The “bottleneck” of current gene therapy approaches is the lack of an efficient gene delivery system. Retroviral vectors have been very useful for ex vivo gene delivery, but suffer two major limitations: (1) inability to infect post-mitotic cells, and (2) following implantation of the transduced cells, the expression of the transgene is “shut off”. The adenoviral vectors can be generated at high titers (>1011–1012 pfu/ml) but also have two major limitations: (1) Generation of CTLs to viral proteins, and (2) humoral response to viral proteins precluding repeat infection. The adeno-associated viral (AAV) vectors have not been exploited extensively partly due to inability to generate high titer helper free recombinant viruses. Furthermore, the size of the transgene that can be accommodated to generate viable vectors is limited.

To overcome some of these difficulties we have constructed lentiviral vectors based on HIV and pseudotyping with VSVG protein. We show that high titer helper free recombinant HIV vectors (∼108–109 CFU/ml) can be generated which can infect non-dividing cells in vitro. Furthermore, they can efficiently transduce monocyte-derived macrophages. We will show that HIV based vectors in contrast to the traditional retroviral vectors can be used to deliver genes directly in the brain, muscle, lung, liver, eye and islets. The efficiency of gene delivery is very high [over 70–80% of the neuronal cells at the site (∼2.5–3.0 mm) of injection] Furthermore, expression of the transgene is detected for over a six-month period of time, the longest time point tested so far. We believe lentiviral-based vectors will prove to be extremely valuable for in vivo gene delivery.