681. Inadvertent Transduction of Recipient Tissues after Transplantation of Ex Vivo Lentivirally Transduced Hematopoietic Cells

Abstract

Hematopoietic stem cells are frequent targets for genetic modification, and readily repopulate the bone marrow after successful ex vivo transduction with lentiviral vectors derived from Human Immunodeficiency virus (HIV-1). Vector particles bear envelope (Env) glycoproteins as a prerequisite for cell surface receptor engagement and as a limitation to successful target cell transduction. However, particle attachment can also occur without cellular uptake, can be non-specific, unrestricted, and reversible. These observations prompted us to investigate the magnitude and kinetics of attachment and release of infectious vector particles from hematopoietic cells after ex vivo transduction culture. We used 3rd generation HIV-1 lentivectors pseudotyped with Vesicular stomatitis virus G (VSV-G) Env, expressing EGFP from an EF-1alpha promoter to transduce whole murine bone marrow cells. Following a 1-hour vector exposure at a multiplicity of infection (MOI) of 5, cells were carefully washed, resuspended and placed alongside 293T cells plated the day prior. Remarkably, the percentage of GFP- expressing 293T cells 72 hours later reached 80%. Twenty percent of secondary targets (293T) expressed GFP even after 4 thorough washes of the hematopoietic target cells (=carrier cells with a ratio to secondary targets of 5:1). Attachment and release kinetics correlated with MOI at transduction of carrier cells, but were independent of pseudotype, duration of ex vivo transduction culture (1 |[ndash]| 24 hours), and direct cell- cell contact. Moreover, vector particle attachment to carrier cells rescued infectivity on secondary targets in the presence of serum, whereas transduction was greatly reduced after serum pre-incubation of particles by themselves. We next examined the potential for in vivo dissemination and host tissue transduction after transplantation of ex vivo transduced whole bone marrow cells (MOI 5, 1 hour exposure, 2 sequential PBS washes). We injected 1|[times]|106 vector exposed CD45.1 (=carrier) cells into isotype mismatched (CD45.2), non-myeloablated C57BL/6 recipients. In serial follow up studies after transplantation and at animal sacrifice we used flow-cytometry to demonstrate GFP marking in recipient CD45.2 peripheral blood cells (up to 11%), bone marrow (2%), and spleen (2%). GFP marking was confirmed by real-time PCR analysis. Additional immunohistochemistry showed CD45-negative (as well as CD45- positive) GFP-expressing cells in liver and spleen - with rare events in muscle, lung and kidney. Vector lot and serum from animals tested repeatedly negative for p24(HIV-gag), excluding gain of replication competency. In summary, non-specific attachment and release result in substantial transfer of infectious particles to secondary targets and lead to serum complement protection in vitro and in vivo. The observed transduction of diverse host tissues in murine recipients after ex vivo gene transfer has obvious and wide-ranging implications from a biosafety perspective. In addition, our observations may affect the interpretation of studies on biodistribution and homing of ex vivo lentivirally-tagged cells and those employing stem cell gene therapy in submyeloablative conditioning regimens.