Lentivirus Vectors Containing a Band 3/γ-Globin Gene Flanked by Distinct Insulator Elements Are Resistant to Gene Silencing in Primary Mouse Erythroid Cells.

Abstract

Effective Gene Therapy of the hemoglobin β-chain disorders, Sickle Cell Disease (SCD) and β-thalassemia (β-thal), requires that viral vectors deliver a β-like globin gene into hematopoietic stem cells and express it at levels > 20% of that of endogenous α-globin. The β-globin gene is poorly expressed without sequences from the Locus Control Region (LCR), however, inclusion of LCR sequences in viral vectors introduces cryptic splicing and polyadenylation signals leading to inefficient virus production. Moreover, the insertion of LCR enhancer elements near oncogenes may result in the undesirable activation of those genes. Our approach to design safer vectors is to use erythroid-specific, enhancer independent, non-globin promoters to express γ-globin. Band 3 (B3) is the most abundant membrane protein of mammalian erythrocytes. We have previously shown that the minimal B3 promoter was localized within 350 bp upstream of the mRNA start site and had no enhancer. In transgenic mice the −350 B3 promoter was sufficient to direct high-level (~7% of α-globin/transgene copy), erythroid-specific expression of a linked γ-globin gene. However, γ-globin gene expression was variegated, reducing the level of B3/γ-globin mRNA and protein. No B3/γ-globin transcripts were initiated at the correct start site, with all of the transcripts initiating at multiple upstream sites. Finally, γ-globin expression was not detected in any of 17 spleen foci containing a retrovirus vector with the −350 B3/γ-globin gene. We hypothesized that the position effects and gene silencing we observed could be overcome by flanking B3/γ-globin gene with insulator elements. We tested this hypothesis in transgenic mice by flanking the B3/γ-globin gene with the chicken β-globin insulator element 5′ Hypersensitive Site (HS) 4 (ch5′HS4). We observed uniform expression of B3/γ-globin at therapeutic levels (19.8% of α-globin/transgene copy), with all transcripts initiating from the correct B3 transcription start site (Frazar et al. Mol Cell Biol 23:4753–63, 2003). Because we and others have found that vectors containing two internal copies of ch5′HS4 were prone to recombination and produced a low virus titer, we hypothesized that flanking the B3/γ-globin gene with distinct insulator elements would improve gene transfer and expression. To search for B3 insulator elements, we have developed a high throughput real-time PCR-based assay to search for HS across the B3 locus. We have identified an HS within exon 1 (~1.3 kb downstream of the transcription start site) and a cluster of HS that map between 5.7 kb and 8.7 kb upstream of the promoter to test for insulator activity. Using transgenic mice and K562 assays, we have also characterized insulator elements located in the Ankyrin (ANK) and α-spectrin (α-Sp) loci. Three different lentiviral vectors containing the B3/γ-globin gene flanked by combinations of the ch5′HS4, ANK and α-Sp barrier elements (e.g. ch5′HS4 and ANK) can be produced at high titer and deliver unrearranged copies of the flanked B3/γ-globin transgene to target cells. After transduction of primitive mouse hematopoietic progenitor cells with these vectors, γ-globin mRNA was detected in 100% of foci. We conclude that the ch5′HS4, ANK and α-Sp insulators improve the expression of the B3/γ-globin gene in viral vectors and that inclusion of B3 barrier elements will further refine these vectors.