Novel retroviral vectors to facilitate expression screens in mammalian cells

Abstract

As tools for functional genomics, expression profiling and proteomics provide correlative data, while expression cloning screens can link genes directly to biological function. However, technical limitations of gene transfer, expression, and recovery of candidate genes have limited wider application of genome-wide expression screens. Here we describe the pEYK retroviral vectors, which maintain high titers and robust gene expression while addressing the major bottleneck of expression cloning--efficient candidate gene recovery. By exploiting schemes for enhanced PCR rescue or strategies for direct isolation of proviral DNA as plasmids in bacterial hosts, the pEYK vectors facilitate cDNA isolation from selected cells and enable rapid iteration of screens and genetic reversion analyses to validate gene candidates. These vectors have proven useful to identify genes linked to cell proliferation, senescence and apoptosis.

Figure 1

Schematic representation and characterization of the pEYK retroviral vectors. All vectors contain the green fluorescent protein (GFP) gene. The pEYK1 vector incorporates subF, a mutated version of the suppressor tRNA gene supF, to serve as unique primer binding or probe sequence. The ble gene confers resistance to the bleomycin/phleomycin family of antibiotics. All pEYK vectors include ∼1.0 kb of retroviral gag gene sequence for enhanced retroviral packaging (19). The two ATGs within the extended gag region capable of generating gag-cDNA fusion proteins by alternative translation initiation were eliminated by site-directed mutagenesis (denoted by **). The 959 LTR (red arrow) consists of unique restriction enzyme sites (N-NotI and A-AscI) and a loxP site. Retroviral titers were determined as averages of triplicate infections with standard deviations. Titers are shown as 10⁶ infectious units per ml. The minimal quantity of genomic DNA needed to recover the GFP insert or the intact provirus from cells containing a single-copy provirus was determined for each vector. For pEYK3.1, amount of genomic DNA required for recovery by restriction digestion and intramolecular ligation (#) or cre-mediated excision and intramolecular ligation (##).

Figure 2

Comparative levels of expression for pEYK retroviral vectors. (A) FACS plots of BaF/3 cells infected with pMX, pEYK2, pEYK2.1 and pEYK3.1 retroviral vectors demonstrate differences in GFP fluorescence (pEYK1 was not included in the overlay to allow better visualization of profiles for other vectors). Expression levels were calculated as fold-increase in geometric mean fluorescence above background fluorescence, and represent averages of triplicate infections with standard deviations. Mutagenesis of the two internal ATGs (**) within the gag region of pMX yielded pEYK2, which showed modest but consistent increases in GFP expression over pMX. This enhanced packaging region was incorporated into the pEYK2.1 and pEYK3.1 vectors. (B) FACS plots of BaF/3 cells infected with pEYK3 and pEYK3.1 demonstrate a 4-fold increase in expression levels with the incorporation of the mutagenized gag sequences into pEYK3.1. The pEYK3 vector lacks the mutagenized gag sequence. For both plots, GFP expression is measured on a logarithmic scale.

Figure 3

Reversion analysis of the pEYK3.1-B/A retroviral vector. (A) The integrated provirus contains the BCR/ABL oncogene and is flanked by loxP sites (red arrows). (B) A population of BaF/3 cells containing the pEYK3.1-B/A provirus was infected with a bicistronic virus expressing both the cre and GFP-3M genes. In the absence of IL-3, the viability of GFP positive cells was 12% after 2 days. (C) Immunoblot analysis of the BCR/ABL-transformed population with an anti-c-ABL antibody shows the loss of BCR/ABL expression after cre-medated excision (lane 2). Densitometric analyses demonstrated an ∼80% decrease in BCR/ABL protein upon cre-mediated excision, with endogenous c-abl serving as a loading control.

Figure 4

pEYK retroviral vectors enable sequential iteration of functional genomic screens using different methods for proviral recovery.