Rapid and efficient selection of human hematopoietic cells expressing murine heat-stable antigen as an indicator of retroviral-mediated gene transfer.

Recombinant retroviruses offer many advantages for the genetic modification of human hematopoietic cells, although their use in clinical protocols has thus far given disappointing results. There is therefore an important need to develop new strategies that will allow effectively transduced primitive hematopoietic target populations to be both rapidly characterized and isolated free of residual nontransduced but biologically equivalent cells. To address this need, we constructed a murine stem cell virus (MSCV)-based retroviral vector containing the 228-bp coding sequence of the murine heat-stable antigen (HSA) and generated helper virus-free amphotropic MSCV-HSA producer cells by transfection of GP-env AM12 packaging cells. Light density and, in some cases, lineage marker-negative (lin-) normal human marrow or mobilized peripheral blood cells preactivated by exposure to interleukin-3 (IL-3), IL-6, and Steel factor in vitro for 48 hours were then infected by cocultivation with these MSCV-HSA producer cells for a further 48 hours in the presence of the same cytokines. Fluorescence-activated cell sorting (FACS) analysis of the cells 24 hours later showed 21% to 41% (mean, 27%) of those that were still CD34+ to have acquired the ability to express HSA. The extent of gene transfer to erythroid and granulopoietic progenitors (burst-forming unit-erythroid and colony-forming unit-granulocyte-macrophage), as assessed by the ability of these cells to form colonies of mature progeny in the presence of normally toxic concentrations of G418, averaged 11% and 12%, respectively, in 6 experiments. These values could be increased to 100% and 77%, respectively, by prior isolation of the CD34+HSA+ cell fraction and were correspondingly decreased to an average of 2% and 5%, respectively, in the CD34+HSA- cells. In addition, the extent of gene transfer to long-term culture-initiating cells (LTC-IC) was assessed by G418 resistance. The average gene transfer to LTC-IC-derived colony-forming cells in the unsorted population was < or = 7% in 4 experiments. FACS selection of the initially CD34+HSA+ cells increased this value to 86% and decreased it to 3% for the LTC-IC plated from the CD34+HSA- cells. Transfer of HSA gene expression to a phenotypically defined more primitive subpopulation of CD34+ cells, ie, those expressing little or no CD38, could also be shown by FACS analysis of infected populations 24 hours after infection. These findings underscore the potential use of retroviral vectors encoding HSA for the specific identification and non-toxic selection immediately after infection of retrovirally transduced populations of primitive human hematopoietic cells. In addition, such vectors should facilitate the subsequent tracking of their marked progeny using multiparameter flow cytometry.