Introduction

The yeast Saccharomyces sensu stricto complex group comprises eight “pure” species (e.g., S. cerevisiae, S. paradoxus, S. kudriavzevii, S. eubayanus) and interspecific hybrid species such as S. bayanus and S. pastorianus (Borneman and Pretorius 2015), as schematically shown in Fig. 1A. A large number of strains belonging to the yeast group play important roles in the production of foods and beverages in addition to other biotechnological applications (Replansky et al. 2008; Sicard and Legras 2011; Steensels et al. 2014). This number is increasing owing to the employment of new isolates from nature (Hisatomi and Toyomura 2021; Minnaar and den Haan 2023). The species in the group indicate intimate relationships with each other, that is, haploid cells can mate with the opposite mating-type cells belonging to the other species in the group, and the resulting hybrid cells are viable (Toyomura and Hisatomi 2021). However, the yeast group exhibits a large variation in their genomic sequences between species. Synonymous codon similarity % of S. mikatae, S. kudriavzevii, and S. bayanus to S. cerevisiae is as low as 55% on average (Cliften et al. 2003), which could ensure the durability of the yeast group in nature and generation of valiant strains for industry.

Fig. 1
figure 1

Simple way to convert Saccharomyces sensu stricto strains accessible to conventional recombinant plasmid resources. A Phylogenetic relationships between Saccharomyces sensu stricto group species. A phylogram drawn by Borneman and Pretorius (2015) was adopted with simplification. “Pure” species names are located at the center, whereas alloploid species names are shown on the right side. Allow marks show from which species the ancestor(s) of an alloploid species obtained genomes. Rectangles highlight the species employed for DNA editing experiments in this study. B Aims of this study

Full size image

Since the reliable yeast transformation method was developed (Beggs 1978; Hinnen et al. 1978), every gene of the whole S. cerevisiae genome was cloned into various Escherichia coli-yeast shuttle vectors and integrative vectors for laboratory strains. For instance, Makanae et al. (2013) developed a series of overexpression plasmids, each of which carries a different protein-coding gene from S. cerevisiae. Various recombinant plasmid resources have been deposited in biological resource centers, and the plasmids are distributed on demand. Therefore, it is desirable to apply these plasmid resources to industrial and natural yeast strains for molecular breeding.

Most recombinant plasmids prepared for S. cerevisiae contain one of several nutrient synthesis genes (e.g., LEU2, URA3, and TRP1) as the transformant selection marker gene. The corresponding auxotrophic laboratory mutant strains (e.g., leu2, ura3, and trp1) are chosen as the host strain (Akada 2002; Gnugge and Rudolf 2017; Sikorski and Hieter 1989). However, industrial and natural isolate strains generally lack any auxotrophic mutation (prototrophic yeasts) and are mostly diploid and allopolyploid. Therefore, it is difficult to develop auxotrophic mutants suitable for transformation with the conventional recombinant plasmids in such strains. It is possible to add an antibiotic-resistance gene for yeast to the plasmids, but the modification of many substrate plasmids is laborious and time-consuming.

Genome editing of host cell chromosomes is an attractive solution method (**ek et al. 2012; Raschmanova et al. 2018; Sakuma and Woltjen 2014; Yang and Blenner 2020) to overcome the difficulties in industrial and natural isolate yeast strains. As a convenient and accurate genome editing methodology, the CRISPR-Cas9 system is popular now. A short target sequence (20 nucleotides) in a single-guide RNA (sgRNA) is sufficient for the CRISPR-Cas9 system to generate a double-strand break (DSB) at the target sequence followed by a protospacer adjacent motif (PAM) (**ek et al. 2012; Ran et al. 2013). In S. cerevisiae strains, the application of CRISPR-Cas9 plasmids containing an sgRNA gene results in the formation of mutations at the target locus. Furthermore, a template DNA that covers the target sequence but has some mutations improves the mutant ratio in DNA editing experiments (DiCarlo et al. 2013; Laughery et al. 2015; Raschmanova et al. 2018). Zhang et al. (2014) reported a Cas9 plasmid harboring a neurothricin-resistance gene and a series of sgRNA-expressing plasmids to target URA3, TRP1, LEU2, and HIS3, respectively. Successive transformation of an industrial S. cerevisiae strain with the Cas9 plasmid and then with one of the sgRNA-expressing plasmids resulted in formation of auxotrophic mutants depending on the guide RNA sequence. Laughery et al. (2015) constructed Cas9 plasmids (pML104 and pML107) containing an sgRNA-expression cassette gene as well as the selection marker gene LEU2 and URA3, respectively. Plasmids simultaneously containing Cas9 and sgRNA genes are categorized as all-in-one type, which is convenient because a single transformation produces mutants.

We aimed to broaden the applicability of the plasmid resources for Saccharomyces sensu stricto complex group. In this study, we focused on LEU2 gene disruption for several reasons. First, LEU2 has been frequently employed as the selection marker gene in yeast plasmid vectors in conventional plasmid resources. Second, among the auxotrophic mutations apt to the conventional yeast vectors, leu2 mutants are the most difficult to achieve in polyploid strains by conventional methods because there is no counter-selective characteristic against chemicals, whereas ura3 mutants are resistant to 5-fluoroorotic acid and trp1 mutants endure the killing action of 5-fluoroanthranilic acid (Akada 2002; Gnugge and Rudolf 2017; Kitamoto et al. 1990). Here, we show all-in-one type CRISPR-Cas9 plasmids with LEU2-gene-targeting guide sequences for application as widely as possible to Saccharomyces sensu stricto species. Dissemination of the Cas9 plasmids to yeast strains belonging to five species in the group produced leu2 mutants at practically desirable frequencies.

Materials and methods

Microbial strains, plasmids, and culture conditions

Table 1 lists bacterial and yeast strains and plasmids used in this study.

Table 1 Bacterial and yeast strains and plasmids used in this study
Full size table

Yeast strains were cultured in YPD medium (1% yeast extract, 2% polypeptone, 2% glucose) at 28 °C. E. coli strains were grown in LB medium (1% Bacto tryptone, 0.5% NaCl, 0.5% Bacto yeast extract) at 37 °C.

Yeast strains were cultured also on synthetic media. Synthetic dextrose (SD) medium consists of 2% glucose and 0.67% Bacto yeast nitrogen base without amino acids. Solid media for yeast were prepared by adding 2% agar. Synthetic galactose solid medium is the same with the solid SD medium but contains galactose in place of glucose. To verify auxotrophic mutant phenotypes in recombinant yeasts, colonies were streaked onto solid SD medium and solid SD medium supplemented with 30 μg/ml leucine.

Plasmid construction

An outline of the plasmid structure is shown in Fig. 2.

Fig. 2
figure 2

Scheme for preparing CRISPR-Cas9 plasmids harboring a 20-bp guide sequence. A Three tool plasmids and processes to construct all-in-one CRISPR-Cas9 plasmids containing a 20-bp guide sequence. The high-copy (ori 2μ) yeast vector pYAMTr2GCas and the low-copy (CEN/ARS) yeast vector pYAMTrGCas lack a guide sequence. The two yeast vectors were linearized by treatment with XbaI and NotI to accept the guide sgRNA expression cassette. The mini-plasmid pBSsgRNA possesses an sgRNA expression cassette lacking a guide sequence. A ds-oligonucleotide DNA containing a 20-bp guide sequence was ligated with BclI/SwaI-digested pBSsgRNA plasmid to place the target guide sequence on the cassette sgRNA expression cassette. The sgRNA expression cassette with the target guide sequence was excised from the guide-sequence-containing mini-plasmids by digestion with PvuII and SphI. Finally, the XbaI/NotI-linearized vector plasmids and the PvuII/SphI fragment harboring the sgRNA expression cassette were introduced into either a yeast strain or an E. coli strain expressing λ Red recombinase. The vector and guide-containing cassette DNAs were fused together in vivo by homologous recombination at a 42-bp overlap** region (left arm) and a 129-bp overlap** region (right arm). B Resultant CRISPR-Cas9 plasmids containing the target guide sequence

Full size image

To build all-in-one type CRISPR-Cas9 plasmids containing the Cas9 gene and a sgRNA expression cassette with a 20-bp guide sequence, we first constructed two vector plasmids as follows. The Cas9 gene and a part of the target-lacking sgRNA expression cassette were excised from pML104 (Laughery et al. 2015) as a 5.1-kbp EcoRI/XbaI DNA fragment. The 5.1-kbp EcoRI/XbaI DNA fragment was ligated with a high-copy-type yeast Agrobacterium-mediated transformation (AMT) vector pYAMTr2G (Kiyokawa et al. 2023) cleaved by EcoRI and XbaI (7.0-kbp) and a low-copy-type yeast AMT vector pYAMTrG (Kiyokawa et al. 2023) linearized by EcoRI and XbaI (6.1-kbp), respectively. The resulting plasmids were named pYAMTr2GCas and pYAMTrGCas, respectively (Fig. 2A).

Second, a mini plasmid harboring the sgRNA expression cassette lacking a target sequence, which was obtained from pML104, was prepared as follows. The target-lacking sgRNA expression cassette was obtained as a 0.9-kbp PvuII/XhoI DNA fragment from pML104. The 0.9-kbp PvuII/XhoI DNA containing the target-lacking sgRNA expression cassette was ligated with EcoRV/XhoI-cleaved pBluescript KS ( +) plasmid DNA (Short et al. 1988) to form a 3.9-kbp plasmid named pBSsgRNA (Fig. 2A).

Three 20-bp guide sequences were designed as described in “Guide sequence for CRISPR-Cas9-mediated genome editing.”

Each ds-oligonucleotide guide sequence was inserted into the sgRNA expression cassette of the pBSsgRNA plasmid as follows: The pBSsgRNA plasmid DNA was prepared using a dam− dcm− E. coli strain JM110 (Yanisch-Perron et al. 1985). The pBSsgRNA plasmid was digested with SwaI overnight at 25 °C and subsequently treated with BclI for 4 h at 50 °C. Each of the three sets of complementary oligonucleotide DNA (Table 3 (A1)) containing the 20-bp guide sequence (Table 2 (A)) was ligated with the 3.8-kbp SwaI/BclI-digested pBSsgRNA. The resulting three plasmids were named pBSsgScLEU2, pBSsgSeLEU2, and pBSsgSc-SeLEU2, respectively.

Insertion of the full-length sgRNA expression cassette from each of the three pBSsgRNA derivatives into the two Cas9 plasmids was performed by in vivo fusion as follows. The Cas9-containing plasmids YAMTr2GCas and pYAMTrGCas were linearized by treatment with XbaI and NotI. The full-length sgRNA expression cassette with a 20-bp guide sequence was taken as a 0.9-kbp PvuII/SphI DNA fragment from pBSsgScLEU2, pBSsgSeLEU2, and pBSsgSc-SeLEU2, respectively. The linearized 12.1-kbp YAMTr2GCas and the 0.9-kbp PvuII/SphI fragment possessing the sgRNA expression cassette were simultaneously introduced into an E. coli strain HB101 RedET expressing λ phage recombinase genes γ, β, and α. In the E. coli cells, the two DNAs were fused in vivo by homologous recombination at 129-bp and 42-bp overlap** sequences between each end of the two DNAs. The three resulting high-copy-type CRISPR-Cas9 plasmids were named pYAMTr2GCsgScLEU2, pYAMTr2GCsgSeLEU2, and pYAMTr2GCsgSc-SeLEU2, respectively.

Construction of low-copy type all-in-one Cas9 plasmids using pYAMTrGCas was performed as described above, but the in vivo fusion in E. coli was replaced with that in yeast strain T55. The resulting low-copy-type CRISPR-Cas9 plasmids were named pYAMTrGCsgScLEU2, pYAMTrGCsgSeLEU2, and pYAMTrGCsgSc-SeLEU2, respectively.

YEp351_PgalGFP was prepared by ligation between SalI- and BamHI-digested YEp351 and 1.5 kbp fragment from SalI- and BglII-digested pFA6a-kanMX6-PGAL1-GFP.

Guide sequence for CRISPR-Cas9-mediated genome editing

We selected 20-bp guide sequences in the target LEU2 gene as follows: Genome sequences of 50 S. cerevisiae strains were retrieved from the Saccharomyces genome database ((Cherry et al. 2012); https://www.yeastgenome.org). To identify common sequences among 50 S. cerevisiae strains in each target gene, multiple sequence alignment was performed using the MAFFT program version 7.463 (Katoh and Standley 2013). Target sequence candidates with low off-target effects were searched using CRISPRdirect (Naito et al. 2015) and CRISPOR (Concordet and Haeussler 2018) programs. Highly conserved regions in the LEU2 gene were used as the query sequence. We examined the applicability of the target sequence candidates to Saccharomyces sensu stricto species using the nucleotide blast program and chose guide sequences (Table 2) to make the applicability as wide as possible.

Table 2 Target sequence in LEU2 gene for CRISPR-Cas9
Full size table

DNA transformation in yeast

The lithium acetate (LiAc) yeast transformation method was performed according to a protocol described by Gietz and Schiestl (2007) with minor modifications. Yeast cells (1.9 × 107 colony forming unit (CFU)) were suspended in a solution containing 100 ng of circular plasmid DNA, LiAc, polyethylene glycol (PEG) 4000, and carrier DNA and incubated for 40 min at 42 °C.

Before selection using G418, the yeast cells were then cultured in liquid YPD medium at 28 °C for 2 h (for experiments using laboratory strains only) or 3.5 h. After the cultivation, yeast cells were collected and spread on solid YPD medium containing 400 μg/ml G418. To enhance the mutant frequency, 50 pmol of a hybridized 90-base oligonucleotide DNA pair (which were named as dsOligo DNAs as listed in Table 3 (B)) was added as the template DNA together with the CRISPR-Cas9 plasmid DNA to the LiAc/PEG 4000/carrier DNA solution.

Table 3 List of oligonucleotides used in this study
Full size table

Transformation frequency was represented as the number of transformants per μg of plasmid DNA per output viable cell number after LiAc/PEG 4000 treatment.

Trans-kingdom conjugation to yeast

A mobilizable LEU2 plasmid pMz1was transferred from a donor E. coli strain HB101 that harbors pMz1 and a helper plasmid pRH220 as shown by Mizuta et al. (2012). After co-incubation between a leu2 mutant yeast and the donor E. coli cells, the cell mixture was spread onto SD solid medium, which allows selective growth of yeast cells that obtained pMz1.

Nucleotide sequence determination

The DNA sequencing was determined using the Sanger method by Eurofin Co. (Tokyo, Japan). Template DNA sequencing reaction was prepared by PCR amplification using primer sets listed in Table 3 (B). In leu2 mutants in aneuploid strains, each of two leu2 genes was amplified using allele-specific primer sets shown in Table 3 (B) and sequenced respectively.

Statistical analysis

Statistical analysis was performed using the R program version 4.0.2 and its expansion package (https://www.R-project.org/). Each datum in a table represents a set of a mean value ± its standard deviation (SD). Individual methods and experimental replicates for statistical comparisons are presented in each table.

Results

Construction of all-in-one CRISPR-Cas9 plasmids for conversion of prototrophic yeast strains to leu2 mutant strains

Figure 2 depicts a scheme to construct all-in-one CRISPR-Cas9 plasmids. Using a Cas9 expression gene and sgRNA expression gene cassette from pML104 (Laughery et al. 2015), we prepared three tool plasmids, namely, a small plasmid pBSsgRNA and two CRISPR-Cas9 vector plasmids (pYAMTr2GCas and pYAMTrGCas), which lacked the target sequence. A 20-bp guide sequence for Cas9 scission was inserted into the sgRNA expression gene cassette in pBSsgRNA, and the cassette with the guide sequence was then moved to one of the two CRISPR-Cas9 vector plasmids. The vector plasmid pYAMTr2GCas is a high-copy and pYAMTrGCas is a low-copy replication type, because the former contains ori_2μ and the latter harbors CEN/ARS (Sikorski and Hieter 1989). Because half of the two vectors were derived from pYAMTrG and pYAMTr2G (Kiyokawa et al. 2023), respectively, they harbored the G418 resistance gene cassette KanMX to enable transformant selection for prototrophic yeast strains and contained a right border sequence (RB), an overdrive sequence, and the pBBR1 replication gene, which as a whole permit mobilization by the AMT method.

Figure 1A illustrates the phylogenetic relationship of yeast species among Saccharomyces sensu stricto group. As listed in Table 2 (A), we designed three 20-bp guide sequences for DNA editing at LEU2 gene, aiming as broadly as possible among Saccharomyces sensu stricto species strains. In DNA databases, LEU2 gene sequence was available for eight species in the group (Table 2 (B)). The three guide sequences matched well with the sequences in the database and were equipped to the two Cas9 vector plasmids, as listed in Table 2 (A).

In this study, we used seven strains belonging to five yeast species. Most of these strains are natural and industrial strains. As a pilot application with a pedestal purpose, we examined the basic behavior of yeast toward the CRISPR-Cas9 plasmids and template DNAs using two S. cerevisiae laboratory strains.

High yield of leu2 mutant strains from S. cerevisiae laboratory strains

The two all-in-one type plasmids pYAMTrGCsgScLEU2 and pYAMTr2GCsgScLEU2 contain a 20-bp guide sequence for S. cerevisiae LEU2 gene (Table 2 (A)). The former plasmid is of the multi-copy type and the latter is of the low-copy type. Introduction of the two plasmids into the haploid laboratory strain T55 generated prototrophic (Leu+) colonies and leucine-auxotrophic (Leu−) colonies. As shown in Table 4, the ratio of Leu− colonies was approximately 90%, irrespective of the plasmid copy difference. The diploid laboratory strain T556 produced fewer recombinant colonies than the haploid strain. The ratio of Leu− colonies by the diploid strain was around 50%, which is a half the ratio of the haploid strain.

Table 4 LEU2 gene disruption in haploid and diploid S. cerevisiae laboratory strains
Full size table

We prepared ds-oligonucleotide DNA (Sc), which is a hybridization of two complementary 90-base sequences that match the target region of ScLEU2 but lacks two nucleotides in the PAM sequence (Table 3 (A2)). Addition of the ds-oligo DNA markedly increased the number of recombinant colonies by more than tenfold, and the Leu− ratio was up to 98% in the haploid strain and approximately 95% in the diploid strain (Table 4).

The mutation(s) that caused the Leu− phenotype were located at the LEU2 locus. First, leucine auxotrophy of the Leu− mutant recombinant colonies was rescued by LiAc transformation with YEp351 (Fig. 3A) and by trans-kingdom conjugation with a donor E. coli HB101 harboring a mobilizable ScLEU2 plasmid pMz1 and a helper plasmid pRH220 (Fig. 3B). Second, crossing haploid Leu− colonies with an opposite mating-type leu2 mutant tester strain IS289-1C failed to complement the Leu− phenotype, while crossing with an opposite mating-type leu1 mutant tester strain HD119-3 complemented the Leu− phenotype (data not shown). Finally, DNA sequencing analysis verified that the mutation is present at the target sequence in LEU2 gene as shown in Fig. 4A. Three haploid and three diploid mutants had a sequence identical to the template ds-oligonucleotide DNA sequence.

Fig. 3
figure 3

Rescue of Leu− phenotype by introduction of ScLEU2 gene. A LiAc transformation with YEp351, which harbors ScLEU2. B Trans-kingdom conjugation using a donor E. coli strain HB101 containing a helper plasmid pRH220 and a ScLEU2-containing mobilizable plasmid pMz1

Full size image
Fig. 4
figure 4

Verification of leu2 mutations by nucleotide sequence analysis. A template 90-bp oligonucleotide DNA sequence was aligned with the corresponding wild-type (WT) and mutant gene sequences. Shade boxes and underlines indicate the 20-bp guide sequence and PAM, respectively. Arrows and filled triangles indicate nucleotide polymorphisms among leu2 mutant strains and between wild-type and template DNA sequences, respectively. A Four mutants derived from S. cerevisiae strain T55 and four mutants from the diploid strain T556. B Three mutants of the sake-producing strain S. cerevisiae Kyokai No. 7 (K7). C Three mutants of S. paradoxus strain CBS432T. D Three mutants of S. bayanus strain CBS380T. E Two mutants from S. kudriavzevii strain CBS8840T. F Three mutants of the beer-producing S. pastorianus strain W34/70

Full size image

Generation of leu2 mutants from a sake (rice-wine) brewing strain and a S. paradoxus strain

We applied the high-copy plasmid pYAMTr2sgScLEU2 and the template ds-oligonucleotide DNA (Sc) to five yeast strains, which consisted of industrial and natural isolate strains.

Sake-producing S. cerevisiae strain Kyokai No. 7 (K7) is a diploid (Akao et al. 2011). The strain showed a much lower Leu− mutant ratio (19%) than the diploid laboratory strain T556 (100%). Conversely, K7 strain yielded a sevenfold larger number of G418R recombinant colonies than T556 (Table 5). We examined three Leu− mutants of the K7 strain using DNA sequencing analysis at the LEU2 locus. All three K7 mutants exhibited a sequence identical to the template ds-oligonucleotide DNA sequence (Sc) (Fig. 4B).

Table 5 Leu− mutant formation induced by CRISPR-Cas9 plasmids in yeasts belonging to five Saccharomyces sensu stricto species
Full size table

S. paradoxus CBS432T is a diploid strain. As shown in Fig. 2B, CBS432T strain’s LEU2 gene has 20-bp guide and PAM sequences identical to those of the ScLEU2 gene, whereas there were single nucleotide polymorphisms (SNPs) near the target sequence in S. paradoxus LEU2 compared with ScLEU2 (data not shown). As indicated in Table 5, the CBS432T strain showed a high Leu− mutant ratio (93%). We analyzed the LEU2 sequence in three CBS432T Leu− mutant strains. All three mutants contained one nucleotide substitution in the target sequence and lacked PAM, which appears to reflect an event of replacement by double crossovers with the template ds-oligonucleotide DNA (Sc) (Fig. 4C). In contrast, the three CBS432T leu2 mutant strains showed small differences from the template DNA in sequences surrounding the target sequence, which is likely to show crossover positions near the scission site between the template ds-oligonucleotide DNA and the chromosomal LEU2 gene in the S. paradoxus strain (Fig. 4C).

In contrast to the S cerevisiae and S. paradoxus strains, application of the same plasmid and the template ds-oligonucleotide DNA produced no auxotrophic mutant in S. kudriavzevii CBS8840T, S. bayanus CBS380T, and S. pastorianus W34/70 (Table 5). This result for S. bayanus and S. pastorianus coincides with the estimation based on their LEU2 gene sequences in the database (Table 2 (B)), whereas no data were available for S. kudriavzevii.

Formation of leu2 mutants from S. bayanus using a guide sequence suitable for S. eubayanus-type LEU2

As shown in Fig. 1A, S. bayanus CBS380T is an allodiploid strain whose hybrid genome originated from S. eubayanus and S. uvarum, whereas S. pastorianus W34/70 is an allodiploid strain with an amalgam genome derived from S. eubayanus and S. cerevisiae (Borneman and Pretorius 2015; de la Cerda Garcia-Caro et al. 2022). The all-in-one CRISPR-Cas9 plasmid pYAMTr2GCsgSeLEU2 contains a guide sequence (SeLEU2) suitable for S. eubayanus LEU2 gene as the target.

The S. bayanus strain CBS380T was transformed simultaneously with pYAMTr2GCsgSeLEU2 and a template ds-oligonucleotide DNA (Se), which is similar to the template ds-oligonucleotide DNA (Sc) but suits S. eubayanus LEU2 (Fig. 4D). As indicated in Table 5, the CBS380T strain produced Leu− colonies upon the transformation. Its Leu− ratio was 100%. DNA sequencing analysis verified that three S. bayanus Leu− mutants had a sequence identical to the ds-oligonucleotide DNA (Se) sequence in the target LEU2 locus (Fig. 4D).

Conversely to S. bayanus, and accordingly with the estimation shown in Table 2 (B), S. pastorianus, S. kudriavzevii, and S. cerevisiae strains failed to generate any Leu− mutants by transformation with pYAMTr2GCsgSeLEU2 and its template (Table 5).

Conversion of S. pastorianus and S. kudriavzevii strains to leu2 mutants using a guide sequence common to S. cerevisiae-type and S. eubayanus-type LEU2

The all-in-one CRISPR-Cas9 plasmid pYAMTr2GCsgSc-SeLEU2 contains a 20-bp guide sequence (Sc-Se LEU2 in Table 2), which is highly conserved between Sc-type and Se-type LEU2.

The S. pastorianus strain W34/70 was transformed with a mixture of pYAMTr2GCsgSc-SeLEU2 and its corresponding template ds-oligonucleotide DNA (Sc-Se). The resulting recombinant W34/70 colonies exhibited a Leu− mutant ratio 17% (Table 5). The same mixture was also applied to S. kudriavzevii strain CBS8840Tas well as to S. cerevisiae T556 and S. bayanus CBS380T strains. This treatment generated Leu− colonies. The Leu− ratio among recombinant colonies of CBS8840T was 7%, whereas that of T556 and CBS380T was 100% (Table 5).

DNA sequencing analyses revealed that all Leu− mutants from W34/70 and CBS8840T strains had a four-nucleotide deletion surrounding PAM, which is likely due to a double-crossover with the template DNA at the target LEU2 locus (Fig. 4E, F). Leu− mutants from S. pastorianus strain W34/70 had some nucleotide polymorphisms, which were probably a consequence of homologous recombination between chromosomal Sc-type and Se-type LEU2 alleles at the target locus (Fig. 4F). One CBS8840T Leu− mutant strain (#Tce39) contained an additional four-nucleotide deletion in the target sequence (Fig. 4E). These results support our assumption that the pYAMTr2GCsgSc-SeLEU2 plasmid, which has the guide sequence Sc-SeLEU2 most common to both Sc-type and Se-type LEU2 genes, is applicable to a broad range of Saccharomyces sensu stricto yeasts to generate leu2 mutant strains.

Easy removal of all-in-one CRISPR-Cas9 plasmid from recombinant leu2 mutants

After DNA editing by plasmid transformation, the plasmid DNA needs to be removed from the resultant mutant transformants for subsequent studies. We analyzed the loss frequency of a resident CRISPR-Cas9 plasmid from recombinant leu2 mutants. First, we cultured leu2 mutant strains induced by pYAMTr2GCsgScLEU2 derived from S. cerevisiae T556 and S. paradoxus CBS432T in liquid YPD medium (without G418) for 24 h and then established colonies on solid YPD medium. Sensitivity to G418 was examined for each colony. All 40 T556- and CBS432T-derived colonies were sensitive to G418, indicating that the resident CRISPR-Cas9 plasmid is easily removable.

One leu2 mutant from each of the seven strains was treated as described above. Consequently, G418-sennsitive colonies were obtained easily. PCR analysis further confirmed the absence of the Cas9 plasmid in a G418S leu2 mutant colony of the respective strains. The leu2 mutant strains free of the Cas9 plasmid are available in two public depositories (Table 1 B).

Expression of a foreign gene by transformation of resulting leu2 mutants

The leu2 mutants were transformed with a LEU2 plasmid YEp351GFP, which contains a green fluorescent protein gene that is located just downstream of the promoter of ScGAL1. As shown in Fig. 5, leu2 mutants derived from S. cerevisiae T55 and T556, S. paradoxus CBS432T, S. pastorianus W34/70 emitted green fluorescence when transformed with YEp351GFP and grew on the galactose medium, while those with the vacant vector plasmid YEp351 did not indicate the fluorescence. Mutants from S. cerevisiae K7 and S. bayanus CBS380T did not emit the fluorescence. The S. kudriavzevii mutants as well as S. kudriavzevii CBS8840T did not grow on the galactose medium. These results suggest that the S. paradoxus and S. pastorianus strains possess a regulatory mechanism similar with that of the S. cerevisiae galactose regulon, while the S. cerevisiae K7 and S. bayanus CBS380T strains have a different mechanism.

Fig. 5
figure 5

Introduction of a green fluorescent protein gene in yeast leu2 mutants. A Yeast leu2 mutants transformed with YEp351 and its derivative plasmid (YEp351GFP) containing a PGAL1::gfp gene were grown on synthetic galactose solid medium. Yeast colonies were irradiated with 365 nm LED light. B The transformed yeast cells harboring YEp351_PGAL1GFP were observed under blue light excitation

Full size image

Applicability prediction of the all-in-one plasmids to other yeast species

In this study, we did not handle strains belonging to five “pure species,” namely, S. cariocanus, S. mikatae, S. arboricolus, S. eubayanus, and S. uvarum. However, it is easy to predict that S. eubayanus and S. uvarum are good substrates for LEU2 gene disruption by the plasmids containing the S. eubayanus type guide (Se), because 100% Leu− ratio was exhibited (Table 5) by the allodiploid S. bayanus strain CBS380T, whose genome was derived from S. eubayanus and S. uvarum (Borneman and Pretorius 2015). Furthermore, the DNA database supplied LEU2 gene sequences for S. eubayanus and S. uvarum (Table 2 (B)). The sequence data indicated a perfect match to the guide sequence Se-SeLEU2 (Table 2 (B)). Similarly, the LEU2 sequences of S. mikatae and S. arboricolus suited exactly to the guide sequence SeLEU2. These data forecast the suitability of the Cas9 plasmids for the five remaining species.

Discussion

In this study, we prepared leu2 mutants from seven prototrophic yeast strains belonging to five species of the Saccharomyces sensu stricto group by DNA editing. Leu− mutant ratios varied among the yeast strains. Three diploid strains, namely, one S. cerevisiae laboratory strain T556 and the type strains of S. paradoxus and S. bayanus, achieved an almost 100% Leu− mutant ratio (Table 5). Contrary to the high efficacy of the three diploid strains, S. cerevisiae Kyokai No. 7 (K7), S. kudriavzevii CBS8840T, and S. pastorianus W34/70 exhibited lower Leu− ratios, 9–19% (Table 5). Despite the unexpectedly low ratio level in the three strains, it was easy to find the target leu2 mutants because the discrimination between Leu+ and Leu− was a facile work for the number of colonies, and all of the Leu− colonies examined by sequence analysis were leu2 mutants (Fig. 4), irrespective of the ratio levels. The guide sequences in the Cas9 plasmids mostly fit the strains (Table 2 (B), Fig. 4). However, the template DNA sequences differed by several nucleotides in some strains (Fig. 4). Therefore, the ratio could be increased by supplying a template DNA that perfectly matches the sequence around the guide sequence for each strain. We discuss an additional scenario due to the presence of the case where the template DNA sequence was not the issue of the template DNA sequence.

Feasible reason for the cases of unexpectedly low Leu− mutant ratio

In the DNA databases, no LEU2 sequence was available for S. kudriavzevii CBS8840T, when we started this study. Sequence analysis of CBS8840T strain’s LEU2 revealed one nucleotide difference at 5’ end region of the target 20 bp compared with the common guide sequence for Sc-SeLEU2 (Fig. 4E). The single-nucleotide difference at the site may have caused the low ratio, whereas one nucleotide substitution near the 3’-end of the 20 bp did not matter (Fig. 4A–D). The allodiploid strain S. pastorianus W34/70 harbors Sc-type and Se-type LEU2 genes, similar to the strain CBS1513T (Table 2 (B)). The guide sequence type Sc-SeLEU2 in the Cas9 plasmid was identical to the corresponding regions in the two LEU2 genes in W34/70 (Fig. 4F) and CBS1513T genomes (Table 2 (B)). One possible reason for the low Leu− ratio is that an imaginary higher copy number of the genomic LEU2 gene caused the lower Leu− mutant ratio. In fact, de la Cerda Garcia-Caro et al. (2022) reported that the S. pastorianus W34/70 strain has multiple copies of ScLEU2 and SeLEU2. It is feasible that the higher copy number of the LEU2 gene in the genome caused the lower Leu− mutant ratio.

In this study, we also predicted applicability of the LEU2-targeting plasmids and the template DNAs to other Saccharomyces sensu stricto species, namely, S. cariocanus, S. mikatae, S. arboricolus, S. eubayanus, and S. uvarum.

In conclusion, the series of CRISPR-Cas9 plasmids constructed in this study enabled the generation of leu2 mutant strains from prototrophic strains of Saccharomyces sensu stricto species, including natural isolate, industrial, and allodiploid strains.