Mechanistic Insights into Forced Chromatin Looping Mediated Activation of Fetal Globin Gene Expression

The β-globin enhancer, called locus control region (LCR), is required for high level expression of all b-type globin genes. The LCR is in physical proximity with the genes it controls, with contacts shifting from embryonic (ε) to fetal (γ) and finally to adult (δ and β) globin gene promoters during development. In prior studies we showed that forced chromatin contacts between the LCR and the β-globin promoter led to transcriptional activation, suggesting that LCR-promoter looping causally underlies β-globin transcription (Deng et al. Cell 2012). In these studies, the transcription co-factor Ldb1 was tethered to the β-globin promoter using artificial zinc finger (ZF) DNA binding proteins, to trigger the promoter-LCR interaction. We subsequently showed that tethering Ldb1 to the promoters of developmentally silenced embryonic or fetal globin genes reactivated their expression in adult erythroblasts in a manner dependent on looped contacts with the LCR (Deng et al. Cell 2014). This work established a novel strategy to raise fetal globin expression in patients with sickle cell anemia.

To examine mechanistically the effects of chromatin looping on gene expression we performed single molecule RNA FISH experiments to precisely measure transcription output at individual alleles before and after enforced LCR-γ-globin looping. The experiments were carried out in primary human erythroblasts, which produce elevated levels of γ-globin when exposed to culture conditions. Preliminary data suggest that the majority of transcripts emerge from the β-globin gene with a smaller fraction of transcripts coming from the γ-globin gene as expected. Among cells producing any type of globin primary transcripts, a significant fraction (25%-35%) of cells co-express γ- and β-globin. Importantly, γ- and β-globin are frequently transcribed from the same allele. Forced juxtaposition of the LCR and the γ-globin promoter increases the number of alleles expressing only γ-globin while reducing the number of alleles expressing only the β-globin gene. This result is consistent with the γ- and β-globin genes competing for LCR activity, and emphasizes the usefulness of this approach in the context of sickle cell anemia in which not only elevated levels of fetal hemoglobin but also reduction of the mutant hemoglobin are desirable. Surprisingly, however, the proportion of alleles co-expressing γ-globin and β-globin remains largely constant. We are testing whether co-expression from the same allele is LCR independent. Finally, our studies suggest that LCR-promoter contacts increase the probability of transcription of a given allele.

We will also present work addressing the critical question as to how alteration of chromatin architecture overcomes the action of transcriptional repressive complexes, such as Bcl11a, which normally maintain embryonic and fetal globin genes in a repressed state throughout adulthood.

In sum, our studies produce a deeper understanding of the interplay of chromatin architecture and gene expression in a system that holds great potential for therapeutic application in patients with hemoglobinopathies.