The Importance of the Human β-Globin Locus Control Region HS1 and HS4 Elements for Therapeutic β-Globin Gene Expression in β-Thalassemic Mice.

Globin gene transfer in autologous hematopoietic stem cells is a promising therapeutic option for subjects with β-thalassemia major. In this approach, high level, erythroid specific transgene expression is needed to correct ineffective erythropoiesis and hemolytic anemia following the delivery of few copies of therapeutic vector per cell. Several groups have successfully treated mouse models of severe hemoglobinopathies utilizing lentiviral vectors encoding β- or γ-globin genes placed under the transcriptional control of the human β-globin promoter and the HS2, HS3 and HS4 elements of the β-globin locus control region. The HS2 and HS3 elements are the most powerful and the best characterized single elements within the LCR. The relative importance of HS1 and HS4 is less well defined. We show here the major roles played by HS1 and HS4, which although not seen in MEL cells, are striking in β-thalassemic mice. The effect of HS1 element was tested in vectors, derived from the previously published TNS9 vector, that harbor different globin promoters (either 265, 615 or 1555bp in length). Addition of HS1 to vectors containing the 615bp or 1555bp promoters had no effect on average transgene expression per vector copy (VC) and even decreased average transgene expression from 38±3% (n=32 MEL cell pools) to 26±2% (n=23) of endogenous β-globin levels (p<0.001) in the context of the 265bp promoter. In vivo, however, addition of HS1 had a dramatic effect on globin expression. Transgene expression increased from 27±6% of the endogenous β-globin mRNA to 41±9% for vectors harboring the HS1 element (p<0.001), after normalization to vector copy number. On the Hb level, the vectors without HS1 element provided 4–6g/dl/VC, while addition of HS1 increased this value to 9g/dl/VC. To evaluate the effect of HS4 on gene expression, we created panel of vectors with truncations of 3′ or 5′ flanking regions of HS4. In vectors harboring LCR HS1-4, the 5′ truncation significantly decreased mean in vivo globin expression from 26±2.5% to 20±2% of the endogenous β-globin (p<0.001). A similar effect was observed for 3′ or 5′ truncations in vectors lacking HS1 element. The 5′ flanking region of HS4 was also replaced with an unrelated DNA spacer, the same size fragment of HS3 flanking region, or the human IFN-β S/MAR element. Only the addition of the S/MAR element rescued the function of HS4, restoring average globin expression to the level of vectors encoding the full HS4 element, which suggests that this region may contain a functional S/MAR element. This analysis underscores the importance of carefully analyzing the size and relative positioning of transcriptional control elements within tissue-specific vectors, as well as the critical importance of assessing these elements in animal models of disease. Based on this analysis, we are proceeding to a phase I clinical trial in subjects with β-thalassemia major, utilizing the TNS9.3 vector, which harbors the 615bp human β-globin promoter and HS2–3–4, providing curative levels of hemoglobin at 1 to 2 copies per cell. Addition of HS1 is a promising alternative strategy if higher levels of expression are eventually needed.