Improved human stem-cell transduction: the cat's meow

Comment on Lucas et al, page 51

An important objective in human gene therapy is improving the transduction efficiency of human hematopoietic stem cells. In this issue of Blood, Lucas and colleagues demonstrate that a significant improvement in transduction efficiency can be obtained using retroviral vectors pseudotyped with envelope proteins derived from feline retroviruses.

Despite the fact that hematopoietic stem cells (HSCs) from mice can be efficiently transduced with retroviral vectors, human HSCs have remained relatively refractory to genetic manipulation. This discrepancy is due to several biologic differences between species. For instance, mouse HSCs are more frequently in cycle and therefore more susceptible to integration of retroviral vectors. Restricted expression of certain retroviral receptor proteins on human HSCs is the other main cause of inefficient transduction. Mouse HSCs express relatively high levels of the ecotropic retroviral receptor, an amino acid transporter located on the cell surface, and are thus easily transduced. The problem is that the human homologue of this receptor differs sufficiently enough that it does not bind ecotropic virus envelope protein. Human HSCs express other retroviral receptors, including the amphotropic and gibbon ape leukemia virus (GaLV) receptors; however, vectors designed to enter HSCs via these proteins have yielded only modest levels of gene transfer.

In this issue of Blood, Lucas and colleagues have focused on this problem by testing new envelope/receptor combinations derived from feline leukemia virus type C (FeLV-C) and RD114 endogenous feline retrovirus. They found that human CD34⁺ CD38^– bone marrow cells expressed relatively high levels of the receptor for the RD114 envelope, and even higher levels of the FeLV-C receptor. This result was further evaluated by comparing HSC transduction efficiencies of oncoretroviral vectors pseudotyped with GaLV, RD114, or FeLV-C envelopes. Repopulation with transduced human HSCs was assayed in a stringent xenograft model based on engraftment in fetal sheep. The RD114 vectors gave improved transduction relative to the GaLV-pseudotyped vectors, confirming previous results by other investigators in alternative repopulation assays.^1,2  The key new finding was that transduction efficiency was substantially increased with the FeLV-C vectors, consistent with the receptor expression data in HSC-enriched populations. This represents a significant breakthrough for designing approaches to human gene therapy of hematopoietic disorders.

Several important questions arise from these results. Given the extensive data that lentiviral vectors, such as those based on the human immunodeficiency virus, have distinct advantages for gene therapy, it will be important to determine whether lentiviral vectors can be pseudotyped with FeLV-C envelope. An independent issue receiving considerable attention is the problem of insertional mutagenesis for clinical applications of HSC-based genetherapy. This study does show that the vector integration pattern is not fundamentally altered by the feline pseudotypes, a potential concern; however, the risk for insertional mutagenesis in applications other than X-linked severe combined immunodeficiency (XSCID) gene therapy is not yet clear. One thing is for sure, investigators will be evaluating the use of FeLV-C–pseudotyped vectors for gene therapy now that the cat is out of the bag. ▪