The targeted introduction of heterologous DNA to genomic locations by a simple polymerase chain reaction (PCR)-based strategy has been widely used for research, particularly with the fungi Saccharomyces cerevisiae and Schizosaccharomyces pombe (Bahler et al., 1998; Baudin et al., 1993; Knop et al., 1999; Krawchuk and Wahls, 1999; Longtine et al., 1998; Schneider et al., 1995; Tasto et al., 2001; Wach et al., 1994, 1997). These strategies have been shown to be powerful tools in systematic gene deletion, protein localization and protein complex purification (Gavin et al., 2002; Ho et al., 2002), as well as for single gene-function analysis. The strategy requires: (a) a pair of primers that contain within their 5′ region sequences of homology to the genomic target location; and (b) PCR-cassettes (also termed ‘modules’) that can be amplified using these primers. To make the technique most powerful and cost-efficient, we constructed a series of new cassettes and included in all of them identical primer-binding sites, which allow the amplification of all C-terminal tags with only one pair of primers per gene. An additional primer is needed for gene deletion (Knop et al., 1999) and a fourth primer for the introduction of sequences at the N-terminus (Figure 1).
In addition to the previously published 12 cassettes for C-terminal epitope tags (Knop et al., 1999), we present here a wider range of C-terminal tags as well as two new selection markers, both carrying dominant antibiotic-resistance genes. We also describe new cassettes that allow the replacement of the promoter of a given gene, with the optional addition of an N-terminal epitope tag to the gene. Nine promoters, five of them inducible, were cloned into different cassette plasmids.
The construction of PCR-cassettes is straightforward and can be done via standard cloning strategies (details provided upon request). Therefore, it will be easy to create new cassettes, e.g. to introduce new combinations of tags, makers and promoters (in the case of N-terminal tagging) by simple cloning procedures.
Materials and methods
Cassette plasmid construction
Standard techniques were used for DNA manipulations (Sambrook et al., 1989). The construction of the PCR-cassette pYM1-12 is described in Knop et al. (1999). The construction of the new cassettes is summarized in Table 1; the primers used are listed in Table 2 (further details can be obtained upon request). A comprehensive overview of all available C-terminal tagging cassettes, with regard to selection marker and tag, is provided in Table 3.
Table 1. Properties and construction of the new cassette plasmids
Table 3. Systematic table of all available pYM plasmids for C-terminal tagging and deletion
BFP is a very weak fluorescent protein. So far, we have not yet successfully used the BFP-modules. However, we provide the cassette since some strongly expressed proteins might be well detected when tagged with BFP.
A set of four primers allows to amplify all N- and C-terminal tags and to generate gene deletions. The principle of the primer design is explained in Figure 2. The amplification of the modules can cause problems, because the annealing sites for S1, S2 and S3 primers (Figure 2), which were chosen initially for the EUROFAN project, lead to self-annealing of the primers. Another problem is the high GC content of the natNT2 marker. To circumvent these problems, different PCR conditions have been used (Goldstein and McCusker, 1999). We present here one particular condition, which works well in several laboratories. One other reason for the failure of the PCR is often linked to the quality of the primers (see Discussion).
The pipetting scheme for a 50 µl reaction and the PCR cycle scheme are visualized in Figure 3A/B. A successful PCR gives a very strong band at the estimated size (Table 1, Figure 3C), when 3–5 µl of the PCR were analysed on a standard agarose gel. Some natNT2 cassettes might cause problems. The use of another PCR-buffer (Figure 3C) circumvents this problem.
For transformation of S288c- or W303-derived strains, usually 5 µl of a PCR were used. For some other strain backgrounds (such as SK-1), a 10-fold higher amount of DNA was used. For this purpose, the PCR product was ethanol-precipitated and dissolved in water (1/10 of the original volume).
Yeast strains and growth conditions
YPD and synthetic drop-out media were prepared as described (Sherman, 1991). For antibiotic selection markers, the following concentrations of antibiotics were added to standard YPD-plates (www.duke.edu/web/microlabs/mccusker/; Goldstein and McCusker, 1999): kanMX4, geneticin (G418, GibcoBRL), 200 mg/l; hphNT1, hygromycin B (Cayla, Toulouse, France; www.cayla.com), 300 mg/l; and natNT2, nourseothricin (ClonNAT, Werner BioAgents, Jena-Cospeda, Germany; www.webioage.com), 100 mg/l.
The antibiotics were added after autoclaving and cooling of the medium to approximately 60 °C. In the case of ClonNAT, a sterile filtered stock-solution was prepared prior to addition to the medium, while for geneticin and hygromycin B, the powder and the solution provided by the manufacturer were used directly.
Yeast transformation and testing
Yeast transformation using frozen competent cells was based on the LiOAc method (Schiestl and Gietz, 1989), however with several modifications. A detailed description of the method is given in Knop et al. (1999).
For klTRP1 or HIS3MX6 selection, after transformation cells were resuspended in 200 µl sterile PBS and plated directly onto plates containing synthetic medium lacking the respective amino acid (SC-HIS, SC-TRP; Sherman, 1991).
For kanMX4, hphNT1, natNT2-selection: after transformation, cells were resuspended in 3 ml of YPAD medium and incubated on a shaker for at least 5–6 h at 30 °C, than sedimented and plated onto the selection plates.
Selection for positive transformants on plates containing antibiotics often requires replica plating of the plate after 2 days at 30 °C, because of the high background of transiently transformed cells, which makes it difficult to recognize the correct integrants (Knop et al., 1999; Wach et al., 1997). The success of the integration was tested by colony PCR using a quick chromosomal DNA isolation procedure (Finley and Brent, 1995), immunoblotting or by immunofluorescence, as described previously (Knop et al., 1999). For immunoblotting, protein extraction was done using the NaOH/βME/TCA-protocol (Knop et al., 1999).
For the detection of epitope-tagged proteins, tag-specific antibodies were used: HA-tag, mouse monoclonal 12CA5 (Roche Boehringer-Mannheim), 16B12 (Babco); Myc-tag, mouse monoclonal 9E10 (Boehringer-Ingelheim); Protein A/TAP-tag, rabbit PAP (DAKO); Don1p, affinity purified rabbit anti-Don1p (Rabitsch et al., 2001); GFP, affinity-purified sheep anti-GFP. For ECL detection (Amersham), goat anti-mouse, -rabbit or -sheep secondary antibodies coupled to horseradish peroxidase (Jackson Immuno Research Laboratories) were used.
The full collection of plasmids and the sequence files will be made available for non-commercial recipients through EUROSCARF (http://www.uni-frankfurt.de/fb15/mikro/euroscarf/index.html). The plasmids have been prepared and tested carefully; however, we cannot guarantee that no error has been made. In case of problems, please do not contact any of the authors unless you are absolutely sure that the problem is associated with the plasmid (use positive controls!).
Two new selection markers: hphNT1 and natNT2
Recently, Goldstein and McCusker (1999) introduced three new dominant drug resistance cassettes that can be used in the yeast S. cerevisiae. The cassettes were constructed in analogy to the pFA6–kanMX4 marker (Goldstein and McCusker, 1999; Wach et al., 1994), thus allowing the use of the established S1/S2-primer annealing sites (Wach et al., 1994; Knop et al., 1999) for amplification. The hphMX4 and natMX4 (Goldstein and McCusker, 1999) markers confer resistance to hygromycin B or clonNat (nourseothricin), respectively, and were cloned in-between the promoter and terminator of the kanMX4 cassette (Wach et al., 1994). The homologous sequences flanking the different marker genes, however, lead to recombination between the markers, if the two markers are used simultaneously in the same yeast strain. To circumvent this problem, we exchanged the terminator of the hphMX4 cassette and replaced it with the terminator of the CYC1 gene. Similarly, we replaced the natMX4 terminator with the ADH1 terminator. The new cassettes were termed hphNT1 and natNT2, respectively (NT = new terminator; Table 1). As demonstrated in a control experiment (not shown), kanMX4, natNT2 and hphNT1 completely failed to recombine with each other.
C-terminal tagging: fluorescent proteins
The availability of a variety of fluorescent proteins, such as yeGFP (Cormack et al., 1997), EGFP, EBFP, ECFP, EYFP (http://www.clontech.com/gfp/excitation.shtml), DsRed (Matz et al., 1999), hcRED (Gurskaya et al., 2001) and RedStar, a much brighter version of DsRed (Knop et al., 2002), consequently led to the construction of new cassettes. The coding regions of the six fluorophores were cloned into tagging cassettes preceded by a spacer sequence that codes for the peptide ‘SGAGAGAGAGAIL’. This spacer peptide can facilitate the correct folding of the fluorescent proteins when coupled to the protein of interest (Miller and Lindow, 1997). Additionally, we provide a cassette containing the red fluorescent protein eqFP611 (Wiedenmann et al., 2002).
The properties of some of the GFP derivatives are summarized in a review article (Tsien, 1998; for spectral properties, see also Table 3). All of them have been successfully used for applications in baker's yeast, such as in vivo double labelling and live cell imaging. The suitability of each of the individual fluorescent proteins for a specific experiment, however, has to be tested each time.
The red fluorescent protein DsRed has been limited to special application in yeast (Pereira et al., 2001), since the formation of the red chromophore (Baird et al., 2000) is not fast enough (T1/2 ∼ 24 h) to allow the detection of de novo synthesized proteins in logarithmically growing cells. This has been partially solved by the construction of a much brighter variant, called RedStar (Knop et al., 2002), or by a faster-maturing but less bright variant named T4-DsRed (Bevis and Glick, 2002). We constructed a combination of the T4-DsRed and the RedStar mutant, which leads to a bright, fast-maturing red fluorescent protein, RedStar2. We provide for several of these DsRed variants cassettes (Table 1), most of which contain yeast codon optimized constructs. The last drawback of DsRed-variants, their strong tetramerization (Baird et al., 2000), has only recently been solved (Campbell et al., 2002), but this monomeric DsRed variant seems to be not yet bright enough for general applications in yeast (unpublished observation). However, the red fluorescent protein eqFP611 (Wiedenmann et al., 2002) largely circumvents this problem.
Double labelling using different fluorescent proteins
For double-fluorescent labelling, different fluorescent proteins can be combined: GFP together with DsRed, GFP and BFP, GFP and CFP, and YFP in conjunction with CFP. The combination of YFP and CFP is frequently used. The tagged proteins can be distinguished with appropriate filters. However, both, CFP and YFP bleach faster then GFP. The CFP signals often appeared weakly fluorescent when observed by eye; however, imaging with a CCD-camera gave nice and strong signals (Figure 4).
C-terminal tagging: HA, MYC and TAP tag
HA and MYC-tags are used for the detection of the tagged proteins by immunoblotting and immunofluorescence microscopy. A combination of two tagged proteins (HA and MYC, respectively) in one strain is widely used to detect protein–protein interaction by co-immunoprecipitation. Furthermore, it became obvious that proteins with low expression levels can be detected when several repeats of the HA or Myc tag (6HA or 9Myc) were fused to the protein. On the other hand, too many tags may interfere with the functionality of the fusion protein. For native protein purification, it has been shown that single HA-tagged proteins can be eluted from anti-HA beads using the HA peptide (YPYDVPDYA), while this was not possible when multiple tags were used. Because of these considerations, we constructed a variety of PCR modules using single, triple and hexa- or nona-tags in combination with a variety of selection markers (Table 3), thus enabling the flexible construction of strains carrying different tags at the same time.
The use of Protein A as an affinity tag has shown to be a powerful tool for the purification of proteins from yeast lysates, especially in combination with a calmodulin-binding peptide (CBP) and a TEV site-specific protease cleavage site. This combination of features, called the TAP tag (Rigaut et al., 1999), has been shown to be very useful for native protein complex purification (Gavin et al., 2002).
An example for the application of the TAP tag PCR module (pYM13) is shown in Figure 5.
Recently, other tags with specific properties became fashionable. The FlAsH tag consists of a small peptide, containing four cystein residues (DCCPGC-CA), that is recognized by specific di-arsenic compounds, which, upon binding, become fluorescent (Adams et al., 2001). We have tested the FlAsH tag and found that it worked also in yeast; however the maximally obtainable level of fluorescence, when compared with the analogous GFP fusion, was less than 5%, thus limiting the usefulness of this tag. Similarly, we also constructed a cassette containing the photo-activatable GFP (PA-GFP; Patterson and Lippincott-Schwartz, 2002). Proteins carrying this tag emitted, when maximally activated, less than 10% of the fluorescence compared to GFP-tagged versions. This limits the usefulness of this tag in yeast.
Promoter replacement and N-terminal tagging
The introduction of a heterologous promoter upstream of the START codon of a gene is a way to control and to modulate gene expression. At the same time, it allows the introduction of a N-terminal epitope tag to the gene.
We constructed a set of cassettes with nine different replacement promoters. Eight of these promoters were well characterized from previous applications in centromeric or 2 µ plasmids (Mumberg et al., 1994, 1995). The replacement of an internal promoter with the constitutive ADH, CYC1, GPD or TEF promoters can be used to modulate the expression of a gene in a permanent manner. For inducible expression, the GAL1 promoter and two truncated (and weaker) derivatives of this promoter, termed GALL and GALS (Mumberg et al., 1994), as well as the MET25 promoter, are provided. All the promoters were cloned into cassettes with kanMX4 and natNT2 selection markers. Additionally, all natNT2 promoter-substitution cassettes were combined with a N-terminal 3HA and yeGFP (Cormack et al., 1997) tag (Table 1). We observed different expression rates of the gene DON1 when controlled by the eight different promoters. The inducible promoters are not always completely repressed in the non-induced state. In the case of the relatively strong MET25 and the GAL1 promoters, a weak expression was observed in the repressed state of the promoter (glucose complete medium; Figure 6). In contrast, the two weaker versions of the GAL-promoter, GALL and GALS, were completely repressed (Figure 6).
Furthermore, we constructed five cassettes containing the CUP1-1 promoter (Table 1). This strong promoter can be induced with CuSO4. We used this system successfully for the regulated induction of gene expression during various phases of the meiotic cell cycle (unpublished data). An example of the expression of Ssp1p under control of CUP1-1 is given in Figure 7.
In the present paper, we describe 37 new cassettes for the C-terminal epitope tagging of yeast proteins, developed by combining existing tags with new marker genes and cloning new tags, namely a variety of different fluorescent proteins of all available colours, and the TAP-tag. Furthermore, a series of 37 N-terminal cassettes has been developed that allow, besides the replacement of the promoter of the target gene, the introduction of N-terminal tags. For one single gene, all these cassettes can be amplified with four unique primers (Figures 1 and 2). The versatility of the primers is a strong advantage not only regarding to the cost of the method. Also, once all four primers have been successfully tested, any concerns about the quality of the primers can be omitted, which can turn out in some cases to be quite important (see below). The cloning strategies for most of the cassettes were based on common restriction sites, which facilitate the construction of further cassettes, if necessary (Table 1; further details available upon request).
PCR amplification and primers
Since the PCR amplification of the cassettes has caused problems in different laboratories, we describe a PCR-protocol suitable for the amplification of almost all of the cassettes. This protocol works well (in several laboratories), and fulfils three major criteria: reliability, fidelity and high yield. It requires, however, a reliable PCR machine that allows time increment programming. For the amplification of natNT2-based cassettes, this protocol needs to be slightly modified due to the high GC content of the coding sequence of this marker gene (see Materials and methods; Figure 3). Another reason why sometimes the PCR does not work is the poor primer quality. We found, that for some suppliers, up to 20% of the primers do not work (e.g. 40% of the PCRs performed), while for other suppliers, less then 5% are non-functional (less than 10% of PCRs performed) with respect to amplification of modules. Testing the primers in combination with established primers can help to nail down the faulty primer (companies normally will provide a free replacement primer).
New selection markers
The use of the hphNT1 and natNT2 cassettes is as robust as the kanMX4 cassettes. Cells selected on antibiotic media tend to form a lawn, due to the growth of transiently transformed cells, which might hinder the identification of positive clones. In such a case the cells were replica-plated after 2 days of growth onto a fresh plate of the same medium. On the new plate, only positive clones grow. Using kanMX4 and HIS3MX6 together in one strain led to recombination events within the marker genes. After the transformation of the second cassette, positive clones must be selected on both, G418 and SC-His plates. The klTRP1 cassette seems to promote a somewhat less-than-wild-type growth rate when used to complement the trp1 mutation; therefore, it is recommended to wait 2 more days in case no colonies appear 2–3 days after transformation. Usually, transformants were confirmed using colony PCR in combination with either immunoblotting using anti-HA, anti-Myc, anti-GFP or PAP (for detection of protein A tags) antibodies or fluorescence microscopy (to visualize fusions with fluorescent proteins) or indirect immunofluorescence microscopy (HA or Myc fusions).
New fluorescent markers
We observed that yeGFP (Cormack et al., 1997) and EGFP (Clontech) do not show observable differences in brightness, although they do contain different mutations compared to the wild-type GFP.
We have also provided a number of different cassettes containing DsRed and mutagenized versions of DsRed. Due to the properties of the DsRed protein, its application is somewhat limited compared to GFP. This is mainly due to its strong tetramerization (Baird et al., 2000), which can interfere with protein function (Knop et al., 2002). Table 4 summarizes some of the properties of the different red fluorescent proteins that are contained in our cassettes.
Table 4. Properties of the red fluorescent protein
Brightness relative to DsRed
The yeast codon optimized sequence of DsRed, RedStar and RedStar2, contain an additional codon at position 2.
A new feature of the presented set of cassettes are the 37 new promoter substitution and N-terminal tagging modules. Apart from the CUP1-1 promoter, which was specifically cloned for N-terminal tagging of proteins that are involved in meiosis, all other promoters were taken from existing yeast plasmids, therefore, their expression levels have already been studied in detail and the promoters can be used according to these data (Mumberg et al., 1994, 1995). The promoter substitution can be applied for the determination of expression level-related phenomena, or simply to deplete a gene product. It was noted that, while the GAL1 and the MET25 promoters were slightly leaky under repressive conditions, the less active GALL and GALS were tightly repressed in glucose medium (unpublished data and Figure 6). The use of the GALS promoter might thus be a better tool than the until now frequently used GAL1 promoter, first because of the reduced leakiness, but also for the lower expression rate in the induced state.
Taken together, the new range of PCR-cassettes allows the use of more selection markers, the combination of more tags in one single strain and the application of fluorescence double labelling with CFP and YFP, but also with GFP and RedStar2, while dsRed and RedStar can be used as fluorescent timers (see above and Table 4). N-terminal tags and promoter substitutions allow to interfere with transcriptional regulation and to conditionally deplete gene products, while the availability of N-terminal tags provides the possibility to label proteins that cannot be tagged at the C-terminus. The need of only four different primers for the use of all cassettes described here and in Knop et al. (1999) makes the tagging cheap, reliable and flexible. However, the ease by which new strains can be constructed by this method should, of course, never prevent us from keeping one key question in mind: how does this manipulation affect the function of the gene?
The work of E. Schiebel was supported by the Cancer Research Campaign UK and of M. Knop by the Max-Planck-Institute of Biochemistry, Department of Molecular Cell Biology, Munich, Germany, and the EMBL, Heidelberg, Germany. C. Janke was supported by an EMBO long-term fellowship (ALTF 387-2001). E. Schwob was funded by CNRS and the Association pour la Recherche sur le Cancer (ARC), France. M. M. Magiera is supported by a PhD fellowship from the French Ministry of Research.