Applied and Environmental Microbiology, May 2002, p. 2095-2100, Vol. 68, No. 5
0099-2240/02/$04.00+0 DOI: 10.1128/AEM.68.5.2095-2100.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.
| MINIREVIEW |
Kluyver Laboratory of Biotechnology, Delft University of Technology, 2628 BC Delft, The Netherlands
|
|
|---|
Several marker genes used in yeast genetics confer resistance against antibiotics or other toxic compounds (42). Selection for strains that carry such marker genes requires the addition of these toxic compounds to the growth media. In addition to their toxicity, the price of many of these compounds precludes their use in large-scale processes (18). Moreover, even in resistant strains, the presence of antibiotics may affect cellular function. An alternative is the use of marker genes that complement specific nutritional requirements. Some of the most commonly applied marker genes are wild-type alleles of yeast genes that encode key enzymes in the metabolic pathways towards essential monomers used in biosynthesis. An example is the URA3 gene, which encodes orotidine-5'-phosphate decarboxylase, an essential enzyme in pyrimidine biosynthesis in Saccharomyces cerevisiae (3). Similarly, the HIS3, LEU2, TRP1, and MET15 marker genes encode essential enzymes for de novo synthesis of the amino acids L-histidine, L-leucine, L-tryptophan, and L-methionine, respectively (4, 8).
Use of these genes as markers is restricted to host strains that are auxotrophic for the nutrient in question due to the absence of a functional chromosomal copy of the marker gene. Unless transformed to prototrophy with a functional allele of the marker gene, auxotrophic yeast strains can be propagated only in media that contain the appropriate growth factor(s). This nutritional complementation may be achieved either by including the growth factors in defined synthetic media or by using complex medium components (e.g., yeast extract and peptone) that are rich in the relevant growth factors. The ease with which auxotrophic yeast strains and the corresponding auxotrophy-complementing genes can be manipulated and the low cost of the chemicals involved have contributed to the construction of myriads of laboratory strains of S. cerevisiae with various combinations of auxotrophic markers.
Upon the completion of the genome sequence of S. cerevisiae, scientists set out to assign biochemical and physiological functions to thousands of newly discovered open reading frames (16). Even a basal understanding of cellular function cannot be based solely on qualitative phenotypic analysis but also requires accurate quantitative approaches (31). These approaches require elimination of experimental "noise" resulting from strain choice or unintended side effects of genetic modifications. In my own research and while serving as an Editorial Board member of this journal, I have frequently struggled with complications arising from the incorrect use of auxotrophic strains and auxotrophy-complementing marker genes in physiological studies on S. cerevisiae and other yeasts. The objective of this minireview is to consider a number of aspects that are relevant for the application of auxotrophic yeast strains in quantitative physiology, functional genome analysis, and industrial applications. Although the focus will be on S. cerevisiae, the general considerations should be applicable to other yeasts and, indeed, to other microorganisms.
Another important application of auxotrophic markers is the introduction of knock-out mutations. Targeted inactivation of yeast genes is now performed almost exclusively via the one-step gene deletion method (35). In this approach, a marker gene is equipped with short 5' and 3' sequences that are identical to the sequences flanking the chromosomal sequences that need to be deleted. If auxotrophy complementation is used as the marker system for one-step gene inactivation, the resulting deletion mutant will become prototrophic for the nutrient involved. This change has two consequences. First, the null mutant will have a set of auxotrophic requirements different from that of the original host (see below). Second, the marker gene used for gene inactivation will no longer be available for further genetic modifications. This restriction may be cumbersome if multiple deletions need to be introduced, e.g., during the deletion of large gene families or as part of the construction of null mutants in aneuploid industrial strains. This problem can be bypassed by marker recovery strategies based on site-specific recombination, mediated, for example, by the cre-loxP recombination system of bacteriophage P1 (17, 36). These strategies involve the introduction of flanking repeat sequences at the 5' and 3' ends of the marker gene, which are recognized by a specific recombinase. By introduction of a helper plasmid encoding this recombinase, the integrated copy of the auxotrophy-complementing marker gene is excised, rendering the strain accessible to new rounds of modification with the same marker gene.
When an auxotrophic strain is complemented by introduction of a single copy of the wild-type allele of the affected gene, as is the case during integrative recombination, no detrimental effects are anticipated, with the exception of effects caused by (extra)chromosomal location. However, after one-step gene inactivation with an auxotrophy-complementing marker gene, the transformed strain has a set of auxotrophies different from that of the original host strain since one of the markers present in the original strain is eliminated by the transformation event. To compensate for this deletion, the auxotrophic host strain must be grown by including the appropriate auxotrophic growth factor in the medium. In a physiological sense, nutritional supplementation may differ strongly from genetic complementation. For example, nutritional and genetic complementation of an S. cerevisiae leu2 null mutant in glucose-grown shake-flask cultures yielded different specific growth rates (Fig. 1). A similar phenomenon was observed in a comparison of nutritional and genetic complementation of a ura3 mutant of a commercial baker's yeast strain (T. Petit, personal communication).
![]() View larger version (15K): [in a new window] |
FIG. 1. Genetic complementation versus nutritional supplementation. Shown are growth curves of the prototrophic S. cerevisiae strain CEN.PK113-7D ( ), the congenic leu2-3 null mutant CEN.PK113-16B ( ), and CEN.PK113-16B transformed with the multicopy vector YEPlac181 (15) that carries the wild-type LEU2 allele ( ). All three strains were grown at 30°C in shake flasks on a defined synthetic medium (47) with 20 g of glucose liter-1 as the carbon source. L-Leucine was added to the nontransformed CEN.PK113-16B culture at a concentration of 500 mg liter-1 (Table 1). Cultures were inoculated from exponentially growing cultures on the same medium. Exponential growth started without a detectable lag phase. Specific growth rates calculated from the exponential growth phase preceding the diauxic shift were 0.40 h-1, 0.34 h-1, and 0.39 h-1, respectively. Data are presented as the average of two independent shake-flask cultures for each strain and, for the points preceding the diauxic shift, differed by less than 4% between the replicate cultures.
|
An elegant demonstration of the impact of nonmatching auxotrophies on growth kinetics is provided by a set of experiments on S. cerevisiae strains in which the HO gene (which has no known cellular role other than mating type switching) was inactivated by replacement with either the antibiotic resistance marker gene kanMX or the HIS3 gene (2). In competition experiments, performed in chemostat cultures grown under several nutrient limitation regimens, the ho
::kanMX null mutation did not confer a significant competitive (dis)advantage relative to a congenic strain carrying the wild-type HO allele. However, in histidine-supplemented chemostat cultures, the fitness of congenic HO his3 and ho
::HIS3 strains (which had nonmatching auxotrophies) differed strongly. Under glucose-limited growth conditions, the ho
::HIS3 strain rapidly outcompeted the histidine-auxotrophic HO his3 strain. Conversely, under ammonia-limited growth conditions, the histidine prototroph disappeared from the mixed culture within 30 generations (2). The dependency of this marker gene effect on growth conditions may be related to nitrogen catabolite repression of amino acid uptake and catabolism in S. cerevisiae (19). Without the kanMX control, the experiments of Baganz et al. (2) might have led to the wrong interpretation that deletion of the HO gene affects growth kinetics in chemostat cultures.
Nonmatching auxotrophies also may result in different stoichiometries of biomass and product formation. For example, based on a comparison of strains with nonmatching auxotrophies, it was initially reported that deletion of the ATH1 gene (encoding vacuolar acid trehalase) resulted in increased biomass yields of S. cerevisiae on glucose (22). This was subsequently shown to be due to the introduction of a wild-type allele of URA3 in the ath1 null mutant (7). Effects on growth stoichiometry may reflect energy costs of uptake and incorporation of auxotrophically required nutrients as well as carbon and nitrogen supplementation effects caused by their inclusion in growth media. Especially under anaerobic conditions, inclusion of amino acids in growth media may significantly affect cellular redox balance and product formation (1). This is mainly due to changes in the net production of NADH in amino acid biosynthesis. These changes are reflected in the production of glycerol, which, in anaerobic S. cerevisiae cultures, is the main sink for NADH redox equivalents (1).
Many standard synthetic medium recipes for cultivation of S. cerevisiae are based on an initial glucose concentration of 20 g liter-1. If glucose repression is alleviated by aerobic, glucose-limited cultivation (43) or via genetic modification (13), biomass yields of up to 0.5 g of biomass g of glucose-1 can be achieved. In aerobic batch cultures of wild-type S. cerevisiae strains, the typical diauxic growth on glucose and ethanol (formed during the first, respirofermentative, phase of growth) leads to overall biomass yields of ca. 0.4 g g-1 (25). Based on these biomass yields and published data on the biomass composition of S. cerevisiae (32), it is straightforward to estimate the minimum requirements for auxotrophically required compounds. As an example, the S. cerevisiae leucine-auxotrophic leu2 null mutants' requirement for leucine will be discussed.
L-Leucine accounts for ca. 10% of the mass of yeast protein, and the protein content of dry yeast biomass is ca. 40% (32). Consequently, a leucine auxotroph will require ca. 40 mg of leucine for the formation of 1 g (dry weight) of yeast biomass. In a medium containing 20 g of glucose per liter, ca. 10 g of biomass can be formed under optimal, aerobic, cultivation conditions (43). To avoid growth limitation by leucine during cultivation of a leucine auxotroph, such a medium should contain at least 400 mg of leucine per liter. This concentration is higher than an experimentally determined value of 240 mg liter-1 (5). However, the possibility that growth in the synthetic medium used in this study (5) may have been limited by a nutrient other than glucose or leucine cannot be excluded.
The protein content of yeast biomass is dependent on environmental conditions (46). Furthermore, in the example given above, part of the leucine added to the media may be deaminated and decarboxylated via the enzymes of the Ehrlich pathway (12). It therefore seems prudent to maintain an excess factor in the design of media. In Table 1, minimum requirements for nutritional complementation of some commonly applied auxotrophies have been listed and recommendations have been given for media containing 20 g of glucose liter-1, based on an excess factor of ca. 25%. Since, under anaerobic conditions, the biomass yield on glucose is ca. fivefold lower than the maximum biomass yield in aerobic cultures (43), these recommended concentrations can be divided by a factor of five for anaerobic cultivation on glucose. In the literature, and even in handbooks for yeast research (34, 37, 40), many instances are encountered in which required auxotrophic nutrients are added at concentrations of 10 to 40 mg liter-1. As illustrated in Table 1, this will often result in starvation foror growth limitation bythe auxotrophically required nutrient. The ensuing metabolic, regulatory, and morphological changes (6, 11) may readily lead to misinterpretation of experimental data.
|
View this table: [in a new window] |
TABLE 1. Estimated contents of commonly applied auxotrophy-complementing compounds in S. cerevisiae biomass and recommended concentrations for aerobic cultivation of S. cerevisiae on defined media containing 20 g of glucose liter-1
|
An elegant system has been developed to allow for the use of the URA3 marker gene even in complex, uracil-containing media. This system is based on inactivation of the FUR1 gene (26). In S. cerevisiae, the essential biosynthetic intermediate UMP (uridine 5'-monophosphate) can be synthesized either by de novo biosynthesis via a pathway involving the URA3 gene product (3) or by direct conversion of uracil via the pyrimidine salvage pathway (24). FUR1 encodes uracil phosphoribosyltransferase, a key enzyme in the latter pathway (24). Strains that contain disfunctional chromosomal copies of both FUR1 and URA3 allow for the use of a plasmid-borne URA3 gene as a selectable marker, even in uracil-containing media. This concept has been successfully applied to the stable expression of heterologous proteins in complex media (26, 45). However, prototrophic fur1 strains have a reduced specific growth rate on glucose; this may be a problem in applications that require rapid growth.
Although rarely investigated, cross-feeding may be a relevant phenomenon during the cultivation of auxotrophic yeast strains. In some cases, prototrophic strains appear to release either the auxotrophically required nutrient itself or related biosynthetic intermediates that are located downstream from the metabolic lesion in the auxotrophic strain. Even when synthetic and apparently selective media are used, this may allow for the establishment of mixed cultures, in which a subpopulation of auxotrophic strains remains metabolically active. This is exemplified by an experiment in which a leu2 strain of S. cerevisiae, transformed with a LEU2-bearing expression vector, was grown in carbon-limited chemostat cultures on a synthetic medium without leucine. After prolonged cultivation, up to 45% of the population consisted of leucine-auxotrophic cells and low concentrations of leucine were detectable in culture supernatants (27). Therefore, even under apparently selective conditions, it is advisable to check plasmid stability, e.g., by performing plate assays.
Cross-feeding of auxotrophic strains is relevant not only when the auxotrophies are employed as selectable markers. It has been proposed that competition experiments with mixed cultures, consisting of large numbers of oligonucleotide-tagged deletion mutants, may be a powerful tool for rapid and quantitative functional analysis of yeast genes (2, 38). However, it is conceivable that in such mixed cultures, in which each deletion mutant makes up only a small fraction of the total population, cross-feeding of metabolic intermediates may hide relevant phenotypes.
An example of phenotype masking by an auxotrophic strain background is provided by a study on a pda1 null mutant (48). When the PDA1 gene, encoding an essential subunit of the yeast pyruvate-dehydrogenase complex, was deleted in the prototrophic strain T2-3D, this resulted in a substantial reduction of the maximum specific growth rate on glucose. This effect was entirely absent in the M5 strain background that carries a defective leu2 allele. This strain-dependent phenotype was shown to be due to a partial leucine requirement induced by the pda1 mutation. This interesting phenotype, which probably results from an altered intramitochondrial acetyl-coenzyme A/coenzyme A ratio (48), would have gone entirely unnoticed if the studies had been confined to a leu2 strain backgroundwhich by definition has to be provided with leucine. Similarly, the requirement of S. cerevisiae zwf1 null mutants for methionine (ZWF1 encodes glucose-6-phosphate dehydrogenase [41]) might have gone unnoticed if the strain background have been auxotrophic for methionine.
As the data intensity and costs of experimental biology increase, it becomes ever more important to eliminate factors that may interfere with data interpretation. As discussed in this minireview, use of auxotrophic strains may tremendously complicate the interpretation of experimental data in functional genomics, physiology, and metabolic engineering. Since good alternative marker genes and marker-rescue systems are available, it is advisable to avoid the use of auxotrophic strains in such researchunless the auxotrophy itself is the subject of study. When auxotrophic strains are used for reasons of practicality or economy, comparison of strains with nonmatching auxotrophies should be avoided at all times and auxotrophically required nutrients should be provided in amounts that prevent them from becoming growth limiting (Table 1). Even when prototrophic strains are used, great care should be exercised in the design of synthetic growth media in order to avoid "hidden" nutrient limitations. I stress that, since nutritional requirements of microorganisms vary as a function of growth conditions and genetic background, it is essential to experimentally verify which nutrient limits biomass formation even when prototrophic strains are used.
It becomes ever clearer that the physiological properties of different S. cerevisiae strains can differ greatly (23, 44). At a time when public domain databases have become vital tools for studies on popular model organisms such as S. cerevisiae, it is important to minimize the confusion that may arise from comparison of quantitative data from different genetic backgrounds. It is therefore worthwhile to strive for a limited number of well-defined and experimentally accessible reference strains. Based on a comparison of four candidate strains, a consortium of European yeast research groups has recently proposed prototrophic strains of the CEN.PK family of S. cerevisiae as a useful platform for integrated genetic and physiological studies on this yeast (43). Similarly, the prototrophic strain CBS2359 has been proposed as a reference strain for the yeast Kluyveromyces lactis (21, 49). Clearly, using standard strains to compile a reference set of quantitative data should not lead researchers to extrapolate "strain behavior" to "species behavior" or to neglect biodiversity within and among species as a source of scientific inspiration and information.
|
|
|---|
-amylase cDNA in an industrial baker's yeast strain. Biotechnol. Prog. 15:459-466.[CrossRef][Medline]
subunit of the pyruvate dehydrogenase complex. Eur. J. Biochem. 209:697-705.[Medline]
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»