Lentiviral vectors based on equine infectious anemia virus (EIAV) stably integrate into dividing and nondividing cells such as neurons, conferring long-term expression of their transgene. The integration profile of an EIAV vector was analyzed in dividing HEK293T cells, alongside an HIV-1 vector as a control, and compared to a random dataset generated in silico. A multivariate regression model was generated and the influence of the following parameters on integration site selection determined: (a) within/not within a gene, (b) GC content within 20 kb, (c) within 10 kb of a CpG island, (d) gene density within a 2-Mb window, and (e) chromosome number. The majority of the EIAV integration sites (68%; n = 458) and HIV-1 integration sites (72%; n = 162) were within a gene, and both vectors favored AT-rich regions. Sites within genes were examined using a second model to determine the influence of the gene-specific parameters, gene region, and transcriptional activity. Both EIAV and HIV-1 vectors preferentially integrated within active genes. Unlike the gammaretrovirus MLV, EIAV and HIV-1 vectors do not integrate preferentially into the promoter region or the 5' end of the transcription unit.

Lentiviral vectors are being developed to satisfy a wide range of currently unmet medical needs. Vectors destined for clinical evaluation have been rendered multiply defective by deletion of all viral coding sequences and nonessential cis-acting sequences from the transfer genome. The viral envelope and accessory proteins are excluded from the production system. The vectors are produced from separate expression plasmids that are designed to minimize the potential for homologous recombination. These features ensure that the regeneration of the starting virus is impossible. It is a regulatory requirement to confirm the absence of any replication competent virus, so we describe here the development and validation of a replication competent lentivirus (RCL) assay for equine infectious anaemia virus (EIAV)-based vectors. The assay is based on the guidelines developed for testing retroviral vectors, and uses the F-PERT (fluorescent-product enhanced reverse transcriptase) assay to test for the presence of a transmissible reverse transcriptase. We have empirically modelled the replication kinetics of an EIAV-like entity in human cells and devised an amplification protocol by comparison with a replication competent MLV. The RCL assay has been validated at the 20 litre manufacturing scale, during which no RCL was detected. The assay is theoretically applicable to any lentiviral vector and pseudotype combination.