Introduction

Ectopic expression in mammalian cell culture systems is commonly used to characterize potential cellular properties and functions of a protein of interest (1,2,3). Ectopic protein expression is usually achieved by transfecting vectors containing a respective expression cassette driven by a constitutively or conditionally active promoter (4,5). With constitutively active promoters, two general expression strategies are applied. Transient transfection systems are used to achieve short-term expression of the protein of interest, and cells are commonly processed for analysis within 1–5 days of transfection. Although not required, selection of transfected cells by resistance to antibiotics that are effective within 24–48 h (e.g., puromycin) (6) aids in analyzing the respective effects of overexpression in transient transfection experiments. In contrast, selection of cells with antibiotics is a prerequisite for the second expression strategy, the generation of cell lines stably expressing the protein of interest. In stable cell lines, one or multiple copies of the respective expression vector have become integrated into the genome allowing persistent expression of the transgene, but only if its expression is not cytotoxic. The respective resistance gene and the protein of interest are usually encoded on the same expression vector, but in most of the available expression systems, their expression is driven by different promoters. This set-up has the disadvantage that only some of the cells selected express the protein of interest (i.e., the promoter driving the expression of the protein of interest is silenced during selection), and this is particularly observed in cases where ectopic expression of the protein of interest is cytotoxic. This shortcoming necessitates the clonal selection and expansion of cells expressing the protein of interest to allow the analysis of its potential function. However, the establishment of clonal cell lines is rather laborious (7). In addition, results obtained with single cell clones may be explained by clonal variation (i.e., the effects observed are not functionally related to the protein of interest), and thus, at least 2–3 clonal cell lines should be established and analyzed to ensure that the effects observed are indeed due to ectopic expression of the protein of interest.

To circumvent the shortcomings of mammalian protein expression systems, expression constructs that make use of internal ribosomal entry site (IRES) elements (8,9) have been developed. In such constructs, the protein of interest and the resistance marker protein are expressed from the same transcript. Thus, it is expected that any cell that is resistant to the respective antibiotic will express the protein of interest. However, since they are translated as individual proteins, expression rates of the protein of interest and of the resistance marker protein may differ markedly (i.e., the translation initiation rate of the protein of interest may be significantly lower than the IRES-mediated translation initiation rate of the resistance marker protein).

In the 1980s, Varshavsky and colleagues showed that a fusion protein consisting of ubiquitin fused to the N terminus of another protein is efficiently and site-specifically cleaved by ubiquitin-specific proteases (USPs), resulting in free ubiquitin and free protein of interest (10,11). In this manuscript, we made use of the ubiquitin-fusion protein approach and designed a eukaryotic expression construct encoding a fusion protein consisting of hemagglutinin epitope (HA)–tagged puromycin N-acetyltransferase (briefly, puro^r) fused to the N terminus of ubiquitin (HA-puro^r-ubi), which in turn is fused to the N terminus of a protein of interest (Figure 1A). The predicted advantage of this expression construct is that any cell that is resistant to puromycin generates the protein of interest and the resistance marker protein in a 1:1 ratio, since puro^r and the protein of interest are not only expressed from the same transcript, but moreover are expressed as a polyprotein that is processed to the respective free proteins by USPs (Figure 1A).

Figure 1. The ubiquitin-fusion protein system as a tool for rapid selection of cells ectopically expressing a protein of interest. (Click to enlarge)

Materials and Methods

Plasmids and In Vitro Translation

The HA-puro^r-ubi cDNA was generated by PCR-based approaches (for details, see Supplementary Material, available at www.BioTechniques.com) and cloned into pcDNA3 (Invitrogen, Carlsbad, CA, USA) and thus flanked by a cytomegalovirus (CMV) promoter, a T7 promoter (not shown), and a poly(A) site to ensure efficient expression in vitro (T7) and in mammalian cells (CMV). As in the original ubiquitin fusion protein system (11), the respective ubiquitin-fusion protein contains a mutated form of ubiquitin (ubiquitinK48R), in which lysine residue 48 was replaced with arginine to prohibit ubiquitination of the HA-puro^r-ubi fusion protein at K48 and diminish the possibility that HA-puro^r-ubi is targeted for proteasome-mediated degradation [K48-linked polyubiquitin chains serve as a recognition signal for the proteasome (12,13)].