The Ldb1 complex, including GATA-1 and TAL1, mediates long range interaction between the beta-globin locus control region (LCR) and gene in adult mouse erythroid cells. In an adult human erythroid cell model in which fetal gamma-globin genes can be robustly re-activated by cytokines, we showed that the Ldb1 complex mediates gamma-globin/LCR proximity through a site 3-prime to the genes. Reactivation of these genes may be therapeutically useful in beta-thalassemia and sickle cell disease and understanding how they are silenced is critical. Previous work in the lab showed that when Eto2, a co-repressor, associates with the Ldb1 complex, gamma-globin expression and LCR interactions are decreased. To identify factors functioning downstream of Eto2 that may regulate switching from gamma to beta-globin transcription, we performed RNA-Seq in control and Eto2 depleted K562 cells revealing over 1000 genes (Padj <0.05) mis-regulated in the absence of Eto2 that are now under study. To determine if Eto2 is required for hemoglobin switching, we established a model of this process in CD34+ umbilical cord blood cells and found that reduction of Eto2 using shRNA resulted in higher levels of gamma globin mRNA and a 2 day delay in the switch to predominant expression of beta globin compared to mock treated controls. In CD34+ bone marrow progenitor cells that normally express high levels of beta globin and low levels of gamma globin, reduction of Eto2 lead to predominant gamma globin expression. We are identifying direct targets of Eto2 using ChIP-Seq to clarify the pathways and genetic interactions mediated by Eto2. In addition, we are generating Eto2 null mice carrying a human beta-globin transgene to determine how Eto2 regulates expression of human beta globin genes in vivo. In other experiments, we are addressing the hypothesis that Eto2 regulates erythropoiesis. To determine the role of Eto2 in erythrocyte development, we analyzed expression of candidate genes in yolk sac from WT and Eto2-/- mouse embryos at 8.5 and 10.5 dpc and fetal liver at 12.5 and 15.5 dpc, the primary sites of erythropoiesis during these stages in development. The data suggest an overall defect in erythropoiesis. We have begun flow cytometric sorting experiments to address the stage in which erythroid development becomes impaired. Bgl3, a non-coding RNA transcript located downstream of the A-gamma globin gene, harbors an Ldb1 site that is involved in LCR looping. Bgl3 transcription parallels that of the gamma-globin in various erythroid cell models, although at a much lower level. To investigate the role of Bgl3 during hemoglobin switching we are following Bgl3 and gamma-globin transcripts by RT-PCR and RNA-FISH. To characterize the Bgl3 gene and to rule out the possibility of Bgl3 being a read-through transcript of the A-gamma globin gene located 3 kb upstream of the known Bgl3 gene sequence, we mapped the 5-prime end of Bgl3 by RACE and RT-PCR. We determined that the 5-prime end of Bgl3 is located 600 bp upstream of the sequence deposited in GenBank and also that Bgl3 and the A-gamma globin gene produce independent transcripts. In addition, to investigate whether the expression of gamma-globin depends on Bgl3 transcripts or its transcription per se, we knocked down and over-expressed Bgl3. We found that expression of gamma globin correlates with the levels of Bgl3. Also, we are examining the effect of manipulating Bgl3 levels on gamma globin long range LCR interactions. As Ldb1 and other erythroid specific transcription factors and the gamma globin gene repressor Bcl11a bind within the Bgl3 gene, we are studying if the Bgl3 locus could act as an enhancer. In Chip experiments we observed that this locus is occupied by the enhancer-specific transcription factor p300 and also it is marked by histone modifications associated with enhancers such as histone H3 K27 acetylation and histone H3 K4 mono methylation. Further, the ratio of H3 K4me1/H3 K4me3 is high which indicates the presence of an enhancer. To further investigate the potential enhancer function of Bgl3 we are performing luciferase reporter assays to quantitatively measure the expression of a reporter gene driven by a gamma globin gene promoter. We are also targeting deletion of Bgl3 sequences by genome editing and investigating its RNA-protein interactome. We are using homologous recombination in mouse ES cells study how individual globin genes establish stage specific enhancer communication. We targeted a region upstream of the mouse embryonic epsilon y gene on one allele in ES cells by homologous recombination. Recombinase mediated cassette exchange (RMCE) was used to insert chromatin insulator human HS5 or a transcription terminator in this position. In differentiating ES cells, insertion of the transcription terminator or hHS5 between the LCR and downstream genes inhibits embryonic ey gene activation in an allele-specific fashion but neither insertion affects expression of the downstream adult beta-globin gene. We have undertaken blastocyst injection of our successfully targeted ES cells. Recently we obtained mouse lines which carry the chromatin insulator or transcription terminator upstream of ey. Analyzing βglobin gene expression from embryonic peripheral blood or fetal liver at different developmental stages showed insertion of the chromatin insulator or transcription terminator disturbed embryonic globin gene activation only. For further investigation, we monitored intergenic transcription between the LCR and ey in WT mice and mice with the insulator or transcription terminator insertion. Presently, we are investigating RNA pol II localization across this region and at globin gene promoters using in vivo samples from WT and mutant mouse lines. Concurrently, we have introduced a mutant transcription terminator at the target position by homologous recombination to investigate as a control. Similar experiments are ongoing to introduce the insulator/transcription terminator at an additional target site that we have created between the embryonic and adult globin genes. Results from these mouse lines we are producing will be used to understand how long range interactions regulate beta-globin gene activation in a stage specific manner in vivo.
Showing the most recent 10 out of 16 publications
