Mystery solved: VSV-G-LVs do not allow efficient gene transfer into unstimulated T cells, B cells, and HSCs because they lack the LDL receptor

Fouzia Amirache, Camille Lévy, Caroline Costa, Philippe-Emmanuel Mangeot, Bruce E. Torbett, Cathy X. Wang, Didier Nègre, François-Loïc Cosset and Els Verhoeyen
This article has an Erratum 123(23):3682

To the editor:

Vesicular stomatitis virus (VSV) G-protein pseudotyped lentiviral vectors (VSV-G-LVs) signify a major advancement in the gene and immunotherapy field as illustrated by successful clinical trials, for example, for Wiskott Aldrich Syndrome and leukodystrophies.1

Although VSV-G-LVs allow efficient transduction of nondividing cells,2 they do not provide efficient transduction of quiescent T cells, B cells, and hematopoietic stem cells (HSCs), which hampers their application in gene and immune-therapy areas where conservation of cell phenotype is essential. Although these hurdles can be overcome in lymphocytes by LVs pseudotyped with measles virus envelope proteins (MV-LVs3-5), the reason as to why VSV-G-LVs were not efficient for gene transfer in these quiescent cells, and in particular in HSCs, remains unclear. Recently, Finkelstein et al revealed a long-kept secret of VSV by identifying its receptor, the low-density lipid receptor (LDL-R), explaining its broad tropism.6 This finding prompted us to evaluate LDL-R levels on unstimulated T, B, and CD34+ cells. Strikingly, we confirmed a very low expression of LDL-R, coinciding with VSV-G-LV–mediated poor transduction in these 3 cell lineages (Figure 1A). Stimulation of T cells through the T-cell receptor or of human CD34+ (hCD34+) cells with “early-acting cytokines” remarkably upregulated the LDL-R surface expression and permitted efficient VSV-G-LV transduction. In contrast, B-cell receptor stimulation augmented LDL-R expression only marginally, in agreement with poor VSV-G-LV transduction levels (Figure 1A and Frecha et al4). Binding of the different cell lineages with VSV-G-LVs was detected by incubation with the VSV-G-LVs followed by HIV capsid (p24) detection. VSV-G-LVs bound efficiently to stimulated T and hCD34+ cells but not B cells and barely attached to their resting counterparts (Figure 1B). In contrast, MV-LVs efficiently attached to both stimulated and unstimulated cells (Figure 1B). Next, we used particles formed by VSV-G protein (gesicles7) incorporating high levels of green fluorescent protein (GFP) through a farnesylation tag to verify fusion of VSV-G protein with 3 cell lineages (Figure 1A, right panels). Resting T, B, and CD34+ cells showed a poor GFP signal upon contact with GFP-loaded gesicles, while the GFP signal was evident for prestimulated cells, except for B cells (Figure 1A), confirming the presence of VSV and thus the VSV-G-LV receptor, LDL-R. Accordingly, VSV-G-LV transduction of resting T and B cells also resulted in very low levels of reverse-transcribed viral DNA.4

Figure 1

Low expression of LDL receptor on resting T cells, B cells, and CD34+ cells limits VSV-G-LV binding, fusion, and transduction of these gene-therapy targets. (A) Unstimulated human T cells, B cells, and CD34+ cells (G0) or 24-hour prestimulated (stim) T cells (anti-CD3 + anti-CD28 + IL-2), B cells (SAC + IL-2), and hCD34+ cells (TPO + SCF + Flk-3L) were transduced with a GFP-encoding VSV-G-LV vector at an MOI = 50 (T and B cells) or MOI = 100 (CD34+ cells) and GFP+ cells were analyzed at day 3 posttransduction by FACS (see supplemental Methods, available on the Blood Web site); for LDL-R detection, freshly isolated or 24-hour prestimulated cells (see above) were incubated with the anti–LDL-R antibody (mouse mAb; R&D Systems) followed by staining with anti-mouse APC antibody (white open histograms), a control incubation with the latter antibody alone was performed (gray filled histogram); for fusion detection, freshly isolated or 24-hour prestimulated cells were incubated overnight with GFP gesicles7 at 4°C to allow only binding or at 37°C to allow binding followed by fusion. The cells were then treated with trypsin to remove the GFP gesicles at the cell surface that did not fuse with the cells. (B) Equivalent quantities of VSV-G-LV or LV particles without envelope (measured by p24 content) were incubated with freshly isolated or 24-hour prestimulated cells (2E5 cells) for 1 hour at 4°C and then washed 4 times to remove unbound vector particles. The cells were pelleted and the cell-associated HIV capsid content (p24) was determined by ELISA (means ± SD; n = 3). The p24 signal for nonenveloped LVs was used as reference. (C) Entry through LDL-R was evaluated by blocking with a monoclonal antibody (C7, aLDL-R at 5 μg/mL; Santa Cruz Biotechnology) or by competition with soluble LDL receptor at 0.5 μg/mL (LDL-R 0.5; R&D Systems) or 5 μg/mL (LDL-R 5). A 1-hour preincubation of the prestimulated T cells, B cells, and CD34+ cells with either blocking agent was performed before transduction with GFP-encoding VSV-G-LVs (MOI 50 for T and B cells; MOI 100 for CD34+ cells) or MV-LVs (MOI of 10) for 48 hours, followed by FACS analysis for detection of GFP+ cells (means ± SD; n = 3). aLDL-R, anti-low density lipid receptor antibody; ELISA, enzyme-linked immunosorbent assay; FACS, fluorescence-activated cell sorter; IL, interleukin; mAb, monoclonal antibody; MOI, multiplicity of infection; SAC, staphylococcus aureus Cowan; SCF, stem cell factor; TPO, trombopoietin. Blood samples were obtained from healthy donors after informed consent and after local ethical committee approval in accordance with the Declaration of Helsinki.

Finally, we confirmed the requirement for VSVG-LV entry and transduction through the LDL-R and its family members using an anti–LDL-R antibody or by competition with soluble LDL-R, resulting in reduction or almost complete inhibition of transduction, respectively (Figure 1C). In contrast, MV-LVs were not sensitive to these LDL-blocking or -competing agents. Interestingly, IL-7–stimulated T-cell VSV-G-LV transduction8 coincided with LDL-R upregulation and was inhibited upon LDL-R blocking. Additionally, low-level transduction in resting cells was lost upon LDL-R blocking (data not shown).

In conclusion, although cellular postentry blocks may still play a role in VSV-G-LV transduction of resting T cells, B cells, and HSCs, we confirmed here that VSV-G-LV entry is compromised by the low expression of the VSV receptor LDL-R and its family members. Therefore, other LV pseudotypes (eg, MV-LVs) are more adapted for gene transfer in these invaluable resting gene-therapy targets.9


Contribution: F.A., C.L., C.C., D.N., and C.X.W. performed and designed experiments; B.E.T. and F.L.C. discussed results; and E.V. coordinated the project, designed and performed the experiments, analyzed the data, discussed results, and wrote the manuscript.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Els Verhoeyen, CIRI/EVIR, ENS de Lyon, 46 Allée d’Italie, 69364 Lyon Cedex 07, France; e-mail: els.verhoeyen{at} or els.verhoeyen{at}; and François-Loïc Cosset, CIRI/EVIR, ENS de Lyon, 46 Allée d’Italie, 69364 Lyon Cedex 07, France; e-mail: flcosset{at}


  • F.A., C.L., and C.C. contributed equally.

  • The online version of this article contains a data supplement.