A β-Globin Locus-Intrinsic Epigenetic Mechanism Underlies Fetal Globin Production in F-Cells

Upregulation of fetal hemoglobin (HbF, α₂γ₂) by reversing the developmental switch to adult HbA (α₂β₂) is a key approach for both pharmacologic and gene targeting therapies in the treatment of sickle cell disease (SCD) and β-thalassemia. HbF expression in healthy individuals, patients with SCD, and those treated with hydroxyurea is restricted to a subset of red blood cells known as F-cells; effective SCD therapy requires increasing the proportion of F-cells expressing sufficient HbF to block sickling. Although these cells have been observed since the 1950s, there have not been previous direct comparisons of F-cells to matched HbF-low A-cells from the same individual. Fetal erythroblasts have distinct global transcriptional programs and distinct long-range chromatin looping at the β-globin locus when compared to adult erythroblasts. An important question is therefore whether F-cells are formed through reversion to a fetal-like state at the transcriptional and epigenetic level. To address this clinically important question, we previously reported development of new techniques for the purification of stage-matched F- and A-erythroblasts from primary human CD34⁺ cell erythroid cultures and their downstream analysis (Khandros et al, Blood 2020). We demonstrated that F-cells in primary erythroid cultures have minimal transcriptional differences with A-cells and that the few differentially expressed transcripts do not overlap with fetal-specific transcripts. Furthermore, treatment with hydroxyurea or pomalidomide did not enhance transcriptional differences between F- and A-cells. Surprisingly, we did not find differences in the expression of any known HbF regulators such as BCL11A, LRF, or NuRD complex members that would account for differential HbF expression. Based on these findings, we hypothesized that F-cells are distinguished by epigenetic variation specifically at the β-globin locus.

Given that fetal erythroblasts differ from adult erythroblasts in the chromatin architecture of the β-globin locus (e.g. Huang et al, Genes and Development 2017), we compared the higher order chromatin organization of the β-globin locus between F- and A-cells by Capture-C, a next-generation sequencing-adapted form of chromatin conformation capture. We found that in F-cells, contacts between the distal enhancer and the promoters of the fetal globin genes HBG1 and HBG2 were increased, while those between the enhancer and adult globin genes (HBB and HBD) were reduced. Other architectural changes associated with fetal globin gene expression, including fetal specific contacts of an intergenic non-coding gene with chromatin domain boundaries at the β-globin locus were also partially enriched in F-cells. We also did not find any differences in promoter-enhancer contacts between F- and A-cells for other developmentally regulated genes BCL11A, LIN28B, and THRB. Together these results are consistent with the concept that epigenetic changes associated with nuclear architecture that occur specifically at the β-globin locus underlie the difference in globin gene expression profiles between F- and A-cells.

In sum our data demonstrate that in adult erythropoiesis, F-cells do not arise through either a wholesale reversion to a fetal-like genetic program or through variation in any known HbF regulators. Instead, modulation of chromatin architecture intrinsic to the β-globin locus, perhaps in a stochastic manner, accounts for elevated fetal globin expression in F-cells. We are currently performing mechanistic studies to elucidate the basis for the epigenetic regulation of the β-globin locus in F-cells. These studies will further our understanding of fetal hemoglobin regulation in adult cells and might inform new therapeutic approaches for SCD and β-thalassemia.