Control of Hemoglobin Switching by BCL11A.

Abstract

Abstract 5All vertebrates switch expression of globin chains during development. In humans b-like globins switch from embryonic to fetal to adult, whereas in the mouse a single switch from embryonic to adult occurs. The switch from human fetal (g) to adult (b) expression is especially critical in the b-hemoglobin disorders, such as sickle cell anemia and the b-thalassemias. Delay of the switch or reactivation of the fetal gene in the adult stage greatly ameliorates clinical severity. Despite intensive molecular studies of the human b-globin cluster over more than two decades, the proteins regulating the switch, and the mechanisms controlling the process, have been largely elusive. Recently, genome-wide association studies identified genetic variation at a chromosome 2 locus that correlates with the level of HbF in different populations. The most highly associated single nucleotide polymorphisms (SNPs) reside in an intron of the BCL11A gene, which encodes a zinc-finger repressor protein. Previously we showed that shRNA-mediated ex vivo knockdown of BCL11A in cultured human CD34-derived erythroid precursors leads to robust HbF expression, consistent with a role for BCL11A in maintaining g-genes in a silenced state in adult cells. To address in vivo roles of BCL11A either in development or in globin gene silencing in an intact individual, we have employed stringent genetic tests of function in mice that carry a complete human b-globin gene cluster as a yeast artificial chromosome transgene (b-locus mice). Knockout of BCL11A in mice leads to failure to silence the endogenous b-like embryonic genes in adult erythroid cells of the fetal liver (>2500-fold derepression). The ratio of human g to b globin RNA in the fetal liver of BCL11A knockout mice is inverted compared to controls, such that g constitutes >90% of the b-like human expression at embryonic day (E)14.5 and >75% at E18.5. These quantitatively striking findings indicate that BCL11A controls developmental silencing of g-globin gene expression. To address by formal genetics the contribution of BCL11A to g silencing in adult animals we have employed conditional inactivation of BCL11A through hematopoietic- and erythroid-specific Cre-alleles. These experiments reveal that BCL11A is also required in vivo for g-gene silencing in adults. We observed that human g-globin expression is persistently derepressed >2000-fold (as compared to littermate controls) in bone marrow erythroblasts of 15-20 week old b-locus mice upon erythroid-specific deletion of BCL11A. Taken together, these findings establish BCL11A as the first genetically validated transcriptional regulator of both developmental control of globin switching and silencing of g-globin expression in adults. The recognition of these roles for BCL11A now permits focused mechanistic studies of the switch. In human erythroid cells, BCL11A physically interacts with at least two corepressor complexes, Mi-2/NuRD and LSD1/CoREST, as well as the erythroid transcription factor GATA-1 and the HMG-box protein SOX6. Rather than binding to the promoters of the g- or b-globin genes as do these latter factors, BCL11A protein occupies the upstream locus control and g-d-intergenic regions of the b-globin cluster (as determined by high resolution ChIP-Chip analysis), suggesting that BCL11A mediates long-range interactions and/or reconfigures the locus during different stages. An in-depth mechanistic understanding of globin switching offers the prospect for design of target-based activation of HbF in adult erythroid cells of patients with hemoglobin disorders. Disclosures:No relevant conflicts of interest to declare.