Molecular mechanisms of retroviral integration site selection

Retroviral replication proceeds through an obligate integrated DNA provirus, making retroviral vectors attractive vehicles for human gene-therapy. Though most of the host cell genome is available for integration, the process of integration site selection is not random. Retroviruses differ in their choice of chromatin-associated features and also prefer particular nucleotide sequences at the point of insertion. Lentiviruses including HIV-1 preferentially integrate within the bodies of active genes, whereas the prototypical gammaretrovirus Moloney murine leukemia virus (MoMLV) favors strong enhancers and active gene promoter regions. Integration is catalyzed by the viral integrase protein, and recent research has demonstrated that HIV-1 and MoMLV targeting preferences are in large part guided by integrase-interacting host factors (LEDGF/p75 for HIV-1 and BET proteins for MoMLV) that tether viral intasomes to chromatin. In each case, the selectivity of epigenetic marks on histones recognized by the protein tether helps to determine the integration distribution. In contrast, nucleotide preferences at integration sites seem to be governed by the ability for the integrase protein to locally bend the DNA duplex for pairwise insertion of the viral DNA ends. We discuss approaches to alter integration site selection that could potentially improve the safety of retroviral vectors in the clinic.     

Vectors derived from simian immunodeficiency virus (SIV)

In contrast to other retroviruses, lentiviruses have the unique property of infecting non-proliferating cells. Thus vectors derived from lentiviruses are promising tools for in vivo gene delivery applications. Vectors derived from human primate and non-primate lentiviruses have recently been described and, unlike retroviral vectors derived from murine leukemia viruses, lead to stable integration of the transgene into quiescent cells in various organs. Despite all the safety safeguards that have been progressively introduced in lentiviral vectors, the clinical acceptance of vectors derived from pathogenic lentiviruses is subject to debate. It is therefore essential to design vectors derived from a wide range of lentivirus types and to comparatively examine their properties in terms of transduction efficiency and bio-safety. Here, we review the properties of lentiviral vectors derived from simian immunodeficiency virus (SIV).     

Transcriptional targeting of B cells with viral vectors

In order to optimize viral gene transfer into hematopoietic stem cells we developed retroviral and lentiviral vectors with B cell-specificity. Using fragments of the human CD19 promoter we demonstrate in mice that upon lethal irradiation and reconstitution with virus-treated bone marrow transgene expression is specific for the B cell-lineage. We compare various viral constructs with different promoter length and with or without B cell-specific enhancer regions in retro- and lentiviral backbones. Our data suggest that B cell-targeting for gene therapy approaches is feasible, leads to stable expression, and can be modulated by using different transduction and expression systems.     

Is it going to be SIN?

Insertional mutagenesis resulting in a leukaemia-like lymphoproliferative disease, as observed in the X-SCID (severe combined immunodeficiency) clinical trial using a gamma-retroviral vector that transferred a functional copy of the defective gene into hematopoietic precursor cells of affected children, sparked a debate about a ban on conventional gamma-retroviral vectors. This commentary summarizes the relevant data on this topic and concludes that there is no preclinical or clinical evidence as yet that SIN vectors, which self-inactivate the retroviral long terminal repeats (LTRs), will indeed show an improved safety profile. Conventional murine leukaemia virus (MLV) vectors can thus be used further in clinical gene therapy trials but require a thorough case-by-case risk-benefit analysis.