Reducing Stem Cell Loss and Ex Vivo Differentiation during Lentivirus Gene Transfer to Murine Stem Cells - an Ultra-Short Transduction Protocol.

Most gene transfer models using Moloney murine leukemia virus (MLV) - derived vectors to target hematopoietic repopulating cells require progenitor cell enrichment and extended ex vivo culture for efficient long-term marking. Both may result in qualitative, and/or quantitative, loss of stem cells thereby limiting gene transfer rates in vivo. This can be a critical obstacle in candidate applications with exhausted autologous stem cell pools, such as Fanconi Anemia. Among the advantages of HIV-derived lentivirus vectors is their ability to transduce non dividing cells, permitting shortened ex vivo culture durations while maintaining gene transfer to long-term repopulating cells. We have previously reported long-term gene transfer rates of 12–40% after VSV-G/ lentivirus vector transduction of murine stem cells by targeting unseparated marrow cells after reduced prestimulation and a single 12 hour vector exposure (Kurre et al., Mol. Ther. 2004 Jun;9(6):914–22). We herein report studies showing maintenance of gene transfer efficiency in this model at drastically reduced ex vivo vector exposure times. In initial in vitro experiments we studied cytokine support, vector particle density, and minimum exposure duration requirements for efficient gene transfer to unseparated marrow cells. We determined that fibronectin fragment support was critical in maintaining minimum gene transfer efficiencies, even during brief 1, or 3-hour exposures. In an effort to extend these in vitro findings targeting a mixed leukocyte population and explore the feasibility in vivo, we next performed repopulation experiments in myeloablated murine recipients. Unseparated marrow cells harvested from donor animals were depleted of red blood cells, washed and immediately transduced on fibronectin fragment in the presence of murine stem cell factor. Following a 1 hour exposure to lentivector (VSV-G/RRLsin-cPPThPGK-EGFPwpre), cells were washed repeatedly, resuspended and injected into myeloablated recipients (n=10). Animals showed ready hematopoietic reconstitution and demonstrated average GFP marking of 31% (range: 17–41.2%) in peripheral blood 20 weeks after transplantation. Gene marking in secondary recipients 9 weeks after reconstitution (n=15, 3 recipient animals per donor) persisted at 29% on average (range 14.9–66%). Results also demonstrate transduction of granulocytes, B- and T-lymphocytes, as well as stable long-term GFP expression in primary and secondary animals. Copy number determination by real-time PCR in marrow cells from primary recipients shows an average of 4 proviral copies (range 2.1–8.1) per GFP-expressing cell.

Our studies confirm that HIV-derived lentivirus vectors are ideally suited for the transduction of murine long-term repopulating cells. We hypothesize that ultra-short transduction actively preserves stem cell content in the inoculum. Moreover, this protocol represents an ideal platform for subsequent in vivo selection to achieve complete phenotype correction and high-level therapeutic chimerism required for some applications. We anticipate that our strategy may prove particularly useful in situations where the target stem cell quantity is greatly limited and cells are of poor ex vivo viability.