells. Both insertions were embryonic lethal in homozygous animals, indicating an inability to progress through primitive erythropoiesis. Ter insertion resulted in LCR-epsilon-y intergenic transcript accumulation upstream and within Ter sequences and detection of RNA pol II and elongation factors Spt16 and Spt5 in the transcribed region. We are performing chromosome conformation capture (3C) to determine the effect of insertions on LCR-globin gene long range interactions in WT and mutant mice. These results have implications for understanding enhancer mechanisms and the function of overall genome folding in gene regulation.
Showing the most recent 10 out of 16 publications
