Pomalidomide downregulates FIZ1 to induce fetal hemoglobin in adult erythroid progenitors

Abstract Reactivation of fetal hemoglobin (HbF) expression has been a target for the treatment of hemoglobinopathies such as sickle cell disease (SCD) and β-thalassemia for several decades. While gene therapy has recently been approved as a cure for these hemoglobinopathies, its high cost limits global implementation. Hydroxyurea thus remains the treatment of choice for the induction of HbF although there is variability in the degree of response in compliant patients. Significant efforts have thus been aimed at identifying additional potent inducers of HbF. In vitro studies have documented immunomodulatory drugs (IMiDs), such as pomalidomide, as potent pharmacological inducers of HbF. The primary mechanism of action of IMiDs is the induction of proteasomal degradation of intracellular proteins mediated through Cereblon-dependent ubiquitination. Specific genes in multiple myeloma and del5q-syndrome have been identified as IMiDs' targets; however, the upstream mediators of pomalidomide leading to HbF induction remain to be defined. Previous studies identified IKZF1, IKZF3, CSK1α and ZFP91 as direct IMiD-mediated targets of Cereblon. In a previous study, we ruled out a potential role for IKZF1 and IKZF3 in HbF induction. To identify novel targets of pomalidomide involved in HbF regulation, we performed a global proteomic analysis on healthy human bone marrow CD34+ cells cultured with pomalidomide during ex vivo erythropoiesis. Among the 12,000 quantified proteins, we confirmed the downregulation of known pomalidomide targets such as IKZF1 and ZFP91. Interestingly, we identified FIZ1 (FLT3-Interacting Zinc Finger 1) as a novel target whose protein level was significantly reduced by pomalidomide and this effect was dose-dependent (starting at 100nM of pomalidomide). Having identified FIZ1 as a novel pomalidomide target, we sought to investigate its function in human erythropoiesis. CRISPR/Cas9 was used to knockout FIZ1 in CD34+ cells at day 2 of differentiation, and editing efficiency was assessed by measuring the indel frequency. At day 4 and day 11 of in vitro differentiation, there were no significant differences in dynamics erythropoiesis between control and sgFIZ1 knockout conditions, as measured by flow cytometry using CD105/CD71 (Day 4, progenitor stages) or Band3/a4-integrin (Day 11, precursor stages) as markers of erythroid differentiation. Western blot analyses demonstrated that the expression levels of key erythroid transcription factors (e.g. GATA1, KLF1) were also unaltered in FIZ1 knockout cells. Having ruled out that FIZ1 knockout had any detrimental effects on adult human erythropoiesis, we assessed its impact on HbF production. At day 14 of culture, we observed a significant increase in the total expression levels of γ-globin, while β-globin was slightly decreased and α-globin was unchanged. Similarly, quantification of F-cells by flow cytometry showed a significant increase in the F-cell population in FIZ1 knockout cells (51% vs 6% in the control). Mechanistically, FIZ1 knockout led to decreased expression of HbF repressors BCL11A, HRI and SOX6 while expression of other HbF regulators such as LRF remained unchanged at day 7 of differentiation. Similar expression patterns were noted in cells treated with pomalidomide and an additive effect was observed when cells were treated with both FIZ1 guides and pomalidomide. In summary, our data indicate that pomalidomide enhances the degradation of FIZ1 leading to increase in γ-globin expression without altering erythroid differentiation and as such is a novel inducer of HbF is erythroid cells.