Generation of a large set of genetically tractable haploid and diploid Saccharomyces strains

Abstract

Saccharomyces cerevisiae has proved to be an invaluable model in classical and molecular genetics studies. Despite several hundreds of isolates already available, the scientific community relies on the use of only a handful of unrelated strains. The lack of sequence information, haploid derivatives and genetic markers has prevented novel strains from being used. Here, we release a set of 55 S. cerevisiae and Saccharomyces paradoxus genetically tractable strains, previously sequenced in the Saccharomyces Genome Resequencing Project. These strains are stable haploid derivatives and ura3 auxotrophs tagged with a 6-bp barcode, recognized by a restriction enzyme to allow easy identification. We show that the specific barcode can be used to accurately measure the prevalence of different strains during competition experiments. These strains are now amenable to a wide variety of genetic experiments and can be easily crossed with each other to create hybrids and segregants, providing a valuable resource for breeding programmes and quantitative genetic studies. Three versions of each strain (haploid Mat a and Matα and diploid Mat a/α all as ura3∷KanMX-Barcode) are available through the National Culture Yeast Collection.

Saccharomyces cerevisiae, Saccharomyces paradoxus, haploid, barcode, resequencing, strains

Introduction

Model organisms have been used for decades in order to elucidate biological processes. Among these organisms, the unicellular Saccharomyces cerevisiae has been widely used in classical genetics, molecular biology, biochemistry and, lately, comparative genomics (Liti & Louis, 2005; Landry, 2006). Its properties, such as small genome size, short cell cycle, easy culturing and rapid gene deletion made possible the implementation of large-scale studies, making the budding yeast the best-characterized eukaryotic organism (Dujon, 1996; Goffeau, 1996; Zeyl, 2000; Giaever, 2002; Klinner & Schafer, 2004; Landry, 2006).

Early advances in yeast genetics were obtained using the heterothallic strain S288c. This strain was obtained through genetic crosses from the EM93 progenitor during the 1950s (Mortimer & Johnston, 1986). The popularity of S288c is mostly because it is a rare natural heterothallic (ho) strain and thus a stable haploid, whereas the large majority of S. cerevisiae strains are homothallic diploids (Mortimer, 2000; Bradbury, 2006; Cubillos, 2009). Furthermore, this strain was the first eukaryotic organism to be sequenced (Dujon, 1996; Goffeau, 1996; Mewes, 1997). From the original S288c background, several other strains were generated through further crosses and genetic manipulations (http://wiki.yeastgenome.org/index.php/straintable). One example is the BY strains series (BY4741, BY4742 and BY4743), widely used to generate complete collections of gene deletion mutants to perform functional genomic analyses (Winzeler, 1999; Giaever, 2002; Warringer, 2003). Similarly, W303 was derived from S288c through genetic crosses with other strains (e.g. D311-3A) and has been widely used in genetic experiments (Rothstein, 1977; Rothstein & Sherman, 1980; Fan, 1996).

A few other lab strains, unrelated to S288c, have been developed as genetic models. These strains include Y55 and SK1, both originally homothallic and commonly used in meiotic studies for their rapid and high-efficiency sporulation (Tauro & Halvorson, 1966; Kane & Roth, 1974; McCusker & Haber, 1988; Bishop, 1992). Linkage analyses on segregants from a cross between SK1 and S288c have found several quantitative trait loci (QTL) explaining the differences in sporulation efficiency (Deutschbauer & Davis, 2005; Ben-Ari, 2006). The quantitative variation in numerous phenotypes among strains represents a great resource in mapping novel gene interactions, epistasis and major loci effects. Examples include the wine strain RM11-1a and the clinical isolate YJM789, where multiple QTL were mapped for DNA damage sensitivity (Demogines, 2008), drug response (Perlstein, 2007) and temperature-sensitive growth (Sinha, 2008). However, this handful set of strains do not represent the wide diversity found in S. cerevisiae and Saccharomyces paradoxus. In this context and with the aim of extending the number of strains available for research studies, the Saccharomyces Genome Resequencing Project (SGRP) has released genome sequences of 72 S. cerevisiae and S. paradoxus isolates collected from numerous geographic locations and sources (Liti, 2009a). Genome sequences were used to resolve the phylogenetic relationship among the strains. While S. paradoxus strains present a clear population structure with four highly diverged lineages from different geographic locations, sequence variation in S. cerevisiae is much lower and comparable to a single S. paradoxus population. Half of the S. cerevisiae strains sequenced fall into five distinct clean lineages with the majority of segregating sites being unique within a population (Liti, 2009a). The remaining strains appear to have originated from crosses between two or more clean lineages, resulting in mosaic genomes. This view of the population structure is also consistent with whole-genome genotyping analyses of 67 S. cerevisiae strains (Schacherer, 2009).

The strong correlation found between genotype and phenotype within the SGRP strains makes them an ideal tool in linkage and association analyses. In this paper, we describe the technical approaches used for the generation of 55 heterothallic haploid, ura3-deleted and barcode-tagged S. cerevisiae and S. paradoxus strains sequenced by the SGRP. Transformation efficiency and sporulation levels were both found to vary considerably among isolates. This new set of strains provides a valuable resource for both classical and molecular genetic studies.

Materials and methods

Strains

Details of the origin of the strains used in this study were reported previously (Liti, 2009a). All the strains, with the exception of W303, were originally wild-type isolates with no selectable auxotrophic markers.

Transformation and gene deletion

Yeast transformation was performed for both S. cerevisiae and S. paradoxus strains using the lithium acetate (LiAc) protocol (Gietz & Schiestl, 2007) with some adjustments. Briefly, cells were grown overnight in 5 mL of yeast peptone dextrose (YPD) until saturation. Then, diluted to an OD of 0.2 and grown for 4 h or up to 0.6–0.8 OD. Each culture was harvested by centrifugation at 2000 g for 5 min and resuspended in 1 mL of sterile water. Cells pellet was washed three times in water and spun at 16 000 g for 1 min between washes. Cells were then washed three times in 0.1 M LiAc and the pellet was recovered after centrifugation. Cells were divided into two samples to which were added 240 μL of polyethyleneglycol (50% w/v), 36 μL of LiAc (1 M), 20 μL of Salmon Sperm Carrier DNA (10 mg mL⁻¹, Invitrogen), 25 μL of water and 1 μg of DNA. Cells were vortexed for 1 min and incubated at 30 °C for 30 min, then at 42 °C for 25 min and finally washed in water and incubated in YPD at 4 °C overnight. The cell suspension was plated on appropriate media and transformants were isolated after 2–3 days. A unique barcode, positioned adjacent to the 5′ KanMX promoter, was inserted into each strain, replacing the URA3 gene. KanMX sequence was amplified from plasmid pFA6a-kanMX4 (Wach, 1994). Each barcode consists of a unique 6-bp sequence, recognized by a specific restriction enzyme. Primers for targeting URA3 were designed, taking into consideration the single-nucleotide polymorphisms (SNPs) present in each strain to avoid possible mismatches. A total of 74 primers were designed for the URA3 deletion and are listed in Supporting Information, Table S1a and b.

The HO gene was originally deleted in Y55 and then used to generate a PCR template for the gene deletion in all the other strains. Two primers were designed around 400 bp upstream of the start codon and downstream of the stop codon on the Y55 HO deletion (Table S2a). The hygromycin resistance gene was used as a selectable marker and amplified from plasmid pAG32 (Goldstein & McCusker, 1999).

Crosses between haploid strains

Haploid strains with opposite mating types were crossed with each other on YPD agar. To confirm successful crosses, we performed mating test using tester strains Y55-2369 (Matα, hoΔ, ura2-1, tyr1-1) and Y55-2370 (Mat a, hoΔ, ura2-1, tyr1-1).

The LYS2 and MET15 genes were deleted in the Matα and Mat a versions, respectively, in strains Y55, YPS128 and DBVPG6044 using URA3 as the selectable marker (Table S2b). The URA3 sequence was amplified from plasmid pNSU114 (Louis & Borts, 1995). Strains Y55 (Matα, lys2∷URA3) and Y55 (Mat a, met1∷URA3) were crossed with BY4741 (MAT a, his3Δ1, leu2Δ0, met15Δ0, ura3Δ0) and BY4742 (MATα, his3Δ1, leu2Δ0, lys2Δ0, ura3Δ0) strains, respectively. Diploid cells were selected by complementation in minimal media.

Barcode amplification and digestion

PCR fragments containing the barcode sequence were generated using the primers listed in Table S1c. PCR products were digested using the specific restriction enzyme (Table 1). A total of 100 ng of DNA was digested using 1 × bovine serum albumin (as indicated by the manufacturer), 1 × enzyme buffer and 1 U of the specific restriction enzyme. Products were separated by agarose gel electrophoresis.

1

Strains obtained in this study: (a) Saccharomyces cerevisiae, (b) Saccharomyces paradoxus

Strain	Barcode	Restriction enzyme	TE	SE	SV	a/α	a	α
(a)
S288c	GGATCC	BamH1	+++	NA	NA			3608
UWOPS87-2421	CTCGAG	XhoI	+++	+++	++	3556	3582	3609
NCYC361	CCATGG	NcoI	+	−	NA
DBVPG6040	TTTAAA	DraI	+++	+	+	3557
W303	TGATCA	Bcl1	+++	NA	NA		3583
378604X	CCGCTC	BsrBI	+++	NA	NA			3610
YS2	AGTACT	ScaI	++	−	NA
YS4	AATATT	SspI	++	−	NA
YS9	AGGCCT	StuI	++	NA	NA		3584
273614N	GCGCAT	HhaI	+++	+++	+++	3558	3585	3611
YIIc17_E5	AGGCCT	AatII	+++	++	+++	3559	3586	3612
322134S	GCATGC	SphI	+++	NA	NA		3587
Y55	GATATC	EcoRV	+++	+++	+++	3560	3588	3613
UWOPS83-787.3	TCTAGA	XbaI	+++	+++	+++	3561	3589	3614
SK1	CGGCCG	EagI	+++	+++	+++	3562	3590	3615
DBVPG1853	GAATTC	EcoRI	+	+	+	3563
BC187	GGGCCC	ApaI	+++	++	+++	3564	3591	3616
YJM978	CTAGTT	BfaI	+++	+++	+++	3565	3592	3617
YJM981	AGCGCT	Eco47III	+++	+++	+++	3566	3593	3618
YJM975	ACCGGT	AgeI	+++	+++	+++	3567	3594	3619
DBVPG1373	CCCGGG	SmaI	+++	+	++	3568	3595	3620
DBVPG1106	ATCGAT	ClaI	+++	+++	+++	3569	3596	3621
DBVPG6765	GGCCAT	HaeI	+++	+++	++	3570	3597	3622
DBVPG1788	CGATCG	PvuI	NA	NA	NA	3571
L-1374	AAGCTT	HindIII	+++	+++	+++	3572	3598	3623
L-1528	CAATTG	MfeI	+++	+++	+++	3573	3599	3624
DBVPG6044	GCTAGC	NheI	+++	++	+++	3574	3600	3625
NCYC110	ACGCGT	MluI	+	+++	+++	3575	3601	3626
UWOPS03-461.4	CCCGGG	XmaI	++	+++	+++	3576	3602	3627
UWOPS05-217.3	CATATG	NdeI	+++	+++	+++	3577	3603	3628
UWOPS05-227.2	GAGCTC	SacI	++	+++	+++	3578	3604	3629
K11	TCGCGA	NruI	++	−	NA
Y9	CTGCAG	PstI	++	−	NA
Y12	GTCGAC	SalI	+++	+	++	3579	3605	3630
YPS606	ATTAAT	AseI	+++	+++	+++	3580	3606	3631
YPS128	GGTACC	KpnI	+++	+++	+++	3581	3607	3632
(b)
CBS432	GAATTC	EcoRI	+	+	+++	3633	3662	3689
T21.4	AGGCCT	AatII	+++	+++	+++	3634	3663	3690
Y7	GGGCCC	ApaI	+++	+++	+++	3635	3664	3691
Q32.3	ATTAAT	AseI	++	+++	+++	3636	3665	3692
Q59.1	TGATCA	Bcl1	+++	+++	+++	3637	3666	3393
Q95.3	AGATCT	BglII	+++	+++	+++	3638	3667	3694
S36.7	ATCGAT	ClaI	+++	+++	+++	3639	3668	3695
Z1.1	TTTAAA	DraI	+++	+++	+++	3640	3669	3696
Y6.5	CGGCCG	EagI	+++	+++	+++	3641	3670	3697
Q62.5	AAGCTT	HindIII	+++	+++	+++	3642	3671	3698
Q89.8	CAATTG	MfeI	+++	+++	+++	3643	3672	3699
Y9.6	ACGCGT	MluI	+	+++	+++	3644	3673	3700
Q74.4	CCATGG	NcoI	+++	+++	+++	3645	3674	3701
Q69.8	TCGCGA	NruI	+	+++	++	3646	3675	3702
W7	CTGCAG	PstI	+	+++	+++	3647	3676	3703
Q31.4	GTCGAC	SalI	+++	+++	+++	3648	3677	3704
Y8.5	AGTACT	ScaI	+	+++	+	3649	3678	3705
Z1	AATATT	SspI	++	+++	+++	3650	3679	3706
Y8.1	AGGCCT	StuI	++	+++	+++	3651	3680	3707
N-17	TCTAGA	XbaI	++	+++	+++	3652	3681	3708
CBS5829	CTCGAG	XhoI	++	+++	+++	3653	3682	3709
DBVPG4650	CCCGGG	XmaI	NA	NA	NA
KPN3828	CATATG	NdeI	+	++	+++	3654
KPN3829	GAGCTC	SacI	+	++	+++	3655	3683	3710
YPS138	GATATC	EcoRV	+++	+++	+++	3656	3684	3711
DBVPG6304	GCATGC	SphI	+	+++	+++	3657	3685	3712
A4	CCGCTC	BsrBI	+	+++	+	3658
A12	GCGCAT	HhaI	+	+++	+	3659	3686	3713
UFRJ50791	GGCCAT	HaeI	−	NA	NA
UFRJ50816	CTAGTT	BfaI	−	NA	NA
N-43	AGCGCT	Eco47III	−	NA	NA
N-44	ACCGGT	AgeI	+	+++	+++	3660	3687	3714
N-45	GCTAGC	NheI	−	NA	NA
IFO1804	CGATCG	PvuI	+	+++	+++	3661	3688	3715
UWOPS91-917.1	CCCGGG	SmaI	−	NA	NA

Strain name, sequence barcode and the corresponding restriction enzyme is shown. Strains were scored for transformation efficiency (TE) as a measure of the number of positive transformed colonies, sporulation efficiency (SE) based on percent of cells sporulated and spore viability (SV) as percent of viable spores. Scores: +++, high; ++, medium; +, low; −, none; NA, not applicable. The last three columns indicate whether the three strain versions, diploid Mat a/α and haploids Mat a and Matα all as ura3∷KanMX-Barcode are available and the corresponding NCYC accession number (no number means strain missing).

Tetrad dissection, marker segregation and mating-type test

Strains were sporulated for 3–5 days at 23 °C on 1% potassium acetate media and tetrads were dissected as described previously (Naumov, 1994). Spores from both rounds of deletions were tested for correct marker segregation in appropriate media. Strains obtained were genotyped for mating type by both diagnostic PCR (Huxley, 1990) and mating with tester strains.

Quantitative (Q)PCR TaqMan assay

Primers and probes were designed using primer 3 software (Untergasser, 2007). The unique barcode was used as a target to design specific probes for the strains YPS128 and DBVPG6044 (Table S3). QPCR reactions were performed as described previously with a few adjustments (Salinas, 2009). A 25-μL reaction mixture containing 12.5 μL of ABsolute QPCR ROX Mix (Thermo Scientific), 400 nM of each primer, 125 μM of probes and 50 ng of DNA was used. Amplifications were carried out in a RotorGene 6000 (Qiagen) using the following programme: 15 min at 95 °C, 40 cycles of 95 °C for 10 s, 55 °C for 15 s and 72 °C for 10 s. The sensitivity was measured by performing a standard curve using variable DNA ratios ranging from 15% to 100% between the two strains. Data analyses were performed using rotor-gene 6000 real-time rotary analyzer software (Qiagen). The fluorescence threshold was adjusted manually by examining the PCR curves generated. Finally, standard curves were obtained by plotting C_t values vs. the log DNA concentration.

Results

Deletion of URA3 gene and barcode insertion

The lack of haploid derivates and auxotrophic markers on the SGRP strains make it both difficult and time consuming to use them in further genetic studies. To overcome this problem, we have manipulated the SGRP strains, making them amenable for genetic experiments. An overview of the steps followed in this work is shown in Fig. 1.

1

Overview of strains generation. A diagram representing the steps followed for the generation of strains containing the ura3 deletion, barcode insertion and HO deletion is shown. Horizontal arrows indicate strain version stored at NCYC. The KanMX cassette inserted (white box) is flanked by the barcode (B, dark grey box) and 80 bp of ura3 sequence (light grey box). Strains were disrupted for HO using HYG and sporulated. Spores obtained were dissected and haploid spores were selected on hygromycin B (HYG) plates. Mat a and Matα cells were genotyped by PCR and confirmed by a mating test.

The original SGRP strains were nonhybrid diploids. A single spore from each isolate was obtained to avoid heterozygosity that would complicate SNP calling and sequence assembly (Liti, 2009a). The deletion of the URA3 gene in these strains was performed with the classic one-step PCR deletion approach using the KanMX gene as a dominant selectable marker (Wach, 1994). We have designed strain-specific forward PCR primers to allow the insertion of a 6-bp unique molecular barcode (Fig. 2a). This barcode was designed as an identifier consisting of a restriction site that allows enzyme cleavage. Initially, 36 out of 37 S. cerevisiae isolates were successfully transformed, showing great variation in transformation efficiencies (Table 1). Detection of positive transformants was confirmed by diagnostic PCR.

2

URA3 deletion and barcode Insertion. (a) Amplification of KanMX gene from plasmid pFA6a-kanMX4 (white box) with URA3 flanking sequences (light grey box). Oligonucleotides for PCR amplification are shown above the plasmid pFA6a-kanMX4. The PCR product containing each barcode (B, dark grey box) were used to transform the isolates and generate ura3∷KanMX-Barcode/URA3 strains. (b) PCR reaction was performed to amplify the barcode insertion using primers indicated by the horizontal arrows. PCR fragments containing the barcode were digested with the appropiate enzyme depending on the strain. The example depicted is strain UWOPS87-2421 digested with XhoI. (c) Barcode restriction in isolates L-1528, UWOPS87-2421 and UWOPS03-461.4. Lanes: M, marker; 1, L-1528 undigested; 2, L-1528 digested with MfeI; 3, UWOPS87-2421 digested with XhoI; 4, UWOPS03-461.4 digested with XmaI; and 5, UWOPS03-461.4 digested with MfeI.

Similarly, 29 out of 35 S. paradoxus isolates were successfully transformed. We found that European strains exhibited a higher transformation efficiency. In contrast, American strains presented a lower number of colonies and several attempts were needed to obtain the URA3 deletion. In six of the S. paradoxus strains, G418 resistant colonies were obtained but none contained a deletion at the URA3 locus.

The functional version of the HO gene is present in most of the wild isolates and allows mating-type switching and autodiploidization (Herskowitz, 1988). Because most of the transformants were derived from homothallic strains, the heterozygotes previously generated (ura3∷KanMX- Barcode/URA3) were sporulated and dissected by micromanipulation in order to obtain homozygote ura3Δ versions (Fig. 1). Isolates presented quantitative differences in their sporulation efficiency and spore viability (Table 1). Viable spores obtained from tetrads with correct 2 : 2 segregations in URA dropout and G418 plates were used for further manipulations. After dissection, homozygous diploid deletions (ura3∷KanMX-Barcode/ura3∷KanMX-Barcode) were generated. All 29 S. paradoxus strains were obtained as ura3 homozygous diploids after tetrad dissection. Of the 36 S. cerevisiae strains transformed, 25 were successfully dissected.

Five strains were already haploids. In the cases of S288c and W303, these strains were known to be ho Matα and Mat a, respectively. Likewise, two clinical isolates 322134S and 378604X are stable haploid Mat a and Matα, respectively. Finally, strain YS9 showed only one mating type (Mat a), however, even though the strain was barcoded at the right URA3 locus, it still grows in URA dropout media, suggesting the presence of an extra copy of the URA3 gene. Five strains [National Culture Yeast Collection (NCYC) 361, K11, Y9, YS2 and YS4] could not be further analysed because they did not sporulate efficiently or had poor spore viability. One strain, DBVPG1788, was found to be highly resistant to all three commonly used drugs Nourseothricin, Hygromicin B and G418 (Wach, 1994; Goldstein & McCusker, 1999). URA3 auxotrophy in this strain was obtained by selecting spontaneous ura3/ura3 mutants from mass-germinated spores.

Generation of stable haploid strains

In order to generate haploid strains, the HO gene was deleted using the hygromycin resistance gene as marker. This gene disruption was initially obtained in the Y55 strain. Positive transformed colonies were obtained with high efficiency, then sporulated and dissected. Later, DNA from this strain was used as a PCR template for transforming all the remaining isolates (including S. paradoxus). Primers were targeted ∼400 bp upstream and downstream of the original insertion, yielding a larger region of sequence homology, which increased the transformation efficiency. After tetrad dissection, Mat a and Matα versions were obtained in 23 S. cerevisiae strains transformed previously, with the exception of isolates DBVPG6040 and DBVPG1853, which did not sporulate. Similarly, 27 out of 29 S. paradoxus strains (except KPN3828 and A4) were obtained as Mat a and Matα haploids (Table 1). These strains are now stable haploid and ura3Δ. One potential use of this set of strains is to generate crosses between them. Using a standard protocol, we efficiently obtained diploids from different pairs of strains with opposite mating type.

We then tested the possibility of introducing further markers to allow complementation between strains and thus selection in minimal media, thereby making large-scale crossing feasible and easy to automate. We deleted MET15 and LYS2 using URA3 as the selectable marker in Y55, YPS128 and DBVPG6044 Mat a and Matα strains, respectively. The Y55 Matαlys2∷URA3 and Y55 Mat a, met15∷URA3 transformants were crossed to the strains BY4741 (Mat a; his3Δ1; leu2Δ0; met15Δ0; ura3Δ0) and BY4742 (Matα; his3Δ1; leu2Δ0; lys2Δ0; ura3Δ0). Several colonies obtained from the mass mating between the Y55 and BY strains were selected in minimal media, tested and identified as correct diploids. Similarly, strains YPS128 Mat a, met15∷URA3 and DBVPG6044 Matα, lys2∷URA3 were efficiently crossed. Using this approach, large-scale crosses can be performed, including crossing to the full BY gene deletion collection. Furthermore, the availability of URA3 as a selectable marker allows marker recycling, facilitating deletion of multiple genes (Alani, 1987).

Barcode for strain identification and quantification

The barcode insertion was performed for specific strain identification and quantification in competition experiments. To test the use of the molecular barcode in strain identification, we digested the amplicon generated from the transformation diagnostic PCR on several strains (Fig. 2b and c). The correct specific cleavage observed demonstrates the easy strain identification procedure.

We then investigated the possibility of using the barcode to quantify the prevalence of DNA from one strain over another using a TaqMan QPCR assay. Two haploid strains were selected: YPS128 (MATα, ho∷HYG, ura3∷KanMX-Barcode) and DBVPG6044 (MATα, ho∷HYG, ura3∷KanMX-Barcode. These two strains contain a 2-bp difference within the barcode sequence, which was utilized to design specific probes. The QPCR reaction showed high specificity with no fluorescence detected when the YPS128 probe was used with DBVPG6044 DNA and vice versa. To test the sensitivity of the assay, known concentrations of DNA from each strain were mixed together in different ratios (15–100%) and used as a PCR template. Standard curves for each strain were plotted using the resulting C_t values as a function of the log DNA concentration. A linear range of detection was obtained with high efficiency (R=0.991) in both cases (Fig. 3). These results show that the barcode inserted in each strain is an ideal target that can be tested with specific TaqMan probes over the same genomic region, reducing optimization time for each specific assay. A major use of this system is the strain quantification during competition experiments.

3

QPCR standard curve for DNA quantification of strains YPS128 and DBVPG6044. Variable ratios of DNA concentration between strains YPS128 (a) and DBVPG6044 (b) were used to generate a standard curve by the TaqMan method using the molecular barcode as target. C_t values obtained were plotted against the log DNA concentration. The graph shows the mean value of two replicates.

Discussion

The SGRP project released genome sequences of over 70 Saccharomyces strains (Liti, 2009a). Although the genomes released are not complete in each strain, an imputation method was applied to correct sequencing errors and fill in missing values. Furthermore, the advent of next-generation sequencing technologies now makes feasible the complete genome sequences of all these strains. The same set of strains was also phenotyped for over 200 traits in order to obtain a precise phenomics map. Both sequence and phenotype information can be used in numerous ways, providing a valuable resource for future studies.

In order to facilitate the use of the SGRP strains in genetic experiments, we created haploid, ura3-deleted versions for the majority of the strains (Fig. 1). URA3 was deleted because it is the most widely used selectable marker in plasmids (Brachmann, 1998; Storici, 2001; Nieduszynski, 2006; Frazer & O'Keefe, 2007; Sadowski, 2008) and can be used for several assays without the possibility of integration at the original locus. The strains produced here were successfully used to assess plasmid-based DNA replication using a gap repair assay (C.A. Nieduszynski, pers. commun.) and by inserting URA3 near chromosome ends to measure telomeric silencing (Liti, 2009b).

In addition to the ura3 deletion, a 6-bp barcode was also inserted upstream of the KanMX cassette (Fig. 2a). Strains can be uniquely identified by enzymatic cleavage, barcode sequencing, HRM-PCR (data not shown) and TaqMan PCR (Figs 2 and 3) proving useful in the case of cross-contamination or in experiments that involve a pool of strains (Shoemaker, 1996). The small size of the barcode allows full sequence of this tag, even when short reads sequencing methods, such as Solexa, are used. Furthermore, the barcode allows accurate quantification of strains, which is essential in competition experiments.

We found that transformation efficiencies varied extensively among strains. In general, all S. cerevisiae and European S. paradoxus strains were transformed with high efficiency, whereas American (except YPS138) and Far Eastern S. paradoxus transformed with notably lower efficiency (Table 1). Most of the S. cerevisiae strains used in this study show high sporulation levels. Higher levels of sporulation are observed in some S. paradoxus isolates. A total of 23 S. cerevisiae and 27 S. paradoxus strains were successfully obtained as Mat a and Matα derivates after tetrad dissection and diagnostic PCR. There is a risk that the two rounds of deletions and sporulation have introduced new mutations into these strains. To alleviate this problem, we used only spores from four spore-viable tetrads displaying correct marker segregations.

In conclusion, we released a large set of S. cerevisiae and S. paradoxus strains in a format that can be easily used in the lab. These strains have been extensively characterized at both the genomic and the phenomics level, thus representing a valuable resource for several studies of yeast biology including association and linkage analyses (Marullo, 2007; Perlstein, 2007; Demogines, 2008; Sinha, 2008), speciation (Liti, 2006; Greig, 2009) and breeding programmes (Attfield & Bell, 2006; Saitoh, 2008; Zara, 2008).

Acknowledgements

We thank Dr Marcus Marvin, Dr Sarah Sharp, Dr Shun Adachi, Dr Steve James and Dr Conrad Nieduszynski for useful comments. This work was supported by The Wellcome Trust and Biotechnology and Biological Sciences Research Council.

Supporting Information

Fig S1. Overview of the number of strains generated.

Table S1. Primers used for URA3 deletion, barcode insertion and barcode restriction.

Table S2. Primers for HO, LYS2 and MET15 deletions.

Table S3. Primers and probes used for the QPCR TaqMan assay.

Author notes

Editor: Teun Boekhout