Dose-dependent transduction of vesicular stomatitis virus G protein-pseudotyped retrovirus vector into human solid tumor cell lines and murine fibroblasts.

We examined the transduction efficiency of a VSV-G (vesicular stomatitis virus G protein)-pseudotyped vector encoding beta-galactosidase (lacZ) into human solid tumor cell lines and murine fibroblasts, compared with that of an amphotropic vector carrying the same RNA sequence. The ratio of cells transduced with the VSV-G-pseudotyped vector corresponded closely to 1 - e(-m.o.i.), as predicted from a Poisson distribution of transduction to the entire cellular population, while this was not the case for the amphotropic vector. Here m.o.i. (multiplicity of infection) is defined as the ratio of input infectious units (titrated on the corresponding cell line) to the number of cells used for the transduction. At high m.o.i.s (values greater than 3), the VSV-G-pseudotyped vector transduced approximately 95% of the culture population of all cell lines examined. The transduction efficiency of the amphotropic vector, however, was not dose-dependent and reached a plateau or even decreased, especially at high m.o.i.; this may be attributable at least in part to the presence of envelope protein and noninfectious particles that compete for the receptor of infectious amphotropic virus. The copy number of integrated vector proviral DNA and the expression level of lacZ increased almost linearly with the dose of the VSV-G-pseudotyed vector, which could readily achieve multiple transduction of more than 10 copies per cell and afforded about 100-fold more transgene product than could be achieved with the amphotropic vector. These features of both the VSV-G-pseudotyped vector and the amphotropic vector were essentially unaffected by purification using centrifugation. These properties of the vector should be highly advantageous for gene transfer into entire populations of human tumor cell lines at a designed dosage. Copyright 1999 Academic Press.