The TRP1 marker has been commonly used for gene disruption experiments and subsequent phenotypic analysis. However, introduction of the TRP1 gene into a trp1 strain markedly affects growth under many conditions used for phenotypic profiling. Therefore, its use in the past should be revisited and utilization of this marker should be avoided in future analyses.
The budding yeast Saccharomyces cerevisiae is one of the best genetically tractable eukaryotic systems because it allows powerful and simple experimental approaches based on both classical and molecular genetics. Genetic modifications such as the targeted inactivation of genes or their controlled expression from centromeric or episomal vectors require the use of selectable marker genes for efficient detection and selection of transformed cells (Botstein & Davis, 1982). The use of auxotrophic markers, mainly URA3, LEU2, TRP1, HIS3, MET15 or LYS2, has been virtually unavoidable for many years and, even after the relatively recent appearance of dominant drug-based resistance markers, such as geneticin (Wach et al., 1994) or nourseothricin (Goldstein & McCusker, 1999), genetic manipulations in S. cerevisiae involving auxotrophic markers are very common. Frequently, auxotrophic markers are used to carry out gene disruption experiments, followed by the characterization of the phenotypes of the resulting strains. This is always made on the basis that the introduction of the selectable marker does not affect the traits under investigation.
During the course of a phenotypic analysis of several S. cerevisiae strains defective in two alcohol dehydrogenases, an increased tolerance was observed to different metals in an adh6::TRP1 disruptant that was confirmed in two additional adh6::TRP1 strains on different genetic backgrounds. However, different tests allowed to recognize that the observed phenotypes were not caused by the deletion of ADH6 but, instead, originated from the use of a TRP1-based disruption cassette, which converted a trp− strain into trp+. Interestingly, a classic paper describing the construction of the BY strain series alluded only to the well-known cold-sensitive phenotype of trp− strains (Brachmann et al., 1998). A thorough literature search yielded several reports pointing out phenotypic traits attributable to the TRP1 deletion. However, even after recent systematic and refined fitness studies (Giaever et al., 2002), the available information was rather fragmentary, scattered and difficult to identify. Therefore, it was decided to specifically evaluate up to which extent the use of this marker in gene disruption experiments could produce artifactual results.
To this end, a first set of experiments were performed to evaluate the effect of loss of TRP genes using the BY4741 strain (a prototroph for tryptophan) and their isogenic trp1 to trp5 single KanMX deletion mutants (Winzeler et al., 1999), which lack individual genes required for the synthesis of tryptophan from chorismate. These strains were grown on YPD and different dilutions of the cultures were spotted on YPD plates and grown under different conditions. As shown in Fig. 1a, trp1 mutants display sensitivity to high pH. This phenotype is shared with similar potency by trp2 and trp3 strains and with less intensity by trp4 and trp5 cells. Existing reports disagree in this regard, as the alkaline phenotype of trp1 has been observed in some cases (Giaever et al., 2002) but not in others (Serrano et al., 2004; Sambade et al., 2005). In a previous work (Sambade et al., 2005), it was described that the entire collection of tryptophan biosynthetic mutants failed to grow on pH 7.5 plus 60 mM CaCl2. It is shown here that trp mutants, with the remarkable exception of trp3, are slightly sensitive to this cation even at standard pH. Because the traits described above are also found in mutants with impaired vacuolar function, trp1-5 cells were stained with the lipophylic fluorescent dye FM4-64 in order to visualize vacuolar morphology. However, deletion of TRP1-5 does not result in aberrant morphology of vacuoles.
It was observed that trp1-5 mutants are not particularly sensitive to cell wall-damaging agents such as calcofluor white (CFW) or Congo red, whereas they display a marked growth defect in the presence of sodiumdodecyl sulfate (SDS) at 10 μg mL−1 (Fig. 1b), that is, a concentration threefold lower than that described in a recent report (Liu et al., 2004). However, the reported growth defect at 1.5 M sorbitol could not be confirmed on YPD plates (Giaever et al., 2002). Lack of TRP1-5 confers significant sensitivity to rapamycin, an inhibitor of the TOR pathway. This phenotype was recently described for the trp1, 3, 4 and 5 mutants (Liu et al., 2004; Xie et al., 2005). As shown in Fig. 1a, the trp3 mutation confers a marked phenotype of sensitivity to caffeine, whereas lack of TRP2 causes a lesser effect. This is in contrast with a previous report, in which no effect of caffeine was observed for trp1, 3, 4 or 5 mutants (Liu et al., 2004). The differential effect of rapamycin and caffeine on trp mutants is striking, because the Tor1 kinase has been recently proposed as a target of caffeine (Kuranda et al., 2006). A remarkable finding is that cells lacking trp1 are somewhat sensitive to metal ions such as iron, copper or zinc. Similarly, trp1 cells are rather sensitive to spermine and Hygromycin B, but only slightly sensitive to TMA. As far as the authors know, these characteristics have not been reported in the past. In contrast, the growth of trp mutants does not differ from wild-type cells under other conditions tested, such as nonfermentable carbon sources (ethanol and glycerol), lithium (0–200 mM LiCl), sodium (0–1.2 M NaCl) or cesium cations (0–100 mM CsCl). A comprehensive catalogue of phenotypes for trp1 cells, tested in the authors' laboratory or/and collected from the literature, is shown in supplementary Table S1.
A second set of experiments were carried out to evaluate the effects of the introduction of the TRP1 gene, which is widely used as a selection marker in plasmid shuttle vectors. The presence of a centromeric YCplac22 (TRP1 marker) plasmid in a trp1 strain (JA100) made the cells more tolerant to alkaline pH, caffeine, rapamycin and diverse metals such as zinc, copper and iron (Fig. 1c). These are phenotypes opposite to those observed upon disruption of TRP1 (see Fig. 1a). It is worth noting that in contrast to that observed in the BY4741 background, in this strain the presence or absence of TRP1 does similarly affect tolerance to both rapamycin and caffeine. In agreement with the previous results, the tolerance to sorbitol, CFW, Congo Red, LiCl, NaCl or CsCl was unaffected by the presence of TRP1.
In conclusion, the presence or absence of the TRP1 auxotrophic marker can substantially alter the phenotypic traits of a given strain. In contrast, the introduction of LEU2 or URA3 markers did not affect any of the phenotypic traits investigated here (not shown). The ‘TRP1 effect’ probably has little repercussion when the marker is located in a plasmid carrying a gene to be phenotypically tested, because in this case the same strain carrying an empty plasmid is commonly used as a negative control. However, when the TRP1 marker is used in gene disruption experiments, often TRP1 and trp1 strains are compared. This may lead to ascribe a given phenotype to the disruption of the gene under study when, in fact, it is the result of the introduction of the marker. An example of this was hinted at by Bauer (2003) during the characterization of azr1 and pdr12 mutants. As a further example (Fig. 1d), mutation of CKB1, encoding a regulatory subunit of the CK2 protein kinase using a ckb1::HIS3 disruption cassette in the DBY746 background, results in sensitivity to high pH (this was confirmed with a ckb1::KanMX cassette). However, the use of a ckb1::TRP1 cassette largely eliminates the alkaline pH sensitive phenotype caused by lack of Ckb1 (Fig. 1d), most probably due to the increased tolerance caused by the presence of the TRP1 gene (see Fig. 1c). Therefore, the utilization of TRP1 as an auxotrophic marker in phenotypic studies should be avoided in future analyses, and it would be wise to revisit results obtained in the past involving its use.
The authors thank C. Vázquez de Aldana (U. Salamanca) for pointing out the intense sensitivity of trp1 mutants to SDS. The excellent technical assistance of Anna Vilalta and María Jesús Álvarez is acknowledged. This work was supported by grant BFU2005-06388-C4-04-BMC to J.A. and BMC2003-09606 to J.B. (Ministerio de Educación y Ciencia, Spain and Fondo Europeo de Desarrollo Regional). A.G. is a recipient of a fellowship from Ministerio de Educación y Ciencia, Spain.
Table S1. Phenotypes ascribed to the lack of TRP1.
This material is available as part of the online article from: http://www.blackwell-synergy.com/doi/abs/10.1111/j.1567.1364.2007.00315.x (This link will take you to the article abstract).
Please note: Blackwell Publishing is not responsible for the content or functionality of any supplementary materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.