Sequence of the yeast protein expression plasmid pEG(KT)

Abstract

The plasmid pEG(KT) is a widely used plasmid for expressing high levels of GST fusion proteins in the yeast Saccharomyces cerevisiae. Unfortunately, a complete sequence file has been lacking, thus complicating efforts to design cloning projects or to modify the plasmid for other uses (e.g. exchanging selection markers, epitope tags or protease cleavage sites to remove the epitope tag). Here, the complete sequence of the pEG(KT) plasmid is reported, thus facilitating its use. Additionally, its use as a vector backbone for high-level expression of a TAP-tagged protein is shown.

Introduction

A variety of systems can be used to express and purify large amounts of proteins for biochemical and structural experiments, including bacteria, yeast and insect cells (Verma et al., 1998). Often, the desired protein is initially generated as a chimeric protein, encoded as a translational fusion to an epitope that facilitates purification of the protein (Arnau et al., 2006). Finally, it is generally useful to control the expression of the protein using an inducer, such as arabinose, IPTG or galactose (Stagoj et al., 2006; Terpe, 2006). Using these various strategies, proteins of many types are expressed and purified for a variety of purposes.

The yeast plasmid pEG(KT) was first reported in 1993 as a tool to express proteins such as Ras2 that require eukaryotic post-translational modification for full functionality (Mitchell et al., 1993). Many labs have since used this plasmid and its derivatives to produce large amounts of proteins for biochemical assays (e.g. Seaman et al., 1998; Goode et al., 1999; Duncan et al., 2001; Zhu et al., 2001). One shortcoming in further application of this very useful plasmid is the lack of a complete sequence file. We have sequenced the pEG(KT) plasmid to overcome this problem.

Materials and methods

Plasmid sequencing

pEG(KT) plasmid was purified using Qiagen Miniprep spin columns. Oligos were obtained from Operon (Huntsville, AL) and resuspended at 100 μM in water. The plasmid DNA was mixed with the appropriate oligos (Table 1) and sent for commercial sequencing, using fluorescent di-deoxy Sanger sequencing. Reads were typically 600–800 bp. Sequences were used in BLAST searches to compare to known vector backbones, and chromatograms of the sequencing runs were inspected to resolve points of conflict. Gene Construction Kit (Textco BioSoftware, West Lebanon, NH, USA) was used to merge all of the sequences into one circular plasmid file.

Table 1.

Oligonucleotides used for sequencing (listed 5′ to 3′)

Oligo No.	Sequence
548	GGGCTGGCAAGCCACGTTTGGTG
749	CGGTGTGGTGGGCCCAGG
750	CCTGGGCCCACCACACCG
753	GGCCGCTTTCATGGCCCTAC
1474	GTAAACGCGGGAAGTGGAGTCAGGC
1475	CGAGGACGCACGGAGGAGAGTC
1476	GGTCGTTCGGCTGCGGCGAGCGG
1477	GTGATTTAGGTGGTTCCAACAGTACCACCG
1478	CCATGAGTGATAACACTGCGGCC
1479	GGCCGCAGTGTTATCACTCATGG
1480	GTCAGAGGTTTTCACCGTCATCACC
1481	AGGTGGGTTGGGTTCTTAACTAGG
1711	CCTGTAGCGGCGCATTAAGCGC
1712	GCGGCATCAGAGCAGATTGTACTG

Yeast expression of proteins from pEG(KT)-based plasmid

BJ2168 protease-deficient yeast cells with the genotype MAT a leu2 trp1 ura3-52 prb1-1122 pep4-3 prc1-407 gal2 (Aris and Blobel, 1989; Jones, 1991) were transformed with plasmids using the LiAc method (Schiestl and Gietz, 1989). For large scale purification, plasmid-containing cells should be grown to an absorbance of at least 2.0 at 600 nm (A600) in synthetic medium supplemented with 8 mg/ml tryptophan and 2% w/v raffinose at 26 °C. Cultures were induced overnight (12–16 h) by the addition of 10 g/l yeast extract, 20 g/l Bactotryptone and 100 ml/l 20% w/v galactose at 26 °C. Harvested cells are then washed in water, resuspended in a small volume of water (~0.2 × pellet volume) and poured in a thin stream into liquid nitrogen for quick freezing. The frozen pellets should be stored at −80 °C. Cell pellets are broken into a powder under liquid nitrogen in a metal Waring blender and the powder reconstituted in a 2 × buffer to generate lysates for protein purification. For the samples used in Figure 2, a smaller-scale expression protocol was followed, which yields results that are qualitatively similar to the large-scale induction. For the small-scale expression approach, plasmid-containing protease-deficient yeast cells were grown to an absorbance of at least 1.0 at 600 nm (A600) in synthetic medium supplemented with 8 mg/ml tryptophan and 2% raffinose at 30 °C, and induced for 4 h in 2% galactose at 30 °C. Whole-cell extracts were then prepared, using precipitation by 10% ice-cold trichloroacetic acid, washing twice in ice-cold acetone, drying, and then breaking the cells open by glass bead lysis in protein sample buffer.

Figure 2.

High levels of Pan1–TAP are produced when growing cells in the absence of uracil and leucine. Cells grown in raffinose to stationary phase were induced with 2% galactose for 4 h and extracts from 0.2 OD₆₀₀ were resolved in each lane by 7.5% SDS–PAGE. (A) Whole cell extracts were stained with Coomassie blue to show relative protein levels. Bio-Rad molecular weight markers were used (bands from the top down correspond to: 200, 116.25, 97.4, 66.2 and 45 kDa). The Pan1–TAP expression in [pEG(KT)] vs. (pRS.426) plasmids is compared; in both cases Pan1–TAP is under the control of the galactose promoter and is expressed in protease-deficient yeast. The additional absence of leucine from the medium (−URA, −LEU) selected for yeast cells with very high plasmid copy number, which leads to enhanced expression of the Pan1–TAP protein relative to cells grown with −URA selection. (B) An immunoblot of a gel similar to that shown in (A) was probed with HRP-conjugated donkey IgG, which binds to the TAP tag, and visualized by chemiluminescence

Protein analysis

Cell extracts were heated to 70 °C for 5 min and then separated by SDS–PAGE on 7.5% acrylamide gels. The gels were then either stained with Gel-Code Blue (Pierce, Rockford, IL, USA) or transferred to nitrocellulose by electrophoresis, blocked with 5% non-fat dried milk in Tris buffer, and incubated with HRP-conjugated donkey anti-rat antibodies to recognize the IgG-binding component of the TAP tag. After washing, the nitrocellulose blot was developed using chemiluminescence reagents (Pierce, Rockford, IL, USA) and the image captured using an Alpha Innotech Fluorochem (San Leandro, CA, USA).

Results

Using 14 oligonucleotides (Table 1), the sequence of the pEG(KT) plasmid was obtained. All restriction mapping that is based on the sequence obtained matched perfectly to the bands observed when the actual plasmid was digested and analysed by agarose gel electrophoresis. A map of the pEG(KT) plasmid is shown in Figure 1, with unique restriction enzyme sites in bold and restriction enzymes with two sites in plain font. The sequence of pEG(KT) is available as a nucleotide file through GenBank (Accession No. FJ526990). In addition, we have generated a sequence file of pEG(KG) using the polylinker sequence reported in Mitchell (1993) (Accession No. FJ526991).

Figure 1.

Map of pEG(KT). The plasmid pEG(KT) contains URA3 and leu2-d markers and a multiple cloning site following the GST gene under the control of the GAL1/10 promoter

In an effort to express high levels of the endocytic scaffold protein Pan1 (Wendland and Emr, 1998), we used a series of restriction enzyme digests and ligations to make modifications to a pEG(KT)-based plasmid that expresses an N-terminally truncated GST–Pan1 protein (Duncan et al., 2001; Miliaras et al., 2004); the end result was a plasmid encoding full-length Pan1 with a C-terminal TAP tag in the pEG(KT) backbone. These modifications were done prior to the sequencing of the pEG(KT) plasmid, and relied primarily upon restriction mapping, along with what could be gleaned from the literature regarding the fragments used to construct the original pEG(KT) plasmid. To generate the Pan1–TAP-expressing [pEG(KT)] plasmid, the GST tag was excised while retaining the intact galactose promoter, a 5′ fragment that included the natural PAN1 start codon and 5′ coding sequence was introduced, and a 3′ fragment encoding the C-terminal part of Pan1 fused to a C-terminal TAP tag was introduced (Toshima et al., 2005). The latter fragment was excised from a more conventional high-copy yeast expression plasmid encoding galactose-regulated full-length Pan1 with a C-terminal TAP tag in the pRS426 backbone (Christianson et al., 1992; Toshima et al., 2005).

The Pan1–TAP [pEG(KT)] and Pan1–TAP [pRS426] plasmids were transformed into the protease-deficient yeast strain BJ2168, using the URA3 selection marker of each plasmid. When cells were grown under inducing conditions in the presence of galactose as the carbon source, the Pan1–TAP protein was produced, while in the presence of the repressing carbon source dextrose, no or low levels of Pan1–TAP protein was translated (Figure 2). Higher levels of Pan1–TAP were synthesized from the pEG(KT)-based plasmid, as is evident from the Coomassie blue band visible near the 200 kDa marker in Figure 2A, and the signal from the Western blot detecting the TAP tag in Figure 2B. When the additional selection of the leu2-d allele, which has a truncated LEU2 promoter (Baldari et al., 1987), was applied by growing the cells in the absence of leucine, the highest levels of Pan1–TAP protein were produced (Figure 2). Under the conditions used for large-scale expression (see Materials and methods), we estimate that ~50–100 mg/l Pan1–TAP protein is expressed, based on Coomassie staining of whole cell extracts.

Now that the complete sequence of the pEG(KT) is available, future manipulations of the plasmid will be much easier to accomplish, including using such convenient approaches as cloning by homologous recombination (Ma et al., 1987).

Discussion

The utility of the pEG(KT) plasmid is extremely high, due to the independent ways in which expression levels can be controlled. First, it is a so-called 2 μplasmid, which imparts a high copy number in yeast cells (Rose and Broach, 1990). Second, its galactose promoter controls the expression of the desired protein product, which also contributes to high level of expression when the cells are grown in galactose-containing medium (Johnston and Davis, 1984). Third, it has the leu2-d allele of LEU2, which due to its very short promoter allows for a further selection of cells that have extremely high numbers of plasmids by growing the cells in medium lacking leucine (Erhart and Hollenberg, 1983).

The utility of the pEG(KT) plasmid is also limited, since it has URA3 as the primary selection marker, GST as the epitope tag, the epitope tag located at the N-terminus of the protein, and the thrombin cleavage site for removing the epitope tag. For some circumstances, it might be desirable to use a selection marker other than URA3, an epitope tag other than GST, a placement of the epitope at the protein’s C-terminus, or a different protease cleavage site, such as a TEV cleavage site (Parks et al., 1994). Thus, one might want to retain the galactose-regulated expression and the leu2-d additional high-level expression selection, but alter other features of the plasmid.

Now that the sequence of pEG(KT) is available, many more investigators may find that protein expression in yeast is a convenient and powerful way to express high levels of proteins that might be toxic in other cells. For example, the central and C-terminal parts of the Pan1 protein are toxic in bacteria (Sachs and Deardorff, 1992; and our unpublished observations), but these same regions can be expressed in yeast cells (Duncan et al., 2001; Miliaras et al., 2004; Toshima et al., 2005).