Pharmacological or genetic induction of fetal hemoglobin (HbF, α2γ2) in adult red blood cells is a proven strategy to ameliorate the clinical symptoms of sickle cell disease (SCD) and β-thalassemia. Therefore, efforts are underway to better understand mechanisms that mediate the perinatal switch from HbF to adult hemoglobin (HbA, α2β2). We performed a CRISPR-Cas9/guide (g) RNA screen to identify novel proteins that regulate HbF production in HUDEP-2 cells, a human erythroid line that normally expresses HbA. We identified UHRF1 (ubiquitin-like with PHD and RING finger domains 1) as a repressor of HbF production. UHRF1 binds hemi-methylated DNA and recruit DNA methyltransferase 1 (DNMT1) to ensure faithful maintenance of DNA methylation during DNA replication. Numerous UHRF1-interacting proteins, including DNMT1, EHMT1/2 and HDAC2 are associated with γ-globin repression. We used CRISPR/Cas9 and RNA interference to validate UHRF1 as a HbF regulator. Compared to non-targeting gRNA UHRF1 disruption using Cas9 + 2 separate gRNAs increased the γ-globin/γ+β-globin RNA ratio from 1.9 to 25.8/27.1% (P<0.01), increased the fraction of HbF immunostaining cells ("F-cells") from 7.5 to 25.1/35.4% and increased HbF protein from 2.10 to 16.3/15.0% (P<0.01) in HUDEP-2 cells. Compared to a control luciferase shRNA, 2 different UHRF1 shRNAs increased theγ-globin/γ+β-globin RNA ratio from 9.68% to 21.59/28.93% (P<0.01), increased the F-cell fraction from 37.9 to 49.8/55.6% and increased HbF protein from 9.1 to 16.18/18.5% (P<0.05) in erythroid cells derived from normal adult peripheral blood CD34+ cells. UHRF1 deficiency did not alter erythroid maturation or expression of key transcription factor genes that regulate HbF expression in HUDEP-2 or CD34+ cells (BCL11A, ZBTB7A, MYB and KLF1). UHRF1 mutant proteins defective in recognizing H3K9me2 (FW237/238AA), binding to hemi-methylated DNA (R491A) or ubiquitination of H3K23 to enhance DNMT1 recruitment (C741A), were unable to repress HBG1/HBG2. These mutations have the most profound effects on maintaining DNA methylation, indicating that UHRF1 represses HBG1/HBG2 in HUDEP-2 cells through this mechanism.

UHRF1 knockout induced genome-wide demethylation including 6 CpG sites located at positions -162, -53, -50, +6, +17, +50 positions relative to the γ-globin (HBG1 and HBG2) transcription start site. Demethylation of these sites is thought to be required for γ-globin de-repression. However, forced demethylation of these cytosines in HUDEP-2 cells using specific gRNAs + dead (d) Cas9-TET1 was not sufficient to activate γ-globin expression when UHRF1 was present. Additionally, dCas9-DNMT3a-mediated methylation of the HBG promoter CpG residues in UHRF1 knockdown HUDEP-2 cells did not inhibit γ-globin expression in UHRF1 knockout HUDEP-2 cells. Based on these studies, we conclude that: 1) UHRF1 regulates γ-globin transcription; 2) demethylation of CpG sites at the HBG gene promoters is neither necessary or sufficient for γ-globin induction; 3) UHRF1 regulates γ-to-β globin gene switching either by methylating DNA regions other than those present around the HBG promoter or through non-canonical activities. Distinguishing these mechanisms will elucidate further our understanding of globin gene switching and could identify new pathways for pharmacological induction of HbF.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution