Targeted Degradation of the Wiz Transcription Factor for Gamma Globin De-Repression

Sickle cell disease (SCD) is a life-threatening genetic condition arising from a point mutation in the β-globin gene that causes hemoglobin polymerization and vaso-occlusive sickling of erythrocytes. Hereditary persistence or pharmacologic induction of gamma-globin (HBG1 or HBG2) expression leads to increases in fetal hemoglobin (HbF) that biophysically oppose hemoglobin polymerization. Indeed, the standard-of-care, hydroxyurea (HU), increases expression of HbF in the erythroblast, and in patients the magnitude of response (9-24% HbF) correlates with a reduction in pain crises and acute complications. However, the use of HU is limited by variable response, myelosuppression, GI side effects and carcinogenicity risk. New targeted therapeutic approaches to induce HbF are therefore being intently pursued, leveraging the increasingly detailed functional characterization of the globin locus. The recent discovery of BCL11A as a repressor of HBG has prompted the widespread innovation of genetically engineered hematopoietic stem cells that suppress BCL11A expression or function. While a major advance in SCD therapeutics, the complexity and clinical implementation of stem cell transplantation may limit access and impact, especially in developing countries where SCD burden is greatest. We therefore undertook to discover a small-molecule HbF inducer, focusing on transcription factors that repress gamma-globin expression. Although transcription factors as a target class have historically been considered "undruggable," recent mechanistic characterization of phthalimide medicines as serendipitous, cereblon (CRBN)-dependent degraders of IKZF1 and IKZF3 by our organization and others, suggested the plausible degradation of new zinc-finger containing transcription factors by this emerging class of molecular glues. Therefore, we elaborated a large library of CRBN-biased ligands for study in target-directed and phenotypic drug discovery campaigns. Here we report for the first time the phenotypic discovery and structure-based optimization of two molecular glue degraders of the transcription factor, WIZ, which lead to pronounced de-repression of HbF. Using a chemical biology approach, the first molecule, dWIZ-1, was identified by phenotypic screening of a CRBN-biased chemical library for compounds that induce HbF without deleterious effect on erythroid cell proliferation and differentiation. Subsequent target deconvolution by global proteomics revealed WIZ as the target. WIZ knockout using two independent sgRNA elevated the proportion of γ-globin mRNA, HbF protein levels and HbF + cells (Figure 1) in vitro in primary human erythroblasts, validating WIZ as a previously unrecognized repressor of HbF. Optimization of dWIZ-1 led to a second tool compound, dWIZ-2, which was suitable to assess WIZ degradation and HbF induction in vivo due to its superior pharmacokinetic properties. dWIZ-2 treatment increased the proportion of γ-globin mRNA, HbF + cells and HbF protein levels in vitro in sickle cell disease patient-derived erythroblasts (Figure 2). Because many IMiDs lack activity in rodents, we used a humanized mouse model to measure the impact of dWIZ-2 on WIZ degradation and HbF expression in human erythroblasts in vivo. NBSGW mice humanized with CD34 + hematopoietic stem and progenitor cells from healthy human donors were orally dosed once daily with dWIZ-2 for 21 days. dWIZ-2 treatment resulted in robust dose-dependent WIZ degradation and an increase in total HbF and the proportion of HbF + human erythroblasts in the bone marrow. To further study the kinetics of HbF induction in the peripheral blood, and assess the effects of dWIZ-2 on steady-state hematopoiesis, we orally dosed naïve healthy cynomolgus monkeys with vehicle or 30 mg/kg dWIZ-2, once daily for 28 days. We detected elevated levels of γ-globin mRNA in the blood on day 28, reaching up to 37% of β-like globins. Additionally, on day 28, we observed up to 95% HbF + reticulocytes.dWIZ-2 was well-tolerated based on the absence of notable clinical observations during the study and consistent body weight for all animals. There were no dWIZ-2-related changes in hematology, coagulation or clinical chemistry measurements, specifically including reticulocytes, hematocrit and neutrophil counts. These findings support the development of an oral WIZ degrader as a globally accessible therapeutic approach for SCD.