We compared the efficiency of transduction by an HIV-1-based lentiviral vector to that by a Moloney murine leukemia virus (MLV) retroviral vector, using stringent in vitro assays of primitive, quiescent human hematopoietic progenitor cells. Each construct contained the enhanced green f luorescent protein (GFP) as a reporter gene. ؉ and CD34 ؉ CD38 ؊ cells (13.5 ؎ 2.5%, n ‫؍‬ 11 and 12.2 ؎ 9.7%, n ‫؍‬ 4, respectively). The lentiviral vector is clearly superior to the MLV vector for transduction of quiescent, primitive human hematopoietic progenitor cells and may provide therapeutically useful levels of gene transfer into human hematopoietic stem cells.

Lentiviral vectors have been proposed as a more efficient alternative to Moloney murine leukemia virus-based retroviral vectors for transduction of human hematopoietic progenitors and stem cells. These studies were designed to evaluate the conditions that influence transduction frequency of CD34(+) progenitors, with the goal of optimizing efficiency of stable gene transfer with lentiviral vectors. CD34(+) human cord blood cells and 293 cells were transduced with a human immunodeficiency virus (HIV)-1 derived lentiviral vector pseudotyped with vesicular stomatitis virus glycoprotein and carrying an internal human cytomegalovirus promoter driving enhanced green fluorescent protein (eGFP) expression. Using fluorescence-activated cell sorting analysis of eGFP, we observed pseudotransduction beginning at the time of vector addition and lasting up to 24 h in CD34(+) cells and up to 72 h in 293 cells. Integrase-defective lentiviral vector caused transient eGFP expression for up to 10 days in CD34(+) cells and for up to 14 days in 293 cells. Protamine sulfate conferred no increase in transduction efficiency of CD34(+) cells on fibronectin-coated plates. Transduction frequency was related directly to vector concentration and not to multiplicity of infection across the ranges tested. First- and second-generation lentiviral vectors transduced CD34(+) cells equally, demonstrating a lack of dependence on HIV-1 accessory proteins. These findings will be useful for the optimal utilization of this new class of vectors for transduction of human hematopoietic stem cells.

Lentiviral vectors have undergone several generations of design improvement to enhance their biosafety and expression characteristics, and have been approved for use in human clinical studies. Most preclinical studies with these vectors have employed easily assayed marker genes for the purpose of determining vector titers and transduction efficiencies. Naturally, the adaptation of these vector systems to clinical use will increasingly involve the transfer of genes whose products may not be easily measured, meaning that the determination of vector titer will be more complicated. One method for determining vector titer that can be universally employed on all human immunodeficiency virus type 1-based lentiviral vector supernatants involves the measurement of Gag (p24) protein concentration in vector supernatants by immunoassay. We have studied the effects that manipulation of several variables involved in vector design and production by transient transfection have on vector titer and infectivity. We have determined that manipulation of the amount of transfer vector, packaging, and envelope plasmids used to transfect the packaging cells does not alter vector infectivity, but does influence vector titer. We also found that modifications to the transfer vector construct, such as replacing the internal promoter or transgene, do not generally alter vector infectivity, whereas inclusion of the central polypurine tract in the transfer vector increases vector infectivity on HEK293 cells and human umbilical cord blood CD34 ؉ hematopoietic progenitor cells (HPCs). The infectivities of vector supernatants can also be increased by harvesting at early time points after the initiation of vector production, collection in serum-free medium, and concentration by ultracentrifugation. For the transduction of CD34 ؉ HPCs, we found that the simplest method of increasing vector infectivity is to pseudotype vector particles with the RD114 envelope instead of vesicular stomatitis virus G glycoprotein (VSV-G). OVERVIEW SUMMARYWe have analyzed several variables involved in the design and production of lentiviral vectors for the effect they have on vector titer and infectivity. Wide variability in the amounts of input transfer vector, packaging, and envelope plasmids, as well as changes to the transfer vector construct, are tolerated with no significant change in vector infectivity. We found that inclusion of the central polypurine tract, concentration by ultracentrifugation, and collection in serum-free medium increase the infectivity of vectors on HEK293 cells and human umbilical cord blood CD34 ؉ HPCs. In addition, use of the RD114 envelope instead of VSV-G dramatically increases the infectivity of lentiviral vectors when they are used to transduce CD34 ؉ HPCs, but the titers and infectivities of RD114-pseudotyped vectors as determined on HEK293 cells is misleading, because of poor performance of the RD114 envelope on these cells.