Comparison of Two Alternative Dominant Selectable Markers for Wine Yeast Transformation

Strains.The wine yeast strains used for this work were as follows: EC1118 (Lallemand Inc., Montreal, Canada), a commercial S. cerevisiae second-fermentation strain that was isolated from the Champagne region (France); IFI473, an S. cerevisiae strain from the collection of the Instituto de Fermentaciones Industriales that was isolated in Sant Sadurni d'Anoia (Spain) and previously shown to be suitable as a starter for second fermentation (28); and T_73-4 (35), a ura3Δ derivative of S. cerevisiae T₇₃ (Lallemand), which in turn is a commercial strain that was isolated in Alicante (Spain). Other industrial strains used for this work were IFI228, an S. cerevisiae baker's yeast strain from the collection of the Instituto de Fermentaciones Industriales, and IFI234, IFI237, and IFI239, all of which are brewer's yeast strains from the same collection. The laboratory strain S. cerevisiae BY4741 (MATahis3 leu2 met15 ura3) was also used for transformation experiments. Escherichia coli strain DH5α (supE44 ΔlacU169 [φ80 lacZΔM15] hsdR17 recA1 endA1 gyrA96 thi-1 relA1) (22) was used for the construction and production of the plasmids used in this study.

DNA manipulations and plasmid construction.The ARO4 and FZF1 genes, including promoter and terminator sequences, were amplified from yeast genomic DNA (BY4741) with the primer pairs ARO4U (5′-TGAATCACGTGATCAACAGC-3′)-ARO4D (5′-CATGATCACCTAATTAGCCGT-3′) (14) and FZF1U (5′-CGGTGCTGGTATCATGGCTTTGATC-3′)-FZF1D (5′-ACTCTTGCCTTTTCTTTGTTTC-3′) (31), respectively. Pfu DNA polymerase (Strategene, La Jolla, Calif.) was used for PCR amplifications, and the amplification products were cloned into pGEM-T (Promega Corporation, Southampton, United Kingdom) according to the instructions of the respective suppliers. In both cases, the insert was released by digestion with NcoI and NotI, the extremes were filled in by treatment with Klenow, and then the genes were cloned into a pRS316 (42) or YEp352 (24) vector that had previously been digested with SmaI. The resulting plasmids were named pCA1, for ARO4 cloned in pRS316; pCF1, for FZF1 cloned in pRS316; pEA1, for ARO4 cloned in YEp352; and pEF1, for FZF1 cloned in YEp352. ARO4-OFP- and FZF1-4-containing centromeric or episomic plasmids were constructed by site-directed mutagenesis of pCA1, pCF1, pEA1, and pEF1 by use of a QuickChange in vitro mutagenesis kit (Stratagene). ARO4 mutagenesis was performed with the primer 5′-GCTTGATACCATTTCTCCTAAATACTTGGCTGATTTGG-3′ and its complementary primer 5′-CCAAATCAGCCAAGTATTTAGGAGAAATGGTATCAAGC-3′, and FZF1 mutagenesis was performed with the primer 5′-GCAAAAAATTTATAAGACCGTACCATCTACGAGTTCACAAATGG-3′ and its complementary primer 5′-CCATTTGTGAACTCGTAGATGGTACGGTCTTATAAATTTTTTGC-3′. The resulting plasmids were named pCA2 (ARO4-OFP in pRS316), pCF2 (FZF1-4 in pRS316), pEA2 (ARO4-OFP in YEp352), and pEF2 (FZF1-4 in YEp352). Maps of these four plasmids are shown in Fig. 1. The plasmids were purified from E. coli cells by use of a High Pure plasmid isolation kit (Roche Diagnostics SL, Barcelona, Spain) and from yeast cells by the method described by Robzyk and Kassir (39). Yeast genomic DNAs were extracted by the method of Querol et al. (36). DNA concentrations were calculated by measuring the optical densities at 260 nm. DNA fragments resolved in agarose gels were purified by use of a QIAquick gel extraction kit (Qiagen GmbH, Hilden, Germany). E. coli was transformed by electroporation (40). All other nucleic acid manipulations were performed as described by Sambrook et al. (40).

• Open in new tab
• Download powerpoint

FIG. 1.

Restriction maps of centromeric and episomic plasmids carrying the resistance genes tested in the yeast transformation experiments. E. coli-derived sequences: ori, replication origin of ColE1; amp, beta-lactamase-encoding gene; lacZα, α fragment of the lacZ gene; P_lac, promoter of lacZ. S. cerevisiae-derived sequences: 2μ, replication origin of the natural 2μm plasmid; ARSH4, autonomously replicating sequence; CEN5, centromeric sequence; URA3, URA3 gene; ARO4-OFP, mutant allele of ARO4 conferring resistance to PFP; FZF1-4, mutant allele of FZF1 conferring sulfite resistance.

Transformation experiments.Yeast cells were grown overnight in YPAD medium (1% yeast extract, 2% peptone, 2% dextrose, and 0.01% adenine hemisulfate) at 30°C and 200 rpm. The total number of cells in each transformation tube was adjusted to 10⁸ by a direct microscopic count, and transformation was performed according to the lithium acetate method described by Ito et al. (25), as modified by Agatep et al. (1). Industrial yeast strains were transformed with 20 μg of DNA, while laboratory strains were transformed with 2 μg of DNA. Before the selection of transformants, cells were diluted 10 times in YPD medium (1% yeast extract, 2% peptone, and 2% dextrose) and incubated for 17 h at 30°C and 200 rpm to allow the expression of resistance alleles (31). The cells that were transformed with pCF2 or pEF2 plasmids were selected on YPD + TA plates (2× YPD containing 150 mM l-tartaric acid for buffering at pH 3.5 and 2% agar and were separately autoclaved in equal amounts, mixed and sulfite was added, in the necessary quantity for each strain, from a sterile 100× solution). The cells that were transformed with pCA2 or pEA2 plasmids were selected on SD + Tyr plates (0.67% yeast nitrogen base without amino acids [Difco Laboratories Inc., Detroit, Mich.], 2% dextrose, 1.67% agar for minimal medium, 0.9 g of l-tyrosine and PFP/liter, in an estimated quantity for each strain, and the required supplements, the last two of which were added from sterile 100× solutions). All plates were incubated at 30°C, and colonies were counted after 2 to 5 days. Additionally, for comparisons of the transformation frequencies obtained by using different selection criteria, the transformants of BY4741 and T_73-4 strains were selected after 3 days of incubation at 30°C on SD plates without uridine (0.67% yeast nitrogen base [Difco Laboratories], 2% dextrose, 1.67% agar for minimal medium, and the required supplements). True and false-positive transformants were distinguished by enzyme restriction analyses of plasmid DNAs extracted from a limited number of putative transformants. The actual percentage of false-positive results was deduced from the transformation frequencies discussed below.

Determination of MICs.The MICs of sulfite and PFP were calculated for each strain in the selection media described above. First, the cultures were adjusted to 10⁷ cells/ml, and 2-μl aliquots of these suspensions or serial dilutions of them (1/10, 1/100, and 1/1,000) were spotted onto plates of the cognate selection media with various amounts of sulfite (0, 1, 2, 3, 4, 5, 6, or 7 mM) or PFP (0, 1, 2, 3, 4, or 5 g/liter) and then incubated for 3 to 5 days at 30°C. In a second round of culture, plates containing amounts of inhibitor differing by 0.1 mM (sulfite) or 0.1 g/liter (PFP) from the previous value were used. This test was repeated twice, and identical results were obtained. Finally, these results were confirmed by plating 4 × 10⁷ cells from a YPD overnight culture on selection medium containing the previously calculated MIC, as well as slightly higher and lower concentrations of sulfite (±0.2 mM) or PFP (±0.2 g/liter), to test for the spontaneous appearance of resistant colonies.

MICs.For each of the wine yeast strains, as well as strain BY4741, the MIC of sulfite or PFP was determined, as described in Materials and Methods, in order to decide the amount of each inhibitor to be used for transformation experiments. Sulfite resistance was clearly higher for wine yeast strains than for the laboratory strain, while resistance to PFP was similar for all of the strains, with the exception of T_73-4, which had a resistance about twice that of the other strains (Table 1).

The high MIC of sulfite in the case of wine yeast strains seems to be related to their niche of origin, because baker's (IFI228) and brewer's (IFI234, IF237, and IFI239) yeast strains showed resistance levels that were similar to that of the laboratory strain (0.8 to 1.5 mM) (data not shown).

Transformation of industrial and laboratory yeast strains.In order to obtain comparative results, we cloned both selective markers into the same vector backgrounds, i.e., YEp352 as a multicopy plasmid (24) and pRS316 as a centromeric plasmid (42). Three wine yeast strains and one laboratory strain were transformed, as described in Materials and Methods, with pCA2, pCF2, pEA2, or pEF2. Several transformation experiments were performed for each combination of strain and transforming plasmid. Control experiments included pRS316, YEp352, and no-DNA transformations. The amount of PFP or sulfite used for each strain was the MIC shown in Table 1. For comparison, ura3Δ strains (BY4741 and T_73-4) were selected in parallel for the URA3 marker that was present in all of the constructions. The results of these transformation experiments are summarized in Table 2.

As expected, the highest transformation frequencies were obtained with the laboratory strain BY4741 and the URA3 marker. Transformation frequencies were also high for BY4741 when ARO4-OFP or FZF1-4 was used as a selection marker, with no clear trend towards the episomic or centromeric constructions (frequencies above 10³ were attained for all four constructions tested [Table 2]).

Among wine yeast strains, the highest transformation frequencies were obtained with strain T_73-4, with very similar values for vectors carrying the URA3 or ARO4-OFP marker. Transformation of the prototrophic industrial strains EC1118 and IFI473 with the URA3 marker could not be tested, but the frequencies were high for pCA2 and pEA2 transformation.

The relative transformation frequencies for the two dominant selection markers for T_73-4 were opposite those for the laboratory strain BY4741. IFI473 could not be transformed to sulfite resistance with either pCF2 or pEF2, while EC1118 could only be transformed with the episomic construction pEF2.

Occasionally, some colonies appeared in the control plates (pRS316, YEp352, or no-DNA transformations) in PFP resistance experiments. Because their numbers were sometimes equivalent to about 1/10 the number of transformants in the pCA2 or pEA2 plates, our transformation frequencies may have been overestimated because of the presence of false-positive transformants. We took advantage of the double markers of our constructs (ARO4-OFP and URA3) and the fact that two of the yeast strains were auxotrophic for uridine (BY4741 and T_73-4) to easily test this possibility. PFP-resistant strains from each transformation experiment were replica plated on SD without uridine and YPD. The results were unequivocal, as 159 of the 160 tested colonies from pCA2 or pEA2 transformations were prototrophic for uridine, indicating that they had actually incorporated the transforming plasmid, while the strains from pRS316, YEp352, or no-DNA transformations were uridine auxotrophs. For the first transformation experiments with strains EC1118 and IFI473, a limited number of colonies was also analyzed in order to verify “true” transformation. This was done by plasmid DNA extraction and restriction analysis of the plasmids obtained. As expected, 14 of the 14 strains analyzed from pCA2 or pEA2 transformations contained the plasmid used for transformation.

Sulfite resistance was a more uncertain selection marker for yeast transformation. Several transformation experiments had to be discarded because of an unacceptable background level resulting in a lawn of colonies for the negative controls. In spite of this random factor, we were able to perform several successful transformation experiments and to test the authenticity of the transformants (as described above for PFP-resistant strains). Two hundred nineteen of the 243 tested colonies from pCF2 or pEF2 transformations were confirmed to be true transformants, either by uridine prototrophy (BY4741 and T_73-4 derivatives) or because they contained a plasmid with the expected restriction map. However, the rate of false transformants in some particular experiments with the FZF1-4 marker was as high as 80%. Only confirmed transformants were taken into account to compute the transformation frequencies shown in Table 2. Transformation of a non-wine industrial yeast strain, brewer's yeast strain IFI234, which had a sulfite tolerance level equal to that of the laboratory strain, was unsuccessful.