An ideal gene therapy vector for the potential gene therapy of b-thalassemia and sickle cell disease would be one with which high-efficiency transduction of primitive human HSCs could be achieved, and following erythroid differentiation, robust levels of expression of the transduced human b-globin gene could be obtained. The development of lentiviral vectors by a number of investigators has indeed achieved these objectives 
, but their long-term safety still remains an open question 
. We and others have described the development of the first generation of recombinant AAV2 vectors for the potential gene therapy of b-thalassemia and sickle cell disease 
, but in retrospect, it has become clear that the use of the WT AAV2 capsid, and the single-stranded nature of the vector genome, were major obstacles to achieving therapeutic levels of the human b-globin gene 
. In addition, the use of murine models of these diseases was not predictive of the potential efficacy of a number of alternative serotypes of AAV vectors 
Based on more recent studies by us and others 
, in which AAV6 was identified to be the most efficient serotype for transducing human HSCs, and our observation that B19p6 is more robust than HS2-βp for mediating erythroid lineage-restricted expression 
, we reasoned that combining these features might lead to the development of an ideal vector for the potential gene therapy of β-thalassemia and sickle cell disease, especially since the safety and efficacy of AAV vectors have now been established unequivocally in a number of Phase I/II clinical trials in humans 
. Indeed, the Y705+731F double-mutant scAAV6 vectors containing the B19p6, described here, were determined to be the most efficient in transducing primary human CD34+
cells, and mediating erythroid lineage-restricted transgene expression, both in vitro
and in vivo
. It is also possible that the transduction efficiency of these vectors can be augmented further, based on our recent observations that site-directed mutagenesis of specific surface-exposed serine and threonine residues improves the transduction efficiency of AAV2 serotype vectors 
, and most of these residues are highly conserved in all AAV serotypes.
The basic underlying molecular mechanism of increased transduction efficiency of the Y705+731F double-mutant scAAV6 vectors in human CD34+
cells is not readily apparent. Based on our recent studies with tyrosine-mutant AAV2 and AAV3 serotype vectors 
, we favor the hypothesis that improved intracellular trafficking and/or nuclear transport lead to the observed effect. However, the alternative hypothesis that a more efficient cellular receptor-mediated viral vector entry also play a role, cannot be ruled out since the extent of transgene expression from the B19p6 promoter in human CD36+
erythroid progenitor cells was ~20-fold lower than the more primitive CD34+
HSCs (compare A, B and C, D). Thus, additional studies are warranted to address these issues, as well as to identify the authentic receptor for AAV6 for human HSCs, since in a recent report, EGFR was recently identified to be the cellular receptor for AAV6 
, and Denard et al
reported that galectin 3-binding protein in human sera agglutinates AAV6 vectors, which resulted in decreased transduction efficiency of these vectors. In our studies, pre-treatment of CD34+
cells with EGF had no effect on the transduction efficacy of AAV6 vectors, and K562 cells, which are known to lack expression of EGFR 
, could be efficiently transduced by AAV6 vectors, which was inhibited by FBS (data not shown).
Although erythroid lineage-restricted transgene expression from the B19p6 promoter in primary human HSCs in vitro
has previously been reported 
, those studies were carried out with the first generation single-stranded AAV2 serotype vectors, which were clearly sub-optimal. We subsequently utilized scAAV1 and scAAV7 serotype vectors, and corroborated the erythroid cell-specificity of the B19p6 promoter in vivo
, those studies were carried out in murine HSCs, which were clearly not predictive for human HSCs. In the present studies, we documented sustained transgene expression in human HSCs, both in primary as well as in secondary transplant recipient mice. However, because of less than 1% engraftment of human cells in secondary transplant recipients, we were unable to document stable integration of the AAV proviral genomes. In this context, it is important to emphasize that the general conclusion that AAV genomes do not integrate, has largely been derived from previously published studies, all of which were carried out with post-mitotic cells and tissues, such as liver, muscle, brain, and retina, in which the AAV genomes remain episomal, although integration in liver has been reported by several investigators 
. In our previously published studies with primary murine HSCs, stable integration of the AAV proviral genomes has been documented in both primary as well as secondary transplant recipient mice 
, and in a more recently published collaborative study, we have also documented long-term transduction and multi-lineage engraftment of human HSCs in a mouse xenograft model 
. Thus, our working hypothesis has been that unlike in post-mitotic cells, AAV vectors do integrate in HSCs. The fact that in a recent report by Weltner et al
, all 4 reprogramming genes were shown to be integrated following AAV vector-mediated generation of induced pluripotent stem (iPS) cells, provides further support to our hypothesis.
The development of the optimized scAAV6-B19p6 vectors described here, with which high-efficiency transduction of human HSCs, and erythroid lineage-restricted expression can be achieved in vivo
, and the possibility that the transduction efficiency of these vectors can be further augmented by introducing additional mutations in surface-exposed specific serine and threonine residues, similar to those described for AAV2 
, bodes well for the eventual use of these vectors in the potential gene therapy of human hemoglobinopathies in general, and b-thalassemia and sickle cell disease in particular.