Previous Article | Next Article ![]()
Applied and Environmental Microbiology, August 2004, p. 4686-4691, Vol. 70, No. 8
0099-2240/04/$08.00+0 DOI: 10.1128/AEM.70.8.4686-4691.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Departamento de Microbiología,1 Departamento de Genética, Facultad de Ciencias, Universidad de Extremadura, 06071 Badajoz, Spain,3 Department of Cell Biology, Institute for Cancer Research, Montebello, 0310 Oslo, Norway2
Received 16 December 2003/ Accepted 30 April 2004
| ABSTRACT |
|---|
|
|
|---|
| INTRODUCTION |
|---|
|
|
|---|
Most commercial wine yeasts are naturally occurring strains of S. cerevisiae isolated from wines and spontaneously fermenting musts. The phenotype of these wild yeasts usually is more variable than that of domesticated laboratory yeast strains, which have necessarily been selected for genetic stability to obtain reproducible research results. In wine yeasts, karyotype changes have been reported during vegetative growth (3, 15, 21), as have genetic changes that modify the yeast's metabolic properties (24). Genetic instability may alter useful properties of industrial yeasts, resulting in problems in biotechnological processes or lower quality of products such as bread, pastry, beer, or wine. For example, loss of the killer phenotype may result in protracted wine fermentation (22). Thus, obtaining genetically stable yeasts from economically interesting industrial and natural yeast strains is of economic importance. Moreover, the mechanisms responsible for genetic instability may help drive the evolution of genetic variability in natural yeast populations and could be used for controlled in vivo manipulation of the yeast genome.
Naturally occurring strains of S. cerevisiae usually are very sensitive to cycloheximide (CYH) and are inhibited by 1 to 1.5 µg/ml. The frequency of spontaneous CYH-resistant (CYHR) mutants is <2 x 105 (6, 23). Of the known CYHR mutations, the best characterized occur at the cyh2 gene, which encodes the 60S ribosomal subunit L28 protein. This mutation changes amino acid Gln (Q) 37 to Glu (E) (13, 33). The CYHR phenotype has been described as recessive (30), semidominant (6, 11, 18, 32), and dominant (27).
We analyzed the genetic instability of heterozygous cyh2R/cyh2S hybrids from naturally occurring wine yeasts. We chose the CYHR phenotype because (i) it is easy to obtain spontaneous CYHR mutants from diploid wine yeasts without altering their physiological and technological properties (23); (ii) by using the CYHR and killer K2 phenotypes, which are common in wine yeasts, heterozygous hybrids of homothallic yeasts can be obtained without the need for crosses with laboratory strains, thereby preserving the yeast's natural properties such as the genome instability itself (27); and (iii) it is easy to detect the homozygous yeasts (cyh2R/cyh2R or cyh2S/cyh2S) that originate from the heterozygous hybrids (cyh2R/cyh2S).
Our objectives in this study were to characterize and quantify the genetic instability in natural wine yeasts and hybrids synthesized in the laboratory due to LOH at the cyh2 locus. We show that the heterozygous yeast populations, when grown under nonselective laboratory conditions, can change their phenotype more rapidly than was previously known. By understanding this phenomenon, the sudden loss of interesting properties of industrial yeasts can be avoided and our knowledge of how these organisms change and evolve in industrial and field settings can be increased.
| MATERIALS AND METHODS |
|---|
|
|
|---|
50 µg/ml) and thus are easy to work with and because the mutation is thought to be located in the cyh2 gene (30). Isolation of spontaneous CYHR mutants and MIC determination were performed as described by Pérez et al. (23). 73R11D, 85R4A, 85R6B, and 85R23D are cyhR/cyhR killer-sensitive spore clones from 73R and 85R, respectively. Laboratory yeast strains (Table 1) were used for genetic mapping and the killer assay. Standard culture media were used for yeast growth and phenotype tests in the genetic mapping (10). YEPD+CYH is yeast extract-peptone-dextrose (YEPD) agar supplemented with CYH to the desired final concentration (2 µg/ml unless a different concentration is given). Minimal medium for auxotroph analysis was Difco yeast nitrogen base (without amino acids, with ammonium sulfate; Difco, Detroit, Mich.).
|
Detection of genetic instability.
Standard genetic crosses were performed to map the cyhR mutations (10). Hybridization of natural homothallic wine strains was performed as described by Ramírez et al. (27). All of the hybrids from each cross are isogenic because the parent strains are spore clones from homothallic wine yeasts. Genetic instability was assayed by measuring LOH in the populations of newly obtained heterozygous cyhR/cyhS hybrids before and after manipulation (Fig. 1). All of the hybrids were replicated and grown at the same time on the same YEPD plate (side by side) for each transfer, so all grew under the same nonselective conditions (pH 6, 30°C; absence of CYH and killer activity). The number of population doublings for each step was determined by measuring the initial and final numbers of viable cells on YEPD plates after suitable suspension and dilution of the cultures. The starting material was hybrids selected from different crosses. By the time the hybrids were obtained, their population had already undergone 32 to 49 doublings since zygote formation, depending on when cell conjugation occurred following germination of the mixed spores. The newly obtained hybrids were cultured by serial transfers on YEPD plates at 30°C (nonselective conditions) every 24 h until the population had undergone 100 doublings (from 18 to 20 transfers). The yeast population of the hybrids (both newly obtained and after 100 doublings) was analyzed as follows.
|
Liquid culture seeding on YEPD-CYH and colony analysis.
The hybrids were inoculated (4 x 106 to 6 x 106 cells/ml from a YEPD plate culture) into sterile tubes (18-mm diameter) containing 2 ml of liquid YEPD and incubated at 30°C to saturation (for 2 days with shaking at 250 rpm, 2 x 108 to 4 x 108, about six doublings). A suitably diluted sample of each culture was spread onto a YEPD-CYH plate to obtain isolated colonies. The number of colonies of different sizes was determined. Ten to 20 colonies of each type of colony from each type of hybrid were isolated and subjected to killer phenotype and tetrad analyses.
Colony isolation and YEPD+CYH replica plating.
Colonies isolated in YEPD plates were harvested, replicated on another YEPD plate, incubated at 30°C for 2 days (roughly six doublings), replica plated to YEPD-CYH and YEPD, and incubated for 1 to 8 days at 30°C to determine CYH resistance. Ten to 20 colonies of each phenotype from each type of hybrid were harvested from the original YEPD isolation plate and subjected to killer phenotype and tetrad analysis.
In parallel with the genetic analysis, we analyzed the karyotype (pulsed-field electrophoresis) (2), the restriction pattern of mitochondrial DNA (25), and the nuclear and cellular morphology (fluorescence microscopy of 4',6'-diamidino-2-phenylindole [DAPI]-stained yeasts) (4) of the hybrids before and after 100 doublings.
Genetic mapping, PCR, and DNA sequencing of CYHR mutants.
The cyhR mutations were mapped by analyzing crosses of single-spore clones from spontaneous CYHR mutants with a group of haploid yeast strains in which all of the S. cerevisiae chromosomes were genetically marked (Table 1). Yeast DNA was extracted as previously described (25). DNA samples for sequencing were amplified in a Pharmacia LKB-Gene ATAQ Controller (Pharmacia, Uppsala, Sweden) with the Ready-To-Go PCR Beads kit (catalog no. 27-9555-01) in accordance with the manufacturer's protocol. The primers used were F1 (5'-GAGACGCAAACGTTTTTCCTCGCA-3'; similar to a sequence located 157 to 181 bp upstream of the cyh2 start codon) and B1 (5'-GATAAAAACGTTGGGATCTGCCAC-3'; similar to a sequence located 185 to 209 bp downstream of the cyh2 gene translation stop codon [TAA] with the opposite orientation). PCR products were run in 1x Tris-acetate-EDTA electrophoresis buffer-1% agarose gels. As expected, the PCR with F1 and B1 gave a single 1,301-bp band. Purified PCR fragments were sequenced (at the Centro de Secuenciación de DNA, Facultad de Farmacia, Universidad Complutense de Madrid, Madrid, Spain) by standard methods in an ABI PRISM 377 DNA Sequencer (AME Bioscience Ltd., London, United Kingdom) by using the BigDye Terminator Cycle Sequencing ready reaction kit from PE Biosystems (Foster City, Calif.).
Virus (ScV-LA and ScV-M2) dsRNA extraction, purification, and agarose gel electrophoresis.
Rapid yeast double-stranded RNA (dsRNA) extraction and purification was performed as previously described (8). dsRNA molecules were separated in 1x TAE-1% agarose gel for 60 to 75 min.
| RESULTS |
|---|
|
|
|---|
Stability of newly obtained heterozygous cy2hR/cyh2S hybrids.
Heterozygous cyh2R/cyh2S hybrids were obtained by crossing cyh2R/cyh2R (85R6B) with cyh2S/cyh2S (88P1A and 85P4D) yeast spore clones. There were two classes of hybrids that could be distinguished following replica plating on YEPD+CYH: papilla (Pa) hybrids and lawn (Ln) hybrids (Fig. 2). The Ln hybrids grew slightly slower than the cyh2R/cyh2R parent at CYH concentrations of >50 µg/ml. Of the 334 hybrids analyzed, 149 were Ln hybrids and 185 were Pa hybrids. Each hybrid was a CYHR killer K2 clone from a single cell (Fig. 1) and was not cultured with CYH (unless indicated).
|
|
Changes to cyh2R/cyh2R homozygosity under nonselective conditions.
All hybrids maintained the original CYHR phenotype (lawn or papillae) after 100 doublings on YEPD with no CYH or killer activity. During this time the Ln hybrids lost the killer phenotype while Pa hybrids maintained it (Table 2). All of the cyh2R/cyh2R single-cell colonies from all of the hybrids became killer sensitive, i.e., they retained ScV-LA but lost ScV-M2, while all of the cyh2R/cyh2S and cyh2S/cyh2S colonies remained killers, i.e., they retained both ScV-LA and ScV-M2 (Fig. 3).
|
Among the Pa hybrids (Table 2), cyh2S/cyh2S cells increased for H1-1 and cyh2R/cyh2R cells increased for H2-1, while in the other hybrids (data not shown) there either was no change or the proportion of cyh2S/cyh2S yeasts decreased. This variation (always <15%) could be due to random changes in the hybrid populations that occurred during the doublings they underwent during the genetic analyses. In general, among the Pa hybrids the proportion of cyh2R/cyh2S cells decreased slightly and the proportion of cyh2S/cyh2S and/or cyh2R/cyh2R cells increased slightly.
Stability of parts of the genome other than cyh2 in Ln hybrids.
No differences were detected between the different hybrid populations in studies of pulsed-field gel electrophoresis, mitochondrial DNA restriction pattern, or nuclear and cellular morphology (data not shown). Therefore, it is unlikely that the genetic instability affects the gross organization of the cell's DNA.
| DISCUSSION |
|---|
|
|
|---|
Heterozygous hybrids of these wild yeasts maintain the genome instability. Under nonselective conditions, about half of the hybrids (Ln hybrids) have high LOH and become homozygous cyh2R/cyh2R. The other half of the hybrids (Pa hybrids) are genetically stable. Although two classes approximately equal in frequency result from this process, the stability is not heritable as a Mendelian character. The high LOH in the Ln hybrids could explain the apparently dominant behavior of the CYH2R phenotype in some wine yeasts (27) and the two phenotypes (lawn and papillae) of the heterozygous hybrids. It also makes it possible to use a recessive selectable genetic marker (such as cyh2R) to obtain new hybrids from natural homothallic yeasts; i.e., the diploid heterozygous hybrids can easily be detected if they are genetically unstable because they frequently become homozygous for the marker (27).
The LOH rate depended upon the hybrid, the time of analysis, and the calculation method but was approximately 0.7%/diploid cell/generation on the basis of the method of Puig et al. (24). This value is 3 to 5 orders of magnitude greater than the highest LOH rates previously reported (7, 24) and greater than the estimated mitotic recombination frequency in the region of chromosome VII in which cyh2 is located (7). A similar phenomenon has been reported for spontaneous homozygous LEU1/LEU1 revertants from leu1-12/leu1-12 cells with a frequency of 0.01%/diploid cell/division (7).
Near the cyh2 gene there are many repeated sequences, e.g., retrotransposon long terminal repeats (see http://www.ncbi.nlm.nih.gov/mapview/maps.cgi?org=scerevisiae&chr=VII), a chromosomal origin of replication, and some genes involved in DNA recombination (rad54 and rad6) that could be responsible for the genetic instability and LOH we observed.
Theoretically, the cyh2R/cyh2R cells could originate from mitotic gene conversion, loss of a chromosome followed by reduplication, mitotic crossing over, or a general hyperrecombination phenotype. However, the reciprocal homozygote cyh2S/cyh2S also should occur at the same frequency because of these mechanisms, but this phenotype was never observed. To test these possibilities genetically, we made hybrids between the unstable wine yeasts and laboratory yeast strains with appropriate genetic markers, but only Pa hybrids resulted. Thus, the high genetic instability phenotype is not maintained after genetic crosses are made with domesticated laboratory yeasts. No segregation of genetic instability was observed in 16 crosses of spore clones from the parents 85R6B and 88P1A (eight spores, i.e., two tetrads, from each parent were crossed with one spore from the other parent) (data not shown). Neither did we detect any significant change in the frequency of hybrid phenotypes as the result of possible homozygosity of an instability determinant, an expected result since these wine yeasts all are homothallic and generally homozygous for their entire genomes.
The LOH mechanism appears biased toward cyh2R/cyh2R homozygosity. For example, in sectored colonies of the H6-7 Ln hybrid, sectors of cyh2R/cyh2R killer-sensitive yeasts were detected but sectors of cyh2S/cyh2S yeasts were not (Fig. 3). One possible mechanism is that in the heterozygous hybrids there is a sequence close to cyh2 at which frequent DNA double-strand breaks can occur. These breaks are repaired in the Ln hybrids, but not in Pa hybrids, by break-induced replication (14, 16, 17, 31). If the double-strand break always occurs in the chromosome carrying the cyh2S allele, then the homologous cyh2R region would serve as the template for repair events, resulting in gene conversion of cyh2S to cyh2R. Alternatively, there could be a sequence that facilitates break-induced recombination that could explain the asymmetry of the repair event. Such a sequence would be cis acting and, in our system, functional only for the chromosome carrying the cyh2S allele (17). After the initial double-strand break, both fragments could act as the invasive 3' end in break-induced replication. In one case, both the sequence that facilitates break-induced recombination and the cyh2S allele are converted to yield a cyh2R/cyh2R strain. In another case, the resulting recombinants will be cyh2R/cyh2S and the sequence that facilitates break-induced recombination would still be located on the chromosome carrying the cyh2S allele and a new break-induced recombination event could occur. For each break-induced recombination event, the number of cyh2S alleles decreases by half, and eventually the Ln hybrid population becomes effectively cyh2R/cyh2R homozygous.
Break-induced recombination also could explain the 50:50 Ln/Pa ratio if break-induced recombination occurred just once after hybrid formation, resulting in half unstable Ln hybrids and half stable Pa hybrids. One or more of the genes involved in the break-induced replication mechanism (or its regulation) (31) and/or sequence differences between homologous chromosomes could be responsible for the observed phenotype.
We observed directly that the genetic instability was propagated through subculture of the Ln hybrids until they became fully homozygous. When four crosses were made with each of four spore clones from a tetrad of Ln hybrids after 100 doublings (cyh2R/cyh2R killer-sensitive) and the original cyh2S/cyh2S killer K2 parent, the crosses all yielded 25 to 75% single-cell clones with the genetic instability phenotype (data not shown). Thus, the cause of the genetic instability also persists in the Ln hybrids for at least 100 doublings.
As previously described (3), it is possible to obtain genetically stable heterozygous hybrids, e.g., Pa hybrids, for industrial use. However, we recommend that homozygous single-spore clones from heterozygous hybrids with the desired technological properties (27) be selected, since these homozygotes should not incur the sudden phenotypic changes that can ruin an industrial yeast strain. This process could eliminate recessive growth-retarding alleles from industrial yeast populations and improve fermentation vigor (28).
In conclusion, some natural wine yeasts and their hybrids have a genetic instability that causes asymmetrical LOH at the cyh2 locus and loss of ScV-M2 virus. The heterozygous yeast populations can change their phenotype under nonselective conditions more rapidly than previously observed without affecting the gross organization of the cell's DNA. This phenomenon may cause important, sudden phenotypic changes in industrial and pathogenic yeasts. Understanding the molecular basis of this process also could help us explain how these organisms change and evolve in industrial and field settings.
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| J. Bacteriol. | Microbiol. Mol. Biol. Rev. | Eukaryot. Cell | All ASM Journals |
|---|