One important part in our system is the viral vector we use for transferring genes into target cells. We succeeded to stably integrate our gene of interest into the genome of mammalian target cells by using a vector derived from the murine leukemia virus. The advantages of this vector are: its specificity for murine cells, making the viral work safe and easy; a very high efficiency for infection; and the ability of stable gene transfer into target genomes. Here we present results on these three qualities of our viral vector and how we optimized virus production and cell infection.

For optimal integration of the target gene, we tested how long the viral vector need for transducing murine cells. We incubated murine cells (NIH3T3) with the virus for different time intervals, washed the virus away, and counted fluorescent cells after 48 h. A high transduction rate was already reached after incubating the cells for five hours with the viral particles (Fig. 3).

We increased the transduction efficiency of our virus to over 80% in murine cell lines and achieved high infection rates in other cell lines expressing the murine CAT-1 receptor.

Protocols for virus production and target cell transduction were optimized. As producer cell line for viral particles we used Phoenix cells which stably express most of the viral genes. Before transfection with pMIG, the final plasmid to initiate virus production, we split the Phoenix cells several times to increase viability and push them into the exponential growth phase during production of the viral vector. In addition, supernatant containing the vector was harvested at distinct time points optimized for high viral titer. Target cells were infected by viral supernatant with polybrene which increases the probability for particles to reach their target cells. Since murine leukemia viruses only infect dividing cells, the phase of maximum division rate of the target cells was matched to the time point of infection.

Under these optimized conditions, the efficiency for transducing murine cell lines with our viral vector was over 80% (Fig. 5, Fig. 6). Since we had only 16% transduction efficiency at the beginning of the project, we would consider this a significant advance. To our knowledge, an optimized infection protocol with murine leukemia virus has not been published to date. The exact procedure can be found here (Methods).

Since the viral titer has direct impact on the infection rate of target cells, we assessed the transduction efficiency of the viral particles after dilution and concentration. Concentration was acheived by incubating the viral supernatant with hexadimethrine bromide (polybrene) as well as with chondroitin sulfate, leading to formation of complexes that were big enough for pelleting by centrifugation. Pellets contained the viral particles and were resuspended in a small volume of growth medium. Infecting murine cells with the concentrated vector lead to a higher transduction rate compared to the unconcentrated virus or the diluted virus (Fig. 7).

In addition, we determined the decay curve of the viral particles. Incubation at 37°C leads to a strong decrease of the infection rate with a half time of approximately 6.5 hours (Fig. 8). This short half life provides an enhanced safety regarding work and handling of our viral vector.

An important aspect for the function of our system as well as for its safety is the specificity of the vector regarding infection of different kind of cells. Cells that do not have this specific receptor are not targeted by the vector (Fig. 9; Fig. 10). To test the specificity of the system, different kinds of cells were incubated with the vector containing EGFP. EGFP is stably integrated by the system. We tested human cell lines, human embryonic kidney cells as well as human lung epithel carcinoma cells and breast cancer cells, for their susceptibility of being infected by the vector. In none of them, green fluorescent cells were observed (Fig. 10). Murine cells (NIH3T3) that express the mouse CAT-1 receptor were tested as a positive control (Fig. 10). Quantification by flow cytometry indeed showed that human cells were not transduced, whereas a large fraction of the mouse control cells were transduced with the same virus batch (Fig. 11, Fig. 12). 9E-G).