Lentivirus-delivered stable gene silencing by RNAi in primary cells

Abstract

Genome-wide genetic approaches have proven useful for examining pathways of biological significance in model organisms such as Saccharomyces cerevisiae, Drosophila melanogastor, and Caenorhabditis elegans, but similar techniques have proven difficult to apply to mammalian systems. Although manipulation of the murine genome has led to identification of genes and their function, this approach is laborious, expensive, and often leads to lethal phenotypes. RNA interference (RNAi) is an evolutionarily conserved process of gene silencing that has become a powerful tool for investigating gene function by reverse genetics. Here we describe the delivery of cassettes expressing hairpin RNA targeting green fluorescent protein (GFP) using Moloney leukemia virus-based and lentivirus-based retroviral vectors. Both transformed cell lines and primary dendritic cells, normally refractory to transfection-based gene transfer, demonstrated stable silencing of targeted genes, including the tumor suppressor gene TP53 in normal human fibroblasts. This report demonstrates that both Moloney leukemia virus and lentivirus vector-mediated expression of RNAi can achieve effective, stable gene silencing in diverse biological systems and will assist in elucidating gene functions in numerous cell types including primary cells.

FIGURE 1.

Sequence-specific knockdown of GFP expression by retrovirus-delivered GFP hairpin RNA expression. (A) Schematic presentation of Retrohair-slGFP, a replication incompetent retrovirus that contains a pol-III promoter (U6), a stem-loop cassette followed by a 5T pol-III termination signal, cloned 3′ to a truncated gag lacking a translation start sequence (ATG). Retrohair can be selected by addition of puromycin. (B) The stem-loop encoded hairpin RNA has sequence identity to a 21-nt region of GFP mRNA. (C) FACs analysis of GFP expression in HeLa-GFP infected with (1 and 3) empty vector controls, (2) Retrohair-slGFP1, (4) Retrohair-slGFP2, or (5) Retrohair-slGFP2mut. Overlay histograms in panels 2, 4, and 5 demonstrate the level of knockdown in cells infected with Retrohair-containing stem-loop cassettes (red) compared to cells infected with empty vector controls (green). (D) Northern blot analysis of GFP expression in HeLa-GFP cells described in C using a GFP-specific probe. Northern blot with a β-actin probe served as a loading control.

FIGURE 2.

Time course of selection of GFP knockdown in RNAi-stable HeLa-GFP cells. (A) FACS analysis of GFP expression in HeLa-GFP cells infected with Retrohair-lacking a stem-loop (empty, top panels) or with Retrohair-slGFP1 (-slGFP1, bottom panels). GFP expression was measured in HeLa-GFP either preselection, with 3 d, 2 wk, or 4 wk of puromycin selection or in HeLa-GFP after 2 wk of growth with puromycin selection followed by 2 wk of growth without puromycin selection. (B) Northern blot analysis of GFP expression in HeLa (lane C), in HeLa-GFP infected with Retrohair at (lane 1) 3 d (lane 2) 2 wk, (lane 3) 4 wk, or infected with Retrohair-slGFP1 at (lane 4) 3 d (lane 5) 2 wk, (lane 6) 4 wk, or at (lane 7) 2 wk with selection and 2 wk without puromycin selection. β-Actin expression served as a loading control. (C) Modified Northern blot analysis of the samples described in B using an RNA probe (sense strand) with complementarity to the antisense strand of the GFP hairpin RNA. Samples were normalized by total RNA content. Titration of GFP siRNA was used as a size marker and for quantitation by densitometry of small RNAs. (D) FACS analysis of GFP expression in HeLa-GFP infected with Retrohair-slGFP1 4 wk after selection on puromycin (green) or infected with Retrohair-slGFP1 4 wk after selection on puromycin and then superinfected with Retrohair-slGFP2 (red).

FIGURE 3.

Knockdown of GFP expression in DC-GFP. (A) Photomicroscopy of DC-GFP visualized by white light or by fluorescent light after infection with Retrohair or with Retrohair-slGFP1(slGFP1). (B) FACS of GFP expression in DC-GFP infected with the Retrohair or Retrohair-slGFP1 retrovirus. A bar graph depicts the percent decrease in GFP expression as measured by MFI. Error bars represent the average (±SD) of six experiments. (C) Modified Northern blot analysis of Retrohair and Retrohair-slGFP1-infected DC-GFP using an RNA probe (sense strand) with complementarity to the antisense strand of the GFP hairpin RNA. Samples were normalized by total RNA content. Titration of GFP siRNA was used as size marker and for quantitation by densitometry of small RNAs.

FIGURE 4.

Knockdown of GFP expression by lentivirus-delivered RNAi. (A) Schematic presentation of the lentiviral construct. Lentihair is a replication incompetent, self-inactivating retrovirus generated by cloning the U6 promoter and stem-loop cassette targeted to luciferase (Lentihair-slLUC) or GFP (Lentihair-slGFP1) 5′ of the cPPT. A puromycin resistance gene was cloned 3′ of the human hPGK promoter. (B) FACS analysis of GFP expression in HeLa-GFP cells (1) uninfected or infected with (2) Retrohair, (3) Retrohair-slGFP1, (4) Lentihair-slLUC or (5) Lentihair-slGFP. (C) FACS analysis of GFP expression in (1) uninfected, (2) Lentihair-slLUC, or (3) Lentihair-slGFP infected DC-GFP.