Transfer of Contaminants in Adeno-Associated Virus Vector Stocks Can Mimic Transduction and Lead to Artifactual Results

Abstract

The potential of adeno-associated virus (AAV) vectors for gene transfer and gene therapy applications is currently being intensively investigated. Although much progress has been made in defining AAV vector biology, inconsistencies remain in the literature regarding the efficiency of AAV transduction in various cell types. In the course of exploring these differences, we have identified a problem associated with the use of AAV vector stocks that results in overestimation of gene transfer efficiencies. We show here that biologically active vector-encoded proteins can contaminate AAV vector stocks, especially cell lysate preparations that have not been further purified, and can be transferred in a virion-independent manner to target cells, a phenomenon called pseudotransduction. This observation is significant because impure cell lysate stocks have been widely employed in the AAV literature, and we demonstrate here that this phenomenon can occur with commonly used reporter proteins such as β-galactosidase and alkaline phosphatase. We conclude that although there are many potential explanations for apparently conflicting results in the literature, the possibility of pseudotransduction must be considered, especially when cell lysate stocks of AAV vectors have been employed. This artifact can be avoided by further vector purification.

Overview summary

We have found that marker gene protein contaminants present in AAV vector preparations can be taken up by cells exposed to the vectors and mimic true vector-mediated transduction. We documented this effect by following the procedure for making AAV vector stocks from cells expressing marker gene proteins while leaving out the viral components needed to make AAV vectors, and by showing that exposure of cells to such preparations resulted in marker protein transfer in the absence of virus. This phenomenon of pseudotransduction has previously been described for a retroviral vector with a coat protein derived from vesicular stomatitis virus, and we show here that pseudotransduction can also occur with an adenoviral vector. Pseudotransduction can be very efficient, resulting in the majority of cells being positive for the marker protein. These results document an artifact that complicates the interpretation of gene transfer experiments, but that can be avoided by further purification of viral vectors.