A Cell-Based High-Throughput Screen For Novel Inducers Of Fetal Hemoglobin For Treatment Of Sickle Cell Disease, Cooley’s Anemia and β-Thalassemias

Decades of research has established that the most effective treatment for sickle cell disease (SCD) and Cooley’s anemia is increased fetal hemoglobin (HbF). Certain β-thalassemias may also benefit from fetal hemoglobin induction. Fetal hemoglobin normally accounts for less than 0.5% of total hemoglobin in adults; increasing levels to approximately 10% alleviates much of the pathophysiology associated with SCD. Hydroxyurea is the most widely available treatment for SCD that results in enhanced HbF production, but this drug is highly pleiotropic in its action and does not exclusively modulate γ-globin gene expression. Identification of a drug specific for inducing or reactivating γ-globin expression in pediatric and adult patients, with minimal off-target effects, continues to be an elusive goal. One hurdle has been an assay amenable to a high-throughput screen (HTS) of chemicals that displays a robust γ-globin off-on switch to identify potential lead compounds. An assay system developed in our lab to understand the mechanisms underlying the γ- to β-globin gene expression switch during development allowed us to generate a cell-based assay that was adapted for a HTS of 121,085 compounds from the libraries of the KU-HTS Laboratory (Prestwick, MicroSource, CMLD, Chembridge and ChemDiv compound libraries) and LCGC (OCL compound library). Transgenic mice were produced using a modified 213 Kb human β-globin locus yeast artificial chromosome (β-YAC). Two gene fusions were introduced into the β-YAC via homologous recombination in the host yeast, firefly luciferase was fused to the ^Aγ-globin promoter and Renilla luciferase was fused to the β-globin promoter. The resultant YAC was microinjected into fertilized mouse oocytes to produce transgenic mice. We used these mice to derive chemical inducer of dimerization (CID)-dependent bone marrow cells (BMCs) containing the γ-luc/β-luc β-YAC, which were employed in the HTS. We identified 232 primary screen actives that induced γ-globin 2-fold or higher. A 4-assay, 10-point dose-response secondary screen using the same CID-dependent γ-luc/β-luc β-YAC BMCs reconfirmed that 211 of these active compounds induced γ-globin ≥2-fold with minimal or no β-globin induction, minimal cytotoxicity and did not directly inhibit purified luciferase enzyme. Additional secondary assays in CID-dependent wild-type β-YAC BMCs and human primary erythroid progenitor cells confirmed the characteristics of seven of these 233 hits that were cherry-picked for further analysis. Four of the compounds were particularly promising, numbers 7, 42, 87 and 208. In CID-dependent wild-type β-YAC BMCs using the optimal dose for each compound, γ-globin mRNA induction ranged from 3- to 42-fold compared to 10-fold with sodium butyrate as measured by real-time qRT-PCR; F-cells ranged from 9.9-29.9% compared to 0.7% untreated and 15.9% treated with sodium butyrate as measured by flow cytometry. In human primary erythroid progenitor cells, the mRNA change was 1.6- to 3-fold compared to 1.75-fold with sodium butyrate and F-cells ranged from 9.1-29% compared to 5.7% untreated and 39.4% treated with sodium butyrate. Lead compounds will be tested in a pre-clinical β-YAC transgenic mouse model to determine their ability to induce HbF in vivo to aid development of these compounds for future clinical applications in hemoglobinopathies.