ABSTRACT
Autolysis of Saccharomyces cerevisiae is the main source of molecules that contribute to the quality of sparkling wines made by the traditional method. In this work the possibility of accelerating this slow process in order to improve the quality of sparkling wines by using genetically engineered wine yeast strains was explored. The effect of partial or total deletion of BCY1 (which encodes a regulatory subunit of cAMP-dependent protein kinase A) in haploid and diploid (heterozygous and homozygous) yeast strains was studied. We proved that heterozygous strains having partial or complete BCY1 deletions have a semidominant phenotype for several of the properties studied, including autolysis under simulated second-fermentation conditions, in contrast to previously published reports describing mutations in BCY1 as recessive. Considering the degree of autolysis, ethanol tolerance, and technical feasibility, we propose that deletion of the 3′ end of the open reading frame of a single copy of BCY1 is a way to improve the quality of sparkling wines.
Sparkling wines made by the traditional “méthode champenoise” are the result of two main fermentative steps. First, a base wine is made by standard procedures used for the production of most high-quality white wines. For the second step, called “prise de mousse,” a combination of yeasts (Saccharomyces cerevisiae or Saccharomyces bayanus), sucrose (usually 25 g/liter [final concentration]), and eventually clarifying agents is added to the base wine. The mixture is then bottled and allowed to ferment and age. At the end of the aging period, yeast lees are removed from the bottles by disgorging (25). Before the bottles are closed for marketing, the final character of the wine is adjusted by adding wine, liquors, and sugar in various proportions. The amount of sugar added determines the sweetness of the wine.
The contribution of yeasts to the properties of sparkling wines during “prise de mousse” also takes place in two steps. First, a secondary fermentation of the added sucrose leads to the production of ethanol, carbon dioxide, and minor secondary products (25). At the end of this step, the sparkling wine has an ethanol concentration of about 9.5 to 11.5% (vol/vol), and the pressure in the bottle can reach 5 to 6 atm (25). After fermentation has been completed, there is an aging period. During aging, yeast cells die and undergo autolysis, which releases intracellular compounds, mostly peptides and amino acids, into the external medium. Improvement in the sensorial quality of the wines has been correlated with the products of the hydrolytic degradation of yeast cells, including free amino acids, peptides, mannoproteins, nucleic acid derivatives, and lipids (27, 30, 31, 39). However, autolysis in enological conditions is a very slow process that leads to long aging periods involving costly storages of the wines.
Two main methods have been devised in order to accelerate the acquisition of aging-like properties during sparkling wine production: adding yeast autolysates to the wine and increasing the temperature of aging (8, 12). Both procedures cause sensory defects in the final product, which are often described as toasty (37). A third procedure for accelerating autolysis, based on the use of killer yeast strains, has been proposed; however, the effect on wine quality has not been evaluated (48). Finally, several authors have suggested that autolytic yeast strains could be used as a way to improve the quality of sparkling wines by accelerating the acquisition of aging-like characteristics (27, 28, 46). These autolytic strains were obtained by sporulation, by random mutagenesis, or by genetic engineering of wine strains (7, 16, 46). The latest strategy was based on the observation that autophagy takes place during “prise de mousse” (6), since a gene inducing constitutive autophagy (43) was overexpressed in an industrial wine yeast strain (7).
Because cell death seems to precede autolysis during sparkling wine production (48), a potential alternative method to accelerate autolysis is the construction of wine yeast strains that quickly lose viability under stationary-phase or starvation conditions. Several S. cerevisiae genes whose mutation leads to reduced viability in response to nutrient limitation have been described previously (13); these genes include bcy1 (4, 5, 18, 34, 47), ard1, ubi4, slk1, ils1-1, ypt1, gis1, and rvs16-1 (10, 22, 34, 50, 51). We are interested in bcy1-defective mutants because they start losing viability during the diauxic shift (34), when glucose becomes exhausted from the medium (22, 50, 51). This is important because it allows the engineered strains to complete the secondary fermentation before cell death.
BCY1 encodes the regulatory subunit of cAMP-dependent protein kinase A (PKA) (4, 5, 18, 29, 34, 45, 47, 52). The PKA pathway is activated by numerous environmental signals and plays a major role in the control of metabolism, stress resistance, proliferation, and filamentous growth (33, 34, 45, 49). PKA activity is essential for glycolysis (24) and cellular proliferation (36, 44). Conversely, PKA must be inhibited for growth on nonfermentable carbon sources (24), sporulation (4), stress resistance (2, 17, 21, 44), glycogen and trehalose accumulation (20, 44), stationary-phase adaptation (9, 35, 41), and autophagy (3).
The PKA holoenzyme is a tetramer consisting of a homodimer of two catalytic subunits, encoded by TPK1, TPK2, and TPK3 (18, 52), and two regulatory subunits, encoded by BCY1 (4, 5, 18, 34, 47, 52). PKA is inactive when the regulatory and catalytic subunits are associated. In the presence of glucose, cAMP levels are high and PKA is activated. Two cAMP molecules bind to the C-terminal end of each regulatory subunit, which results in a conformational change (53). This change causes PKA holoenzyme dissociation, resulting in the release and activation of the catalytic subunits (18, 34, 47, 53).
BCY1 is not an essential gene, and null mutants display a variety of phenotypes, including an inability to grow on nonfermentable carbon sources, slow growth on glucose, stress and starvation intolerance, and inabilities to sporulate and to accumulate glycogen or trehalose (4, 5, 18, 47). BCY1 mutations have been classified according to the level of tolerance to starvation (4, 34). The “early acting class” of mutations (including null mutations) results in cell death soon after glucose exhaustion, while the “late acting class” results in cell death during the stationary phase (7 to 12 days after inoculation into rich medium) (34). Unlike null mutants, late acting mutants do not show growth defects during the exponential growth phase. In this study, we performed a more detailed phenotypic analysis of diploid heterozygous and homozygous mutants having partial or total deletions in BCY1 in order to evaluate the potential of BCY1 defective mutants for the acceleration of autolysis during sparkling wine production. The rationale behind this approach was that strains undergoing accelerated loss of viability under starvation conditions would readily undergo autolysis under sparkling wine production conditions, once the carbon source had been completely fermented.
MATERIALS AND METHODS
Microbial strains.Escherichia coli DH5α (supE44 ΔlacU169 [φ80dlacZΔM15] hsdR17 recA1 ndA1 gyrA96 thi-1 relA1) (19) was used for production of the different plasmids constructed in this study. The genotypes of the S. cerevisiae strains used in this study are shown in Table 1.
S. cerevisiae strains used in this work
Yeast media, culture conditions, and phenotype assays.Yeast strains were grown in liquid YPD medium (1% yeast extract, 2% peptone, 2% glucose) or on YPD medium plates (YPD medium plus 2% agar). Minimal medium plates contained 0.67% yeast nitrogen base, 2% agar, auxotrophic supplements, and a cognate carbon source at a concentration of 2%, as follows: glucose for SD medium, galactose for SG medium, fructose for SF medium, sucrose for SS medium, glycerol for SGy medium, maltose for SM medium, raffinose for SR medium, pyruvate for SP medium, and potassium acetate for SA medium. Synthetic base wine contained 0.17% yeast nitrogen base without amino acids and ammonium sulfate (Difco), 0.6% malic acid, 0.3% tartaric acid, 0.03% citric acid, 0.05% ammonium sulfate, the compounds required for auxotrophic growth, and ethanol at concentrations ranging from 0% to 6% (vol/vol); the pH was adjusted to 3.5 with KOH. Sugar (glucose or sucrose) was added to a final concentration of 2%. Unless indicated otherwise, yeast phenotype assays were started with cells grown overnight at 30°C and 180 rpm on YPD medium and washed three times with a saline solution (0.87% [wt/vol] NaCl).
For carbon source utilization assays, known cell numbers, as determined by microscopic cell counting, were spotted onto SD, SG, SF, SS, SGy, SM, SR, SP, or SA medium plates. The plates were incubated for 2 to 5 days at 30°C (depending on the specific medium) and photographed.
For fermentation assays 50-ml portions of synthetic base wine containing different amounts of ethanol were inoculated to obtain a final concentration of 106 cells/ml. Cultures were maintained at 30°C without agitation, and samples were withdrawn periodically. Yeast cells were removed by centrifugation, and the supernatant was stored at −20°C. Residual reducing sugars were quantified with 3,5-dinitrosalicylic acid as described by Bernfeld (1). Before quantification, sucrose-containing samples were incubated overnight at 37°C with invertase (2.8 mg/liter; Sigma-Aldrich Química, Tres Cantos, Spain) to ensure complete hydrolysis.
Survival under carbon starvation conditions was studied by inoculating a known number of cells (107 to 108 cells/ml) onto 0.67% yeast nitrogen base with the compounds required for auxotrophic growth (S medium). The suspensions were incubated at 30°C and 180 rpm, and samples were taken at different times and inoculated onto YPD medium plates at the appropriate dilutions. The plates were incubated 2 days at 30°C. The results were expressed as the viability compared to the viability of the control strains for each time.
The release of amino acids under accelerated second-fermentation conditions was studied by inoculating 106 cells/ml into 50 ml of synthetic base wine (4% [vol/vol] ethanol). The cultures were maintained at 30°C without agitation, and samples were withdrawn at different times. Yeast cells were removed by centrifugation, and the supernatant was stored at −20°C until amino acids were quantified by the modified Cd-ninhydrin method described by Doi et al. (11). The residual reducing sugars were quantified with 3,5-dinitrosalicylic acid as described by Bernfeld (1).
For stress resistance assays stationary-phase yeast cells were obtained by inoculating cells from a fresh culture into YPD medium to obtain a final concentration of 5 × 105 cells/ml and incubating the resulting preparation for 24 h at 30°C and 200 rpm. Yeast cells were subjected to different stress conditions in suspensions containing 107 cells/ml for specific times, as follows: for thermal stress, 2 h in YPD medium at 4°C, 37°C, 42°C, 47°C, and 50°C; for oxidative stress, 2 h in YPD medium with H2O2 at a final concentration of 5 mM; and for osmotic stress, 3 h or overnight in YPD medium with KCl at a final concentration of 3 M. Known numbers of cells obtained by serial dilution were spotted onto YPD medium plates and incubated for 2 to 3 days at 30°C.
Experiments were performed three to six times, and, when appropriate, data were analyzed by one-way analysis of variance and C-Dunnett test for comparisons of means. Differences were considered significant at a P value of <0.05. The calculations were carried out by using the SPSS program for Windows, release 13.0, run on a personal computer.
Molecular biology techniques.Unless indicated otherwise, all DNA manipulations were performed as described by Sambrook et al. (42). A PCR was performed using Pfu TURBO DNA polymerase (Stratagene, La Jolla, CA) by following the instructions of the supplier. Restriction enzymes were obtained from Roche Diagnostics SL, Barcelona, Spain. Yeast genomic DNA was extracted by the method of Querol et al. (40). DNA fragments resolved in agarose gels were purified with a QIAquick gel extraction kit (QIAGEN GmbH, Hilden, Germany). E. coli was transformed by electroporation (42), and plasmids were purified from E. coli cells by using a High Pure plasmid isolation kit (Roche). For transformation of S. cerevisiae we used the lithium acetate method described by Ito et al. (23), as modified by Gietz and Woods (15). Transformants were selected in YPD medium plates containing 200 μg/ml Geneticin (G418) (Life Technologies, Paisley, Scotland). Southern blot analyses were performed using a DIG High Prime DNA labeling and detection starter kit II (Roche, Mannheim, Germany) by following the instructions of the supplier.
Western blotting.Protein extraction and Western blotting were performed as described by Leber et al. (26), using a rabbit monoclonal antibody antiacetaldehyde dehydrogenase (Rockland, Gilberstville, PA) and the ECL detection system (Amersham Biosciences Europe, Barcelona, Spain). This antibody simultaneously reacts with Ald6p and Ald4p, providing an internal control for extract concentration and gel loading (32).
Construction of BCY1 defective strains.The cassettes for interruption of BCY1 in the genome of S. cerevisiae were constructed as follows. A selection-excision cassette containing the gene coding for S. cerevisiae FLP recombinase and an expression cassette for G418 resistance (KanMX), flanked by two copies of FRT (FLP target), was PCR amplified from pHFKH using primers PromBCY1-preFRT and TermBCY1-postFRT (Table 2). Plasmid pGPCR3 was the result of cloning this 3.5-kb amplicon into pGEM-T (Promega) by following the instructions of the supplier. In order to extend the BCY1 homologous regions in pGPCR3, the following S. cerevisiae genomic DNA fragments were PCR amplified: tBCY1 (positions 1328 to 1855 from the BCY1 start codon), iBCY1 (positions −500 to 18 from the BCY1 start codon), and sBCY1 (positions 653 to 1211 from the BCY1 start codon). The sequences of the primers used for these amplification reactions are shown in Table 2. The PCR fragments were inserted downstream (tBCY1) or upstream (iBCY1 and sBCY1) of the selection-excision cassette in pGPCR3 by using the strategy described by Geiser et al. (14). The resulting plasmids were designated pITGPCR3 (containing tBCY1 and iBCY1) and pSTGPCR3 (containing tBCY1 and sBCY1) (Fig. 1). These plasmids were digested with XbaI before yeast transformation. Correct integration of the insertion cassettes was verified by Southern blot analysis using a DIG High Prime DNA labeling and detection starter kit II (Roche, Mannheim, Germany) by following the instructions of the supplier.
(A) Plasmid pITGPCR3 containing iBCY1 (BCY1 initial region from position −500 to position 18 with respect to the first ATG) and tBCY1 (BCY1 terminal region from position 1328 to position 1855 with respect to the first ATG). (B) Plasmid pSTGPCR3 containing sBCY1 (BCY1 region from position 653 to position 1211 with respect to the first ATG) and tBCY1. FLP, FLP gene that encodes FLP recombinase; TEFp, TEF gene promoter; Kan, gene of transposon Tn903 coding for aminoglycoside phosphotransferase; TEFt, TEF gene terminator; FRT, FLP recombinase recognition sequences; LacZα, E. coli beta-galactosidase gene alpha peptide; ORI, E. coli pMB1 replication origin; AmpR, E. coli beta-lactamase gene; F1ORI, F1 phage replication origin.
Primers used in this work
When necessary, the selection-excision cassette was excised from transformants as follows. Yeast cells were grown for 14 days in YPD nonselective medium at 30°C and 180 rpm. Every day, culture was diluted (104) in fresh medium in order to avoid cultures with BCY1 deleted entering the stationary phase. Appropriate numbers of cells were inoculated onto YPD medium plates and incubated for 24 h at 30°C. The colonies were replica plated onto YPD medium plates containing 200 μg/ml Geneticin and 40 μg/ml BCIP (5-bromo-4-chloro-3-indolylphosphate). Sensitive strains were identified by poor growth and blue staining due to the release of intracellular phosphatases. Selection marker excision was verified by Southern blotting as described above for the primary transformants.
RESULTS
Construction of strains having partial or complete deletions of BCY1.The first attempts to disrupt BCY1 in S. cerevisiae were performed by transforming BY4741 or BY4743 with the 3.5-kb amplicon obtained by using pHFKH as the template and primers PromBCY1-preFRT and TermBCY1-postFRT (see Materials and Methods). The PCR fragment contained 90-bp terminal extensions homologous to the BCY1 open reading frame (ORF) flanking regions, which directed homologous recombination (long terminal tails were used because the Saccharomyces Genome Deletion Project [http://www-sequence.stanford.edu/group/yeast_deletion_project/deletions3.html] reported that systematic deletion of BCY1 required the use of such long tails). Because no transformants were recovered, plasmid pGPCR3 was constructed and sequenced as described in Materials and Methods. The sequence of the plasmid obtained indicated that the long terminal tails of the synthetic primers included mutations and deletions that precluded their use for homologous recombination. We decided to incorporate longer homologous regions into pGPCR3, as described in Materials and Methods, which resulted in plasmid pITGPR3 for total BCY1 deletion and pSTGPR3 for partial BCY1 deletion. The limits of the partial deletion were chosen in order to mimic mutation bcy1-53, which has been described as a member of the “late acting class” by Peck et al. (34). The following strains were constructed by this strategy: LT11 (haploid, BY4741, total BCY1 deletion), LT21 (diploid, BY4743, total deletion of one BCY1 allele), LS11 (haploid, BY4741, partial BCY1 deletion), and LS21 (diploid, BY4743, partial deletion of one BCY1 allele). LT22 (diploid, BY4743, total deletion of both BCY1 alleles) was constructed by excision of the selection marker cassette from LT21 and transformation of the resulting strain with pITGPCR3.
BCY1 deletion behaves like a semidominant mutation.As a first step toward technological characterization of the strains with BCY1 deletions, all the strains generated during this work (LS11, LT11, LS21, LT21, and LT22), as well as Y27300, were assayed to determine known phenotypes of BCY1 defective strains. These phenotypes included survival after a heat shock, osmotic stress, and oxidative stress, as well as growth in different fermentable and nonfermentable carbon sources. The results of this analysis are shown in Tables 3 and 4. As expected, the haploid strains having a complete BCY1 deletion (LT11) or diploid strains homozygous for the same deletion (LT22) were hypersensitive to all the stress conditions analyzed (Table 3). Even though BCY1 loss-of-function mutations have been described as recessive (47), the results obtained for strain LT21 clearly suggest that a strain with a BCY1 deletion behaves like a strain with a semidominant mutation, exhibiting intermediate responses to heat shock and osmotic stresses and a response similar to that of LT22 for oxidative stress. Surprisingly, Y27300, a heterozygous strain in which the BCY1 ORF was completely deleted in the same genetic background (BY4743), exhibited a slightly better response to heat shock stress than LT21 exhibited.
Stress responsesa
Growth on different carbon sourcesa
The effect of partial deletion of BCY1 was less pronounced than the effect of total deletion of the ORF. The stress tolerance observed for the haploid strain with the partial deletion, LS11, paralleled that of the heterozygous LT21 strain, while the stress tolerance of the strain heterozygous for the partial deletion, LS21, was indistinguishable from the stress tolerance of the wild-type BY4743 strain, except that the tolerance to oxidative stress was somewhat reduced.
The phenotypic differences observed between the strains were confirmed by testing growth with different carbon sources (Table 4). LT11 and LT22 showed the strongest dependence on easily fermentable carbon sources. The unexpected differences previously found between Y27300 and LT21 were confirmed, and the parallelism between LS11 and LT21 was again observed.
All these results challenged the accepted view that BCY1 mutations are recessive (47) and prompted us to continue technological characterization of BCY1 mutants by also including heterozygous strains. This had practical consequences that are discussed below.
Ethanol tolerance under simulated second-fermentation conditions.The tolerance to ethanol of the different BCY1 defective strains was tested under conditions that resembled those of second-fermentation experiments. For this purpose synthetic base wines containing different amounts of ethanol and 2% glucose or sucrose were prepared. The fermentative capacity of each strain was studied by inoculating these media and recording sugar consumption at least until the sugar was exhausted in all cultures containing the control strains. Strain LT11 was impaired for the fermentation of sugar added to the base wine in the absence of ethanol, and it was completely unable to ferment sugar in the presence of 2% or 4% ethanol (almost 1.5% residual sugar at the end of the experiment). The results for sucrose consumption in the presence of 4% ethanol are shown in Fig. 2. In contrast, LT21, a diploid strain heterozygous for the same deletion, completely fermented glucose in the absence of ethanol (data not shown) and left less than 0.5% residual sugar in the presence of 2%, 4%, or 6% ethanol, showing again that BCY1 deletion had a semidominant effect. The results for sucrose consumption in the presence of 6% ethanol are shown in Fig. 3. No apparent reduction in fermentative power was observed for the strains with partial BCY1 deletions, LS11 and LS21. These results are relevant for practical application of the findings described in this paper, as discussed below.
Fermentative capacities of haploid strains on synthetic base wines containing 4% ethanol and 2% sucrose, as described in Materials and Methods. Residual reducing sugars were quantified at several times, as described in Materials and Methods. The values are means ± standard deviations (error bars).
Fermentative capacities of diploid strains on synthetic base wines containing 6% ethanol and 2% sucrose, as described in Materials and Methods. Residual reducing sugars were quantified at several times, as described in Materials and Methods. The values are means ± standard deviations (error bars).
Loss of viability under carbon starvation conditions.Loss of viability under carbon starvation conditions was studied to further assess the differences between the deletion strains which were studied. This test was found to be more sensitive than the plate assays used in the first step of this characterization. It also had the advantage of focusing on the actual objective of this work, identifying genetic modifications that potentially lead to accelerated autolysis in S. cerevisiae. The results of the analysis are shown in Fig. 4. Both mutations tested (complete BCY1 deletion and partial BCY1 deletion) had a clear detrimental effect on cell survival either for haploid strains or for diploid heterozygous strains. After 1 day under starvation conditions the decrease in viability was noticeably greater for all the mutant strains tested than for the cognate controls. As expected, the ability to survive under starvation conditions was especially reduced for LT11 (viability was almost completely lost after 2 days). Again, the behavior of LS11 and the behavior of LT21 were more or less similar, and the relative viability stabilized at 20 to 30% of the wild-type viability after 4 days. This confirmed the semidominant character of BCY1 deletion. Interestingly, this test also revealed the semidominant behavior resulting from partial deletion of BCY1, even though LS21 was almost indistinguishable from the wild-type strain in all the previous assays.
Relative loss of viability by haploid and diploid strains under carbon starvation conditions. Cells were inoculated into S medium as described in Materials and Methods. The number of viable cells was calculated by plating on YPD medium. (A) Haploid cells. •, BY4741; ▪, LS11; ▴, LT11. (B) Diploid cells. •, BY4743; ▪, LS21; ▴, LT21. The values are means ± standard deviations (error bars).
Release of amino acids into the external medium.The release of amino acids under accelerated second-fermentation conditions (Fig. 5) was tested for a number of mutant strains as described in Materials and Methods. Because of its low ethanol tolerance, LT11 was excluded from this analysis. The results of this assay confirmed the conclusions deduced from the experiments performed under starvation conditions. All the mutant strains experienced accelerated autolysis compared to their wild-type counterparts (BY4741 and BY4743). After 3 days, before sugar was completely exhausted, LS11 had released more amino acids into the medium than all the other strains tested had released, and the difference between this strain and the BY4741 control strain reached a maximum after 10 days of incubation. The heterozygous strain LS21 having the same partial deletion in one of the alleles of BCY1 showed similar kinetics of amino acid release, although the amounts released were intermediate between the amounts for the control and the amounts for LT21. The disparities were less pronounced with prolonged incubation, especially for the haploid strains. All the differences discussed above were confirmed by analysis of variance (P < 0.05). This experiment confirmed that strains with partial or total BCY1 deletions undergo accelerated autolysis. The codominant character of the BCY1 partial and total deletions was also confirmed by this experiment.
Release of amino acids by haploid and diploid strains under accelerated second-fermentation conditions. The results are expressed as milligrams of leucine equivalents per milliliter. The values are means ± standard deviations (error bars).
Autophagy.It has been shown previously that Ald6p is preferentially degraded during autophagy, in contrast to Ald4 (32). The availability of antibodies that simultaneously recognize both enzymes provides an excellent experimental tool for monitoring autophagy. In order to strengthen the observations described above, the effect of BCY1 mutations on yeast autophagy was tested for diploid strains LS21, LT21, and LT22 compared to wild-type strain BY4743. The results are shown in Fig. 6. As expected, a clear reduction in the relative Ald6p level was observed for the control BY4743 strain after 10 days of incubation under nitrogen starvation conditions. The total lack of BCY1 activity in LT22 resulted in clearly higher relative Ald6p levels under the same conditions (Fig. 6). Again, the intermediate relative levels of Ald6p in LT21 proved the semidominant behavior of BCY1 deletion (Fig. 6). In this assay, the Ald6p levels were very similar for BY4743 and LS21 (Fig. 6).
Ald6p and Ald4p levels of diploid strains BY4743, LS21, LT21, and LT22 under nitrogen starvation conditions, as detected by Western blotting, as described in Materials and Methods.
DISCUSSION
Two main observations can be highlighted from the results presented above. First, neither total deletion of the ORF nor partial deletion involving the last 13 codons is completely recessive, as previously described, but the mutations are codominant. Second, heterozygous strains having one of the mutant BCY1 versions mentioned above undergo accelerated autolysis under conditions resembling those of the second fermentation during sparkling wine production. Both findings have practical implications for genetic improvement of second-fermentation wine yeast strains. The second finding has practical implications because it suggests a genetic engineering strategy for industrial strains aimed at accelerating autolysis, using BCY1 as a target. The first finding has practical implications because in situ deletion of a single copy of BCY1 in industrial strains would be technically much simpler than consecutive deletion of two or more copies of the gene. In addition, heterozygous strains would have the advantage of an intermediate phenotype, especially in the case of the partial deletion, exhibiting little or no increase in ethanol susceptibility but still undergoing accelerated autolysis compared to the wild-type strains.
One of the main expected consequences of PKA constitutive activity of BCY1 defective strains is delayed or reduced autophagy, and this was confirmed for several of the strains analyzed in this work. Hence, the results that we obtained are in contrast to the previous observation that strains specifically impaired for autophagy do not exhibit accelerated autolysis (7). This is probably due to the pleiotropic effects of BCY1 mutations, and impaired autophagy is just one of these effects.
We have not found a convincing explanation for the different behaviors of Y27300 and LT21. Both strains were constructed in the same genetic background, BY4743, in both strains the ORF of BCY1 is complete deleted, and in both strains the interruption cassettes carry the gene conferring resistance to Geneticin. The only appreciable difference is the fact that the cassette used for construction of strain LT21 carries additional sequences that facilitate excision of the cassette for marker recycling. Examination of the genomic context in order to find essential genes whose expression could be impaired by changes in the BCY1 locus failed to identify any essential gene adjacent to it.
The information obtained in this work has been used to design two alternative genetic engineering strategies for the improvement of industrial second-fermentation yeast strains. In one of these strategies, we are trying to sequentially delete the 3′ end of all the copies of BCY1 in one or two of the commercial strains available in our collection. Indeed, preliminary results obtained with an industrial strain heterozygous for this deletion indicate that it is able to complete the second fermentation while undergoing accelerated autolysis (data not shown). The second strategy, based on the observations of Portela et al. (38), consists of making phenocopies of heterozygous strains with BCY1 partial deletions by nondirected insertion of a one or a few copies of the truncated version of the gene into the genomes of industrial strains.
ACKNOWLEDGMENTS
We are grateful to Victoria Santamaría and J. M. Barcenilla for technical assistance. We thank V. Bruschi for providing plasmid pHFKH.
This work was supported by the Spanish Ministerio de Ciencia y Tecnología (grants AGL2003-01762 and AGL2002-01109). L. Tabera is the recipient of an I3P fellowship from the Spanish Council for Scientific Research.
FOOTNOTES
- Received 11 July 2005.
- Accepted 16 January 2006.
- American Society for Microbiology
