Abstract
We rapidly produced and isolated novel yeast hybrids by using two-color flow cytometric cell sorting. We labeled one parent strain with a fluorescent green stain and the other parent with a fluorescent orange stain, and hybrids were selected based on their dual orange and green fluorescence. When this technique was applied to the production of hybrids by traditional mating procedures, more than 96% of the isolates were hybrids. When it was applied to rare mating, three hybrids were identified among 50 isolates enriched from a population containing 2 × 106 cells. This technology is not dependent on genetic markers and has applications in the development of improved industrial yeast strains.
Strains of the yeast genus Saccharomyces are used in some of the largest and oldest biotechnology industries (16), including baking, brewing, distilling, and winemaking. Improvements in the performance of the yeast strains used in these processes have come about as a result of the development of strains with novel genotypes. The methods used to obtain these improved genotypes include genetic engineering (22), protoplast fusion (21), and mutation-selection techniques (10). However, in many situations, traditional techniques involving mating followed by selection are still effective for strain improvement (5). In these techniques, spores derived from parental Saccharomyces strains are isolated, germinated, and allowed to mate. The hybrids produced from these matings can be screened to identify novel strains that have desirable industrial traits (9, 13).
If parent strains, haploids, and hybrids are all able to grow on the same media, separating hybrids from both parents and haploids can be difficult. In laboratory studies of yeast genetics, when two haploid yeast strains with complementary genetic markers and opposite mating types are mixed, they mate, and the hybrids formed can be identified by growth on selective media. Alternatively, hybrids can be physically isolated with a micromanipulator (3).
In general, industrial yeast strains used for baking or brewing lack selectable genetic markers (4, 14), making identification of hybrids by genetic complementation impossible. Genetic markers can be introduced by mutation into industrial strains; however, this is difficult due to polyploidy and is undesirable due to possible effects on industrial performance. Furthermore, many industrial strains sporulate at low frequencies, and a high proportion of the spores produced are not viable (7). Due to these problems it is difficult to produce and isolate the large number of new strains required to identify industrial yeast strains with overall improved characteristics. One method used to overcome this problem has been to introduce antibiotic resistance markers into a yeast, which allows workers to identify hybrid strains (15). However, this approach is limited by the small range of suitable markers and the labor-intensive nature of the procedures. In addition, the presence of antibiotic resistance markers in industrial yeast strains is considered undesirable because the products are released live into the environment.
In this study, we produced and isolated rare mating hybrids between an industrial baker’s yeast strain and a laboratory yeast strain without using selective markers.
MATERIALS AND METHODS
Strains and culture conditions.
All of the strains used in this study are available from the Australian Nation Reference Laboratory in Medical Mycology (AMMRL), Royal North Shore Hospital, Sydney, New South Wales, Australia. Studies were performed with three strains of Saccharomyces cerevisiae, PB1 (= AMMRL 57.9) (MATa trp1 his1 MAL6T::lacZ) (1), SMC19-A (= AMMRL 57.11) (MATα MAL2-8c MAL3 leu1 SUC3) (17), and industrial baking strain N1 (= AMMRL 57.10). The yeast strains were grown on a variety of media. Rich medium contained (per liter) 20 g of glucose (Oxoid, Sydney, Australia), 5 g of yeast extract (Oxoid), 10 g of peptone (Oxoid), 3 g of KH2PO4 (Sigma-Aldrich, Sydney, Australia), and 20 g of agar (Oxoid). Two types of minimal media were used. Maltose minimal indicator medium contained (per liter) 20 g of maltose, 6 g of Na2HPO4, 6 g of KH2PO4, 20 g of agar, 6.7 g of yeast nitrogen base (YNB) (Difco, Sydney, Australia), and 40 μg of 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside, and glucose minimal indicator medium contained (per liter) 20 g of glucose, 20 g of agar, and 6.7 g of YNB without amino acids supplemented with CSM (Bio 101, Inc., Vista, Calif.) lacking histidine, leucine, and tryptophan (as recommended by the manufacturer). Cultures were prepared for mating by growing them for 18 h in 5 ml of rich medium in capped 50-ml centrifuge tubes at 30°C with shaking at 200 rpm. After this, cells were washed and resuspended in rich medium to a density of approximately 108 cells/ml.
Staining with CT dyes.
Cell Tracker (CT) probes were obtained from Molecular Probes, Inc., Eugene, Oreg. Stock solutions (10 mM) of CT-Green (CMFDA), CT-BODIPY, CT-Orange, CT-Yellow-Green, and CT-SNARF were prepared with dimethyl sulfoxide (99.9 atom%; Sigma, Sydney, Australia) from a freshly opened flame-sealed ampoule. Dye stock solutions were stored frozen at −50°C in single-use aliquots and were sealed to prevent exposure to moisture or light. After defrosting, any unused dye from each aliquot was discarded. Staining for flow cytometry was performed in 1-ml (final volume) reaction mixtures, each of which consisted of 25 μl of a suspension of an overnight yeast culture added to 975 μl of YNB. Thus, during staining the cell density was approximately 107 cells/ml. For staining, cells were typically incubated at 30°C for 45 min in the dark with dye at a working concentration of 10 μM. This concentration was determined after preparations were stained with dyes at working concentrations ranging from 0.5 to 25 μM (data not shown). Unbound dye was removed by centrifugal washing for 1 min at 12,000 × g, the accessible supernatant was removed with a pipette, and the pellet was resuspended in 1 ml of YNB. This procedure was performed three times. To allow time for slow leakage of unbound dye, cells were incubated for an additional 30 min at 30°C in YNB in new microcentrifuge tubes, and this was followed by three centrifugal washes.
Staining with PKH-26.
PKH-26 (Sigma) was stored and used to stain cells as described in the manufacturer’s instructions. Dye concentrations ranging from 2 × 10−6 to 8 × 10−6 M were used.
Staining with BR.
We prepared aqueous dye stock solutions of Beljian Red (BR) (1a). Staining reaction mixtures (final volume, 150 μl) were prepared by pelleting 25 μl of cells from an overnight yeast culture, resuspending the cells in 50 μl of YNB, and then adding 100 μl of a BR stock solution. Thus, during staining the cell density was approximately 108 cells/ml. Staining reaction mixtures were incubated at room temperature (21 to 25°C) for 30 min. Unbound dye was removed by three centrifugal washes, the preparations were incubated for 30 min at 30°C in YNB in new microcentrifuge tubes, and then three additional centrifugal washes were performed.
Mating procedure.
After cells were stained and washed, they were resuspended in 500 μl of 10× YNB. Two parents were transferred to a 1.5-ml microcentrifuge tube and vortex mixed. After centrifugation for 1 min at 12,000 × g, the cells were incubated statically at 20°C for 16 h in the dark.
Flow cytometry.
To disrupt aggregates, all samples were vortex mixed for 10 s immediately prior to flow cytometry. A FACSCalibur-Sort flow cytometer (Becton Dickinson, Lane Cove, New South Wales, Australia) was operated with Isoton II (Coulter Electronics Ltd., Brookvale, New South Wales, Australia) diluted 1:1000 with filtered (pore size, 0.2 μm), purified (MilliQ filter; Millipore, Sydney, Australia) distilled water as the sheath fluid. ImmunoCheck beads (Coulter Electronics Ltd.) were analyzed each day to ensure that the cytometer was correctly aligned. The flow rate was adjusted to keep the total data rate below 1,000 events per s during analysis or below 300 events per s during sorting. The detection threshold in the forward scatter channel (FSC) was set at a level just below the level of the lowest yeast cell signals. The excitation light (wavelength, 488 nm) was light from a 15-mW argon ion laser. Fluorescence was monitored in fluorescence channel 1 (FL1) (CT-BODIPY, CT-Green, CT-Yellow-Green) or FL2 (CT-SNARF, BR, PKH-26). The compensation controls consisted of unlabeled and single-dye-labeled cells, and these cells were prepared each time that staining was carried out for compensation setting. The actual settings used depended on the physiological condition of the cells and the dye concentrations and were different for each mating pair. However, typical settings are shown in Table 1.
TABLE 1.
Typical flow cytometer settings used for hybrid selection
| Channel(s) | Voltage (V) | Gain | Compensation setting (%) |
|---|---|---|---|
| FSC | E00 | Linear, 1 | |
| SSC | 300 | Linear, 1 | |
| FL1 | 600 | Log | |
| FL2 | 600 | Log | |
| FL1-FL2 | 5 | ||
| FL2-FL1 | 75 |
Characterization of cell type.
Sort regions were defined on an FL1-versus-FL2 dot plot. To determine the nature of cells sorted from the defined regions, sorted cells were examined by microscopy, and regions were modified until sorting accuracy was confirmed. The sorter was always set to single-cell mode. To sort rare mating hybrids, the first round of sorting consisted of collecting 20,000 events in bovine serum albumin-coated (1) 50-ml sterile Falcon tubes (Bacto, Sydney, Australia). Sorted cells are recovered in high volumes from catcher tube sorters, such as the FACSCalibur sorter. Therefore, cells were concentrated by centrifugation at 4,000 × g for 20 min. The supernatants were carefully removed, and the pellets were resuspended in 2 ml of YNB. The resuspended cells were then placed in the cytometer for a second round of sorting, and 50 cells were collected in 50-ml tubes. Cells obtained from this second round were concentrated under sterile conditions (laminar flow cabinet) on 47-mm-diameter, 0.22-μm-pore-size membrane filters (Millipore) and inoculated onto rich medium. The resulting plates were incubated at 30°C for 48 h before colonies were analyzed.
For common mating, the first round of sorting was performed as described above, and 20,000 cells were collected in 50-ml tubes and resuspended in 2 ml of YNB. The resuspended cells were placed in the cytometer for a second round of sorting, and 150 cells were collected and inoculated directly onto both rich medium and glucose minimal indicator medium in triplicate. The resulting plates were incubated at 30°C for 48 h. A comparison between the number of colonies on the rich medium and the number of colonies on the minimal medium revealed the efficiency of hybrid isolation. For confirmation, 100 colonies from the rich medium were picked randomly and transferred to maltose minimal agar. On this medium, hybrids were identified by their ability to grow and to synthesize β-galactosidase (2). This is because strain PB1 has lacZ linked to the MAL6 promoter and integrated into the genome (2). As a result, lacZ is expressed at high levels on a maltose-based medium.
Microscopy.
An Optiphot II epifluorescence microscope (Nikon, Sydney, Australia) fitted with ×12 eyepieces and ×20 and ×40 objectives (models Fluor20 and Fluor40, respectively) was used to examine sorted populations. For clear visualization of bright-field images, Normaski differential interference contrast optics was used. The excitation source was a 50-W Hg vapor arc lamp. A type B2A filter block (excitation at 450 to 490 nm, examination at 520 nm) was used for visualization of green fluorescence. A type G2A filter block (excitation at 510 to 560 nm, examination at 590 nm) was used for visualization of orange-red fluorescence.
PCR fingerprinting.
To distinguish parent strains from rare mated hybrid strains, PCR fingerprints were obtained by using commercially available primers (Yeast Mutilplex PCR primers; Bresatec, Sydney, Australia). Band patterns were determined visually after agarose gel electrophoresis of the PCR products (18).
RESULTS
Selection of cell tracking dyes.
The criteria used to select the most appropriate dyes were as follows. Two dyes had to have different spectral emission properties (e.g., one green and the other orange) such that labeled cells could be readily discriminated by flow cytometry. Each dye had to be bright enough to allow discrimination between parental cell types. Once the yeast cells were labeled, the dyes had to be retained by the cells of the parent strain for the duration of the mating reaction. In addition, they could not rapidly leak from one strain to the other. A number of potentially suitable dyes were tested, including the CT dye range from Molecular Probes Inc. (8) and tracking dye PKH-26. Another dye, BR, was developed by our group specifically for tracking of yeast cells (1a). Three dyes, BR (orange), CT-Green (green), and CT-BODIPY (green), were found to label cells with sufficient fluorescence so that labeled cells and unlabeled cells could be distinguished. The dyes were retained for up to 25 h after staining and washing; after this measurements were not obtained (Table 2). After cells were labeled and washed, there was some leakage of all three dyes to unstained cells, but two populations (stained and unstained) could still be discriminated. CT-Orange, CT-Yellow-Green, CT-SNARF, and PKH-26 were also tested but did not brightly stain yeast cells under any of the incubation conditions tested.
TABLE 2.
Retention of cell tracking dyes by S. cerevisiae N1
| Time (h) | BR orange fluorescencea | CT-Green BODIPY green fluorescence | CT-Green green fluorescence |
|---|---|---|---|
| 0 | 590 (59) | 600 (83) | 500 (154) |
| 4 | 490 (48) | 640 (78) | 300 (88) |
| 25 | 430 (56) | 690 (132) | 180 (63) |
CT-Green BODIPY and CT-Green were measured in the green fluorescence channel (FL1) (525 ± 5 nm), and BR was measured in the orange fluorescence channel (FL2) (575 ± 5 nm). All values are expressed in arbitrary fluorescence units and are averages (standard deviations) of data from three experiments. Unstained cells had typical mean autofluorescence values of 8 to 12 in both FL1 and FL2.
Mating of haploid strains having complementary mating types.
Two brightly stained parent strains (Fig. 1A and B) were mixed together under conditions suitable for mating. Shortly after mixing (5 min), two distinct populations of cells were present in the mating reaction mixture as determined with dot plots of green fluorescence (FL1) versus orange fluorescence (FL2) (Fig. 1C). Cell sorting and microscopic observation confirmed that one of the populations (strain SMC19-A) was fluorescent green and the other population (strain PB1) was fluorescent orange. After 16 h, the fluorescence of both of these strains had decreased, and a third, dual-stained population was detected on dot plots of FL1 versus FL2 (Fig. 1D). Cell sorting and microscopic observation confirmed that this population contained cell clusters containing at least one cell of each parent, including cell pairs displaying “shmoo” morphology, which is characteristic of mating cells (12). Sorting revealed that the clear majority of the clusters in the dual-stained region were clusters containing more than two cells (actual proportions were not determined). Such multicell agglutination is typical of mating reactions and is commonly used as an indicator of mating type (6).
FIG. 1.
Mating of haploid strains of S. cerevisiae. (A) Strain SMC19-A stained with CT-Green BODIPY. (B) Strain PB1 stained with BR. (C) Preparation 5 min after two strains were mixed together under conditions suitable for mating. (D) After 16 h the gated region was used to sort mated cells. All values are expressed in arbitrary units.
Based purely on the FL1-versus-FL2 fluorescence characteristics, cell clusters could not be distinguished from doublets or mating pairs. To enrich for the mating pairs and minimize interference from multicell clusters, the FL1-versus-FL2 dot plot was gated by a region defined on the FSC-versus-side scatter channel (SSC) dot plot (Fig. 2) that included the events with the smallest amount of light scattering. To obtain better resolution, the FSC and SSC amplifiers were set to linear. This low-scatter-value population was expected to correspond to the smallest clusters, which should have included mated pairs. Cell sorting and microscopic observation of the low-scatter-value gated dual-stained population confirmed that the gating strategy used excluded large clusters and included a large proportion of mating pairs.
FIG. 2.
Gating region used for isolation of small clusters of mating cells. Linear gains were used for both SSC and FSC parameters to provide sufficient resolution for size discrimination. Values are expressed in arbitrary units.
Hybrid strains were identified by the formation of colonies on glucose minimal indicator medium and by the expression of lacZ derived from strain PB1. Using this method of analysis, we found that prior to sorting 33% of the population were hybrids. After one round of sorting the proportion of hybrids in the mixed population had increased to 70%, and after two rounds it had increased to 96%.
Rare mating of a polyploid industrial strain with a laboratory haploid strain.
Shortly after mixing, two distinct populations representing the unmated parent strains were detectable in the mating reaction mixture (data not shown). Cell sorting and microscopic observation confirmed that one of the populations (strain N1) was fluorescent green and the other population (strain PB1) was fluorescent orange. After 16 h a small proportion (<1%) of cells formed a third dual-stained population (Fig. 3). Cell sorting and microscopic observation confirmed that this dual-stained population contained cell clusters containing at least one cell of each parent. This region was sorted twice, and 50 isolates were obtained. A total of 3 of these 50 isolates had PCR fingerprints that were consistent with being the result of hybridization between the two parental strains. The three putative rare mated products had the phenotypes of both parents. Their characteristics included lacZ expression, derived from PB1, and prototrophy, derived from N1.
FIG. 3.
Rare mating of an industrial polyploid strain with a laboratory haploid strain of S. cerevisiae. Strain PB1 (haploid) stained with BR and strain N1 (polyploid) stained with CT-Green BODIPY were mixed under conditions suitable for mating. After 16 h cells from the gated region were sorted twice to enrich for rare mated hybrids. All values are expressed in arbitrary units.
DISCUSSION
The method described here was developed for rapid production and isolation of yeast hybrid strains without the need for genetic markers. In this study, the proportion of yeast hybrids isolated following mating between two strains with opposite mating types increased from 33 to 96% after two rounds of sorting. The major advantage of the method described here is that a population highly enriched for hybrids can be produced without the use of genetic markers. This process is useful for producing new strains since isolates selected from the mating reaction are likely to be hybrids.
We also demonstrated that this method could be used in situations in which hybrids are only rarely produced, such as rare mating. Rare mating can occur when a heterozygous diploid or polyploid yeast strain, which does not have a mating type, spontaneously changes into a strain with either an a or α mating type (for example, either by mutation or by recombination) (7, 19). The resultant homozygous strain can mate with a haploid strain or another polyploid strain having the opposite mating type to produce a hybrid yeast strain. In heterothallic yeast strains, which generally include the brewing and baking strains, such an event is rare (4, 7, 14, 20). The reported frequency of spontaneous mating type switching in heterothallic haploid strains is increased by DNA-damaging agents due to gene conversion (19). In this study, three rare mated hybrids were identified among 50 isolates sorted from a mating pool containing more than 2 × 106 cells.
Although the strains used in this study had well-defined phenotypes, the process should work equally well for uncharacterized strains if PCR fingerprinting is used to identify hybrids from the pool of double-sorted cells. Use of this tool for isolating rare mating events should result in production of novel strains, even in situations in which mating is highly inefficient, such as with baking and brewing strains (4, 7). This process could streamline the production of novel strains that combine the useful characteristics of different industrial parent strains.
In another application of the ability to identify rare mating events, the new technique could be used to isolate hybrids resulting from interspecific crosses. Such hybrids may have significant industrial applications since some lager strains (Saccharomyces carlsbergensis) are the result of interspecific hybridization (11). More generally, the ability to rapidly and efficiently isolate hybrids between two gametes need not be applied only to yeasts. It may be possible to apply the technique to other situations in which mating between organisms which do not have convenient genetic markers is required.
REFERENCES
- 1.Becton Dickinson. FACSCalibur users manual. San Jose, Calif: Becton Dickinson; 1996. [Google Scholar]
- 1a.Bell, P., J. Shen, and D. Deere. 1997. Unpublished data.
- 2.Bell P J L, Bissinger P H, Evans R J, Dawes I W. A two-reporter gene system for the analysis of bi-directional transcription from the divergent MALT-MALS promoter in Saccharomyces cerevisiae. Curr Genet. 1995;28:441–446. doi: 10.1007/BF00310813. [DOI] [PubMed] [Google Scholar]
- 3.Berlin V, Brill J A, Trueheart J, Boeke J D, Fink G D. Genetic screens and selection of cell and nuclear fusion mutants. Methods Enzymol. 1991;194:774–792. doi: 10.1016/0076-6879(91)94058-k. [DOI] [PubMed] [Google Scholar]
- 4.Casey G P. Yeast selection in brewing. In: Panchal C, editor. Yeast strain selection. New York, N.Y: Marcel Dekker Inc.; 1990. pp. 65–111. [Google Scholar]
- 5.Evans I H. Yeast strains for baking: recent developments. In: Spencer J F T, Spencer D M, editors. Yeast technology. Berlin, Germany: Springer-Verlag; 1990. pp. 13–53. [Google Scholar]
- 6.Fehrenbucher G, Perry K, Thorner J. Cell-cell recognition in Saccharomyces cerevisiae: regulation of mating-specific adhesion. J Bacteriol. 1978;134:893–901. doi: 10.1128/jb.134.3.893-901.1978. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Gunge N, Nakatomi Y. Genetic mechanisms of rare mating of Saccharomyces cerevisiae heterozygous for mating type. Genetics. 1972;70:41–58. doi: 10.1093/genetics/70.1.41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Haugland R. Handbook of fluorescent probes and research chemicals. Eugene, Oreg: Molecular Probes Inc.; 1996. [Google Scholar]
- 9.Jacobson, G. K., and N. B. Trivedi. December 1984. Improvement of yeast strains and their use in baking. European patent application EP128,524.
- 10.Jones R M, Russell I, Stewart G G. The use of catabolite repression as a means of improving the fermentation rate of brewing yeast strains. Am Soc Brew Chem J. 1977;44:54–59. [Google Scholar]
- 11.Kielland-Brandt M C, Nilsson-Tillgren T, Gjermansen G, Holmberg S, Pederson M B. Genetics of brewing yeast. In: Wheals A E, Rose A H, Harrison J S, editors. The yeasts. London, United Kingdom: Academic Press Inc.; 1995. pp. 244–251. [Google Scholar]
- 12.Lipke P N, Taylor A, Ballou C E. Morphogenic effects of alpha-factor on Saccharomyces cerevisiae a cells. J Bacteriol. 1976;127:610–618. doi: 10.1128/jb.127.1.610-618.1976. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Mondal, A. K., G. S. Prasad, and T. Chakrabarti. December 1997. Strains of yeast of Saccharomyces cerevisiae and a process for the population of such strains of yeast. U.S. patent 5,693,526.
- 14.Nagodawithana T W, Trivedi N B. Yeast selection for baking. In: Panchal C, editor. Yeast strain selection. New York, N.Y: Marcel Dekker Inc.; 1990. pp. 139–184. [Google Scholar]
- 15.Putrament A, Baranowska H, Prazmo W. Induction by manganese of mitochondrial antibiotic resistance mutations in yeast. Mol Gen Genet. 1973;126:357–366. doi: 10.1007/BF00269445. [DOI] [PubMed] [Google Scholar]
- 16.Reed G, Nagodawithana T W. Yeast technology. 2nd ed. New York, N.Y: Van Nostrand Reinhold; 1991. [Google Scholar]
- 17.Rodicio R, Zimmermann F K. Cloning of maltase regulatory genes in Saccharomyces cerevisiae. Curr Genet. 1985;9:539–545. [Google Scholar]
- 18.Sambrook J, Fritsch E F, Maniatis T. Molecular cloning: a laboratory manual. 2nd ed. Cold Spring Harbor, N.Y: Cold Spring Harbor Laboratory Press; 1989. [Google Scholar]
- 19.Schiestl R, Winterberger U. Induction of mating type interconversion in a heterothallic strain of Saccharomyces cerevisiae by DNA damaging agents. Mol Gen Genet. 1983;191:59–65. doi: 10.1007/BF00330890. [DOI] [PubMed] [Google Scholar]
- 20.Spencer J F T, Spencer D M. Hybridization of non-sporulating and weakly sporulating strains of brewer’s and distiller’s yeast. J Inst Brew. 1977;83:287–289. [Google Scholar]
- 21.Spencer J F T, Spencer D M. Yeast genetics. In: Campbell I, Duffus J H, editors. Yeast: a practical approach. Oxford, United Kingdom: Oxford University Press; 1988. pp. 65–106. [Google Scholar]
- 22.Stewart G G. Genetic manipulation of industrial yeast strains. Can J Microbiol. 1981;27:937–990. [Google Scholar]