Increasing lentiviral vector titer using inhibitors of protein kinase R

Lentiviral vectors are increasingly used as vehicles for stable gene transfer ex vivo and in vivo (1). These gene transfer vectors have a large cloning capacity, stably integrate the transgene into the host cell genome, and can transduce nondividing cells such as hematopoietic progenitor cells (2). However, in comparison with the production of retroviral or adenoviral vectors, current methods for producing lentiviral vectors are inefficient, labor-intensive, and time- and reagent-consuming. The mechanisms that limit lentiviral vector production are poorly understood. One limiting factor may be the presence of double-stranded RNA (dsRNA) generated during lentiviral vector production or after transduction of producer cells by released lentiviral vector particles. Indeed, infection of cells by viruses often results in the accumulation of dsRNA in the cytoplasm, which then activates interferon-induced antiviral defense pathways (3). Upon binding to dsRNA, the interferon-inducible protein kinase (PKR) becomes activated. Activated PKR phosphorylates the eukaryotic initiation factor-2α (eIF-2α), resulting in a block in protein synthesis initiation. This leads to the inhibition of viral RNA translation, which affects the stability and processing of lentivirus-specific proteins (4–6). A second group of interferon-inducible proteins, the oligoadenylate synthetases (OAS), bind dsRNA and activate RNase L, which degrades both endogenous and viral RNA. Viruses have developed a variety of approaches to repress the innate mammalian antiviral cellular responses. One example is the adenovirus-encoded virus-associated RNA I (VAI RNA), which is transcribed by RNA polymerase III after virus infection. VAI RNA binds to PKR and prevents its activation by dsRNA (7).

In the present study, we investigated whether lentiviral vector production could be increased by the inhibition of PKR activity. Human immunodeficiency virus (HIV)-based lentiviral vectors were produced by the cotransfection of human embryonic kidney (HEK) 293T cells with three plasmids, using the calcium-phosphate precipitation method (2). pCMVR8.91 (15 µg) encodes the necessary HIV structural genes, pMD. G (5 µg) bears the vesicular stomatitis virus glycoprotein G (VSV-G) envelope cDNA, and pLoxC WGFP (20 µg) is the transfer vector carrying the enhanced green fluorescent protein (EGFP) gene under the control of the cytomegalovirus (CMV) immediate early enhancer/promoter region between modified HIV long terminal repeats (LTRs) (the three plasmids were generously provided by Dr. D. Trono, University Medical Center of Geneva, Geneva, Switzerland). Sixteen hours after transfection, the cell medium was replaced with 15 mL of fresh medium and incubated for an additional 24 h. The conditioned medium was collected and stored at −20°C. Lentiviral vector titers were determined by flow cytometry analysis of EGFP expression in transduced HeLa cells. Vector titers are expressed as transduction units per milliliter (TU/mL).

We employed two approaches to study whether inhibitors of PKR increase lentiviral vector yields. In the first approach, we cotransfected the three lentiviral vector production plasmids with different amounts of the pAdVAntage™ plasmid (Promega, Madison, WI, USA) that encodes VAI RNA, an inhibitor of PKR activation, as well as VAII RNA. In the second approach, we added different concentrations of 2-aminopurine (2-AP) (Sigma, St. Louis, MO, USA), an inhibitor of PKR phosphorylation (8). This inhibitor has been previously shown to increase the yield of different viruses grown on interferon-treated cells (9).

We observed that cotransfection with the pAdVAntage plasmid increased lentiviral vector titers in a dose-dependent manner (Figure 1, A and B). A maximal, two-fold increase was obtained when 10 µg of pAdVAntage plasmid were used. Mean fluorescence intensities in transduced cells were not modified by cotransfection with the pAdVAntage plasmid. This indicates that the presence of pAdVAntage increased the number of lentiviral vector particles but had no effect on the transgene expression level in the transduced cells (Figure 1B). Using 2-AP, we obtained a dose-dependent increase in lentiviral vector production. Compared to controls, almost 4-fold highertiters were obtained using 5 mM 2-AP (Figure 1, C and D). To determine the effect of PKR inhibitors on PKR activity under our experimental conditions, we analyzed the phosphorylation state of eIF-2α. We observed that cotransfection of 293T cells with the pAd-VAntage plasmid or incubation with 2-AP markedly decreased the levels of phosphorylated eIF-2α (Figure 1E). To assess activation of the OAS-mediated interferon response pathway, which ultimately results in RNase L activation, we attempted to compare the levels of OAS1 mRNA in lentiviral vector producing 293T cells with and without pAdVAntage and 2-AP. We were unable to detect OAS1 mRNA in any lentiviral vector-producing 293T cells, even after 50 PCR cycles. In other cell types, OAS1 was readily detectable at PCR cycle 25 (data not shown). Therefore, limitations on lentiviral vector production are not linked to the induction of OAS1 in 293T cells. Our results show that inhibitors of PKR activity can be used to increase lentiviral vector yields in HEK 293T producer cells.