Obtaining Efficient and Specific Gene Delivery

An ideal retroviral vector is one that could specifically home in on its target cell in the body and limit its transduction to only that type of cell. This would allow the in vivo delivery of the vector and greatly facilitate the clinical procedure for gene therapy. The entry of a retrovirus into a cell is determined in large part by the properties of its envelope glycoprotein and the specificity of the interaction of that protein with its receptor. To some extent, these restrictions can be circumvented by the use of heterologous fusion proteins to pseudotype the vector particles. However, other host range restrictions also exist, including several postentry blocks to transduction. An obvious example of this is the requirement for nuclear membrane breakdown for MuLV entry to the nucleus, but other less well-characterized resistance mechanisms are also present in some cells. For example, certain human cell lines that were poorly transduced by MuLV-based vectors were shown to be more susceptible to transduction by vectors based on GALV (32). In addition, the LTR promoter can also be considered a determinant of tropism and if gene expression is to be driven from that promoter, then its function in a particular cell type will also be an important consideration.

Current clinical protocols for retroviral vectors use an ex vivo approach. Because many of the cells to be transduced by the vectors express a high level of the natural amphotropic MuLV receptor and are actively dividing at the time of exposure to the vector (either naturally or as a result of culturing conditions), they are relatively easily transduced by MuLV vectors coated with the amphotropic MuLV Env. An important exception at present are primitive hematopoietic stem cells, which are reported to have a low level of the ampho-tropic receptor and be poorly transducible (41). However, the use of both the GALV and VSV-G fusion proteins has gone some way toward enhancing transduction, as has the use of lentiviral vectors (42,43).

Pseudotyping with natural viral fusion proteins that interact with different cell surface receptors, such as the amphotropic, xenotropic, polytropic, and 10A1 MuLV Env proteins and the GALV and VSV-G proteins, may provide enhanced transduc-tion of a particular cell type ex vivo, but these are all still broad host range proteins that do not provide much specificity. This therefore limits the ability of such vectors to be useful in vivo because introducing the vectors systemically would result in the particles binding to the majority of cells that they encountered and being diluted out before reaching their target cells. The problem can be quantitated. The human body contains approximately 5 X 1013 cells. Using concentrated stocks of retroviral vectors pseuodtyped with VSV-G (which have been reported to give titers of up to 109/mL), and infusing 100 mL of such a vector into a patient, would result in the delivery of about 1011 active vector particles. Even if every vector particle were 100% efficient, only 1 cell in 500 could possibly be transduced, and this scenario does not take into account the sequestering of the vectors into the first tissue that they come into contact with (typically the lungs), or the inactivation of particles in vivo by both innate immunity and more specific humoral responses. In addition, there will very likely be detrimental side effects resulting from the delivery of the vector to nontarget tissues as a result of the broad host range of the vector. An important step towards the in vivo use of retroviral vectors will clearly be the development of retroviral particles that can preferentially bind to and transduce their target cells and can be manufactured at a high titer.

0 0

Post a comment