Construction of a self-cloning sake yeast that overexpresses alcohol acetyltransferase gene by a two-step gene replacement protocol.

The commercial application of genetically modified industrial microorganisms has been problematic due to public concerns. We constructed a "self-cloning" sake yeast strain that overexpresses the ATF1 gene encoding alcohol acetyltransferase, to improve the flavor profile of Japanese sake. A constitutive yeast overexpression promoter, TDH3p, derived from the glyceraldehyde-3-phosphate dehydrogenase gene from sake yeast was fused to ATF1; and the 5' upstream non-coding sequence of ATF1 was further fused to TDH3p-ATF1. The fragment was placed on a binary vector, pGG119, containing a drug-resistance marker for transformation and a counter-selection marker for excision of unwanted DNA. The plasmid was integrated into the ATF1 locus of a sake yeast strain. This integration constructed tandem repeats of ATF1 and TDH3p-ATF1 sequences, between which the plasmid was inserted. Loss of the plasmid, which occurs through homologous recombination between either the TDH3p downstream ATF1 repeats or the TDH3p upstream repeat sequences, was selected by growing transformants on counter-selective medium. Recombination between the downstream repeats led to reversion to a wild type strain, but that between the upstream repeats resulted in a strain that possessed TDH3p-ATF1 without the extraneous DNA sequences. The self-cloning TDH3p-ATF1 yeast strain produced a higher amount of isoamyl acetate. This is the first expression-controlled self-cloning industrial yeast.

Detection of a point mutation in FAS2 gene of sake yeast strains by allele-specific PCR amplification.

To identify yeast mutants with a point mutation, detection of the specific mutant alleles is necessary. For this purpose, we applied allele-specific polymerase chain reaction (PCR) to detect the FAS2-1250S dominant mutant allele that encodes an altered fatty acid synthase in Japanese brewer's yeast strains. These strains are known to produce a higher amount of ethyl caproate in Japanese sake. The mutant strains were supposed to be diploid and to contain heterozygous alleles, including wild-type FAS2 and a dominant FAS2-1250S. A set of oligonucleotide primers was designed to contain different nucleotides at their 3' termini: one type was identical to the wild type and the other to the mutant FAS2. Another set of primers was designed to have an additional mismatch at the second nucleotide from their 3' termini. By testing with control strains, we established PCR conditions for specific amplification. Using these conditions and a simple template preparation procedure with SDS, the presence of the allele was detected in commercially used sake yeast strains. The method presented here will be useful for the identification of specific yeast strains.

Isolation of five laccase gene sequences from the white-rot fungus Trametes sanguinea by PCR, and cloning, characterization and expression of the laccase cDNA in yeasts.

To obtain laccase-gene-specific sequences from the white-rot fungus Trametes sanguinea M85-2, a PCR screening method was used. Degenerate primers were designed based on highly conserved copper-binding regions I and IV of known laccases and used to amplify laccase sequences from T. sanguinea M85-2 genomic DNA. A single 1.6-kbp DNA band was amplified and cloned into a vector. Partial sequences of 21 clones were classified into five groups (lcc1-5) and the deduced amino acid sequences were all homologous to known laccase sequences. Based on the partial sequence of lcc1, the 5'-end of its cDNA was obtained by a PCR termed 5' rapid amplification of cDNA ends (5'-RACE), and RT-PCR was then carried out using the 5'-primer and the poly-dT primer to obtain the full-length lcc1 cDNA. The obtained cDNA encoded a protein consisting of 518 amino acid residues and its first 21 amino acid residues were predicted to be the signal peptide for secretion. The conserved characteristic structures of laccase, such as copper-binding ligands, N-glycosylation sites, and cysteine residues for disulfide bridges, were observed. The genomic DNA sequence of the lcc1 gene was also cloned by PCR method and the sequence revealed 10 introns. The lcc1 cDNA was inserted into yeast vectors for heterologous expression by Saccharomyces cerevisiae and Pichia pastoris. Phenol-oxidizing activity was detected from transformants of the yeasts, indicating that the obtained cDNA encodes a laccase. Previously, two laccase isozymes were biochemically characterized and purified from T. sanguinea M85-2. Using the sequential PCR method presented here, we have obtained partial sequences of at least five laccase genes and one cDNA clone encoding a protein with laccase activity but without any enzymatic information, suggesting that expressed enzymes under restricted conditions may not represent all the isozymes in target microorganisms. PCR cloning and heterologous expression of the cloned genes can be an alternative method of screening enzymes if these enzymes have conserved sequences.

DNA extraction method for screening yeast clones by PCR.

A simple procedure for isolating yeast DNA suitable for use as a template for PCR amplification is described. SDS treatment alone is sufficient for extraction of chromosomal DNA from yeast cells. Cells of a yeast colony are suspended in a small volume (about 20 microL) of a 0.25% SDS solution, mixed vigorously and centrifuged. The supernatant can be directly used as a template after dilution to give an SDS concentration of less than 0.01% in the final PCR mixture.

A positive selection for plasmid loss in Saccharomyces cerevisiae using galactose-inducible growth inhibitory sequences.

Counter-selections for the loss of introduced plasmid sequences are useful for gene manipulations in yeast. We have used GAL10 promoter-mediated overexpression of GIN sequences, which inhibit the growth of cells, to develop a novel counter-selection system. Yeast cells carrying a GIN sequence grow normally on glucose medium but are unable to grow on galactose medium, whereas derivatives that have lost the GIN sequence are able to grow in the presence of galactose. We constructed autonomously replicating, integrating, and disruption plasmids carrying GIN sequences and tested their use to select for loss of the plasmid. The results showed that the GIN sequences provide a selection for efficient loss of plasmids or integrated constructs from yeast during growth on galactose medium, indicating that this system can be used for plasmid shuffling, gene replacements and marker gene recycling. This counter-selection system has wide application, because any Gal+ strain and a wide variety of marker genes can be used. In addition, counter-selection systems using growth-inhibitory sequences should be applicable to other yeasts and possibly to other organisms.

Construction of recombinant sake yeast containing a dominant FAS2 mutation without extraneous sequences by a two-step gene replacement protocol.

A novel two-step gene replacement protocol was developed to construct a recombinant industrial yeast free of bacterial and drug-resistant marker sequences. A yeast strain exhibiting cerulenin resistance conferred by a dominant mutation of FAS2 was previously shown to produce high levels of a flavor component of Japanese sake. A N- and C-terminally truncated portion of the mutant FAS2 gene was subcloned to an integrating plasmid containing an aureobasidin A-resistant transformation marker and a galactose-inducible growth inhibitory sequence (GAL10p::GIN11). The plasmid was targeted into the chromosomal FAS2 locus of sake yeast Kyokai no. 7, resulting in a tandem repeat of inactive FAS2 sequences surrounding the integrated plasmid sequences. Cells containing the integrated plasmid were unable to grow on galactose medium due to the inhibitory effect of GAL10p::GIN11. This growth inhibition allowed efficient counter-selection for cells that had undergone homologous recombination between the FAS2 repeats by their growth on galactose medium. This recombination event resulted in loss of the integrated plasmid sequences and the resulting strains should contain a single copy of either wild-type or cerulenin-resistant FAS2. The selected cerulenin-resistant strains produced approximately 3.7-fold more ethyl caproate, a flavor component, than the Kyokai no. 7 strain. Southern blot and sequence analyses confirmed the presence of the FAS2 mutation and the absence of integrated plasmid sequences in the genome of the selected strain. This gene replacement method provides a straightforward approach for the construction of recombinant industrial yeasts free of undesirable DNA sequences.

Screening and identification of yeast sequences that cause growth inhibition when overexpressed.

To isolate genes that negatively regulate cell growth, we constructed a galactose-inducible expression library with partially digested Saccharomyces cerevisiae genomic DNA fragments inserted downstream of the GAL10 promoter. In all, 240,000 yeast transformants were screened for lethality on galactose medium. From 17 such transformants identified, 16 nonoverlapping DNA sequences were obtained. Restriction mapping and determination of DNA sequences adjacent to the GAL10 promoter indicated that the inserts encoded part or all of the URA2, RBP1, TPK3, SAC7, BOI1, and BNI1 genes, and also open reading frames (ORFs) from chromosomes IV, V, IX, XI, and XIII. Some of the identified sequences lacked the amino-terminal sequences of the ORFs, suggesting that truncated forms of the proteins might be necessary for growth inhibition. The sequence of the pGA108 insert was highly homologous to the telomeric X-element and contained an ARS consensus sequence, suggesting a possible growth inhibitory effect of an RNA molecule. Overexpression of the BNI1 deltaN and BOI1 deltaN genes, which lacked amino-terminal sequences, was associated with phenotypes similar to those of mutants defective in bud formation. Overexpression of the GIN4 and GIN12 sequences induced elongated buds and a G2/M arrest-like phenotype, respectively. The phenotypes induced by the overexpression of our cloned sequences could result from either a dominant-positive or a dominant-negative effect and, unexpectedly, in one case from an effect of an RNA.

Yeast src homology region 3 domain-binding proteins involved in bud formation.

The yeast protein Bem1p, which bears two src homology region 3 (SH3) domains, is involved in cell polarization. A Rho-type GTPase, Rho3p, is involved in the maintenance of cell polarity for bud formation, and the rho3 defect is suppressed by a high dose of BEM1. Mutational analysis revealed that the second SH3 domain from the NH2 terminus (SH3-2) of Bem1p is important for the functions of Bem1p in bud formation and in the suppression of the rho3 defect. Boi2p, which bound to SH3-2 Bem1p, was identified using the two-hybrid system. Boi2p has a proline-rich sequence that is critical for displaying the Boi2p-Bem1p two-hybrid interaction, an SH3 domain in its NH2-terminal half, and a pleckstrin homology domain in its COOH-terminal half. A BOI2 homologue, BOI1, was identified as a gene whose overexpression inhibited cell growth. Cells overexpressing either BOI1 or BOI2 were arrested as large, round, and unbudded cells, indicating that the Boi proteins affect cell polarization. Genetic analysis revealed that BOI1 and BOI2 are functionally redundant and important for cell growth. delta boi1 delta boi2 cells became large round cells or lysed with buds, displaying defects in bud formation and in the maintenance of cell polarity. Analysis using several truncated versions of BOI2 revealed that the COOH-terminal half, which contains the pleckstrin homology domain is essential for the function of Boi2p in cell growth, while the NH2-terminal half is not, and the NH2-terminal half might be required for modulating the function of Bem1p. Overproduction of either Rho3p or the Rho3p-related GTPase Rho4p suppressed the boi defect. These results demonstrate that Rho3p GTPases and Boi proteins function in the maintenance of cell polarity for bud formation.

Genetic relationships between the G protein beta gamma complex, Ste5p, Ste20p and Cdc42p: investigation of effector roles in the yeast pheromone response pathway.

The Saccharomyces cerevisiae G protein beta gamma dimer, Ste4p/Ste18p, acts downstream of the alpha subunit, Gpa1p, to activate the pheromone response pathway and therefore must interact with a downstream effector. Synthetic sterile mutants that exacerbate the phenotype of ste4-ts mutations were isolated to identify proteins that functionally interact with Ste4p. The identification of a ste18 mutant indicated that this screen could identify proteins that interact directly with Ste4p. The other mutations were in STE5 and the STE20 kinase gene, which act near Ste4p in the pathway, and a new gene called STE21. ste20 null mutants showed residual mating, suggesting that another kinase may provide some function. Overexpression of Ste5p under galactose control activated the pheromone response pathway. This activation was dependent on Ste4p and Ste18p and partially dependent on Ste20p. These results cannot be explained by the linear pathway of Ste4p-->Ste20p-->Ste5p. Overexpression of Cdc42p resulted in a slight increase in pheromone induction of a reporter gene, and overexpression of activated forms of Cdc42p resulted in a further twofold increase. Mutations in pheromone response pathway components did not suppress the lethality associated with the activated CDC42 mutations, suggesting that this effect is independent of the pheromone response pathway.

Positive and negative elements upstream of the meiosis-specific glucoamylase gene in Saccharomyces cerevisiae.

The SGA1 gene encoding glucoamylase is specifically expressed late in meiotic development of the yeast Saccharomyces cerevisiae. We found that accumulation of both enzyme activity and transcripts was regulated negatively by both nutritional signals and a haploid-specific negative regulator gene of meiosis. RME1, and positively by the inducer genes for meiosis, IME1 and IME2. To study the role of sequences upstream of the SGA1 gene in its expression and regulation, we generated internal deletions in the 5' non-coding region of the gene and chimeric genes with portions of the upstream sequence inserted into a reporter gene. By analyzing the expression of these genes, we have identified both a 19 bp upstream activation sequence (UAS) and a 49 bp negatively regulating element (NRE). The UAS activated transcription with no requirement for heterozygosity at the mating-type locus, but this activation was still under negative control by nutrients. The NRE showed no UAS-like activity but conferred IME2-dependent (or meiosis-specific) expression on a heterologous promoter. These results suggest that meiosis-specific expression of the SGA1 gene is established by a regulatory hierarchy including positive and negative factors, the actions of which are mediated through the two separate upstream regulatory elements, UAS and NRE, respectively. Also, that two independently acting cascades exist for the regulation of SGA1 expression: one transduces both the mating-type and nutritional signals and includes the IME2 product, which acts to relieve the repression through NRE; and another transduces only the nutritional signal independently of the above pathway and inhibits positive factors acting on UAS.

The Saccharomyces cerevisiae genes (CMP1 and CMP2) encoding calmodulin-binding proteins homologous to the catalytic subunit of mammalian protein phosphatase 2B.

Saccharomyces cerevisiae genomic clones that encode calmodulin-binding proteins were isolated by screening a lambda gt11 expression library using 125I-labeled calmodulin as probe. Among the cloned yeast genes, we found two closely related genes (CMP1 and CMP2) that encode proteins homologous to the catalytic subunit of phosphoprotein phosphatase. The presumed CMP1 protein (62,999 Da) and CMP2 protein (68,496 Da) contain a 23 amino acid sequence very similar to those identified as calmodulin-binding sites in many calmodulin-regulated proteins. The yeast genes encode proteins especially homologous to the catalytic subunit of mammalian phosphoprotein phosphatase type 2B (calcineurin). The products of the CMP1 and CMP2 genes were identified by immunoblot analysis of cell extracts as proteins of 62,000 and 64,000 Da, respectively. Gene disruption experiments demonstrated that elimination of either or both of these genes had no effect on cell viability, indicating that these genes are not essential for normal cell growth.

Initiation of meiosis and sporulation in Saccharomyces cerevisiae requires a novel protein kinase homologue.

SME1 was cloned due to its high copy number effect: it enabled MATa/MAT alpha diploid cells to undergo meiosis and sporulation in a vegetative medium. Disruption of SME1 resulted in a recessive Spo- phenotype. These results suggest that SME1 is a positive regulator for meiosis. DNA sequencing analysis revealed an open reading frame of 645 amino acids. An amino terminal peptide of ca 400 amino acids in the deduced protein was similar to known protein kinases. Transcription of SME1 was regulated negatively by nitrogen and glucose and positively by MATa/MAT alpha and IME1, another positive regulator gene of meiosis. By complementation analysis, SME1 was found to be identical to IME2, which had been shown to be important in meiosis. These results suggest that IME1 product stimulates meiosis by activating transcription of SME1 (IME2) and that protein phosphorylation is required for initiation of meiosis.

Multiple genes coding for precursors of rhodotorucine A, a farnesyl peptide mating pheromone of the basidiomycetous yeast Rhodosporidium toruloides.

Haploid cells of mating type A of the basidiomycetous yeast Rhodosporidium toruloides secrete a mating pheromone, rhodotorucine A, which is an undecapeptide containing S-farnesyl cysteine at its carboxy terminus. To analyze the processing and secretion pathway of rhodotorucine A, we isolated both genomic and complementary DNAs encoding the peptide moiety. We identified three distinct genes, RHA1, RHA2, and RHA3, encoding four, five, and three copies of the pheromone peptide, respectively. Complementary DNA clones were classified into two types. One type was homologous to RHA1, and the other type was homologous to RHA2. Transcription start sites were identified by primer extension and S1 nuclease protection, from which the site of the initiator methionine was verified. A primary precursor of rhodotorucine A was detected as a 7-kilodalton protein by immunoprecipitation of in vitro translation products. On the basis of these results, we propose similar three-precursor structures of rhodotorucine A, each containing the amino-terminal peptide sequence Met-Val-Ala. The precursors contain three, four, or five tandem repeats of the pheromone peptide, each separated by a spacer peptide, Thr-Val-Ser(Ala)-Lys, and each precursor has the carboxy-terminal sequence Thr-Val-Ala. This structure suggests that primary precursors of rhodotorucine A do not contain canonical signal sequences.