Plasmid Construction Using Recombination Activity in the Fission Yeast Schizosaccharomyces pombe

Background Construction of plasmids is crucial in modern genetic manipulation. As of now, the common method for constructing plasmids is to digest specific DNA sequences with restriction enzymes and to ligate the resulting DNA fragments with DNA ligase. Another potent method to construct plasmids, known as gap-repair cloning (GRC), is commonly used in the budding yeast Saccharomyces cerevisiae. GRC makes use of the homologous recombination activity that occurs within the yeast cells. Due to its flexible design and efficiency, GRC has been frequently used for constructing plasmids with complex structures as well as genome-wide plasmid collections. Although there have been reports indicating GRC feasibility in the fission yeast Schizosaccharomyces pombe, this species is not commonly used for GRC as systematic studies of reporting GRC efficiency in S. pombe have not been performed till date. Methodology/Principal Findings We investigated GRC efficiency in S. pombe in this study. We first showed that GRC was feasible in S. pombe by constructing a plasmid that contained the LEU2 auxotrophic marker gene in vivo and showed sufficient efficiency with short homology sequences (>25 bp). No preference was shown for the sequence length from the cut site in the vector plasmid. We next showed that plasmids could be constructed in a proper way using 3 DNA fragments with 70% efficiency without any specific selections being made. The GRC efficiency with 3 DNA fragments was dramatically increased >95% in lig4Δ mutant cell, where non-homologous end joining is deficient. Following this approach, we successfully constructed plasmid vectors with leu1+, ade6+, his5+, and lys1+ markers with the low-copy stable plasmid pDblet as a backbone by applying GRC in S. pombe. Conclusions/Significance We concluded that GRC was sufficiently feasible in S. pombe for genome-wide gene functional analysis as well as for regular plasmid construction. Plasmids with different markers constructed in this research are available from NBRP-yeast (http://yeast.lab.nig.ac.jp/).


Introduction
Construction of plasmids is crucial in modern molecular biology.In many cases, plasmids are constructed in vitro by digesting (cutting) DNA fragments with restriction enzymes at specific sites (restriction sites) and then ligating (joining) the resulting fragments.The constructed DNA is usually amplified in E. coli to analyze its structure.However, a second cloning method using homologous recombination activity (often designated gap-repair cloning or GRC) allows for more design flexibility during construction as restriction sites are not used.The basic GRC procedure is shown in Figure 1.In the budding yeast Saccharomyces cerevisiae, efficient GRC is observed due to the high homologous recombination activity of this organism [1,2], which has been used to construct large-scale systematic plasmid collections [3][4][5].In these cases, the target genes are directly cloned into S. cerevisiae cells to study their function.Thus, GRC is very effective for rapid analyses of gene function.
The fission yeast Schizosaccharomyces pombe is also widely recognized as a model eukaryote [6].Although a previous study has reported GRC feasibility in S. pombe [7], it has not been commonly used because there have been no systematic studies reporting its efficiency in this yeast.If GRC in S. pombe is as efficient as it is in S. cerevisiae, analyses of gene function using plasmids in S. pombe could be performed on a wider scale.Therefore, we studied GRC efficiency in S. pombe and first demonstrated amplification of DNA fragments with short homology sequences (.25 bp) by PCR followed by successfully cloning into a S. pombe plasmid.We then demonstrated that plasmid construction using 3 DNA fragments could be achieved in 70% efficiency without specific selection for recombination.Using lig4D mutant cell as a host strain, where the non-homologous end joining (NHEJ) activity is deficient, the efficiency was increased .95%.Thereafter, we constructed plasmids in S. pombe with different auxotrophic markers using GRC.

Cloning the LEU2 Marker Gene Using GRC
We first evaluated the number of GRC events that could be observed in S. pombe and then constructed a plasmid in it using GRC.We used the plasmid pDblet as a cloning vector because it is stable as a monomer and transforms S. pombe with high efficiency.It also shows high mitotic stability [8].These features are advantageous when the constructed plasmid is recovered from the S. pombe cells to verify its structure.As inserts, we used DNA fragments containing S. cerevisiae LEU2 that complemented S. pombe leu1 so that transformants harboring the plasmids constructed by successful GRC (i.e., pDblet with the LEU2 gene) could be easily distinguished by replica plating.We also tested the effect of (1) the length of the homologous region (,25 bp or ,40 bp) and (2) the distance of the homologous region from the fragment end (Figure 2A).
The fission yeast strain Sp286h+ was transformed with a circular vector, linearized (KpnI-HindIII digested) vector, and with 3 combinations of linearized plasmids with inserts as shown in Figure 2A.Transformants were first selected by the auxotrophic pDblet (uracil) marker in order to estimate transformation efficiency.As shown in Figure 2B, the linearized plasmids resulted in a lower number of transformants than the circular or the linearized plasmid with inserts.We then tested whether these transformants were leucine prototrophs by replica-plating the transformants onto -leucine plates as this would provide an indirect evidence of successful GRC.Almost all of the transformants from the 3 groups, where the linearized plasmids with inserts were combined, showed leucine synthesizing ability (Figure 2C).
These results suggest that GRC can be successfully performed in S. pombe.Because the GRC efficiencies were indistinguishable between the 3 groups where the linearized plasmids with inserts were combined, it was clear that a 25-bp homologous region sufficed for GRC and that the recombination site did not have to be at the end of the fragment.We also constructed linearized vectors by digestion with XhoI to check the effect of the restriction sites used.In this case, the number of uracil and leucine-prototroph transformants observed was similar to that observed with the HindIII-KpnI double-digested vector (data not shown), suggesting that no preferences in restriction site (for single or double restriction enzyme digestions) were observed for GRC in S. pombe.
We next examined the structure of the plasmids within transformants grown on -leucine plates by randomly recovering plasmids and digesting them with restriction enzymes and found that .70% of plasmids recovered from the transformants had a desired structure (Table 1).Plasmids without the desired structure appeared to have the LEU2 gene in the pDblet cloning site but had some mutations in the regions (data not shown).These results indicate that GRC is a useful cloning method in S. pombe.

Construction of Plasmids Containing nmt1 Promoter: EGFP Using GRC
We then investigated whether GRC seen in S. pombe could be applied in more advanced plasmid constructions.The blueprint of the plasmid construction is shown in Figure 3A.In this process, 3 DNA fragments containing the promoter of the nmt1 gene, EGFP gene, and linearized pDblet are connected.EGFP is expressed under the control of the nmt1 promoter when the plasmid is constructed correctly.As shown in Figure 3B, most of the transformants with the linearized plasmids with inserts showed GFP fluorescence; 7 out of 10 randomly selected transformants showed GFP fluorescence (Figure 3C).One in 10 transformants (Figure 3CB) showed weak GFP fluorescence.
All plasmids recovered from the 7 transformants showing GFP fluorescence were verified to have the desired structure (Table 2).We could not recover any plasmids from 2 of the 3 transformants showing no or weak fluorescence (Table 2) indicating that the DNA fragments were integrated into the genome in these transformants.The plasmid recovered from the last transformant that did not fluoresce had a truncated insert.To check if the nmt1 promoter and EGFP were joined correctly, we sequenced the connecting region; 2 out of 7 plasmids with the desired inserts had nucleotide substitutions within the connecting region (the primer region for PCR) (Table 2).Currently we are not sure about the reason of these nucleotide substitutions.

The Efficiency of GRC Is Dramatically Increased in a Mutant Cell without NHEJ Activity
It is reported that the efficiency of homologous recombination event can be increased in a yeast strain without NHEJ activity [9].We thus tested the GRC efficiency in a lig4D mutant strain where non-homologous end joining activity is deficient [10].We performed the same experiment as described in Figure 3 (i.e.DNA fragments containing the promoter of the nmt1 gene, EGFP gene, and linearized pDblet are connected).As shown in Figure 4, 96% (45 out of 46) transformants showed significant GFP fluorescence that was regulated by thiamine, indicating the GRC efficiency was dramatically increased in this strain.In the rest one transformant (arrowhead in Figure 4), GFP was probably integrated into genome because any plasmid could be recovered from the transformant.We recovered the plasmids from the transformants and checked the structure with restriction digestion and sequencing.All recovered plasmid had correct structure, but 3 out of 10 plasmids had nucleotide substitutions within the connecting region (the primer region for PCR).Currently we are not sure about the reason of these nucleotide substitutions.
These results indicate that GRC could be effectively performed in S. pombe (with .70%efficiency in wild type, and with .95% in Figure 2. Cloning the LEU2 marker gene using GRC in S. pombe.(A) Blueprint of the plasmid construction.DNA fragments containing LEU2 genes were amplified by PCR using the indicated primer sets with pRS315 as a template.Homologous regions between the vector plasmid (pDblet) and the insert fragments are shown in the same colors.(B) Result of transformation with the linearized (HindIII-KpnI digested) vector, circular vector, and linearized vector with inserts as shown in A. Transformants of Sp286h+ were selected on -uracil plates (EMM with leucine and adenine).(C) A replica of the transformants on the plate in B on a -leucine plate (EMM with adenine).doi:10.1371/journal.pone.0009652.g002Table 1.Structural verification of plasmids found in the transformants after pDblet + LEU2 construction (Figure 2C). a NHEJ mutant) with 3 DNA fragments without selection for recombination, although constructed plasmids sometimes had nucleotide substitutions within the connecting regions.We thus concluded that GRC is suitable for use as a cloning method in S. pombe for genome-wide gene functional analysis as well as for regular plasmid construction.

Plasmids Used in this Study
pDblet [8] was used as a plasmid vector in the GRC experiment.pRS315 [11] was used as the PCR template for the S. cerevisiae LEU2 gene.pHM004_nmt1-1, a plasmid containing nmt1 promoter (from our laboratory stock), was used as the PCR template for the nmt1 promoter.pKT128 [12] was used as the PCR template for EGFP.

Growth Media and Conditions
S. pombe cells were cultured in a manner similar to [13] at 30uC.EMM medium and plates were prepared using EMM broth, powder (MP Biomedicals).

DNA Methods
High-efficiency transformation of yeast, yeast DNA isolations (Yeast DNA miniprep), and DNA miniprep from E. coli were  performed in a manner similar to [14] by replacing the culture medium with one more appropriate for S. pombe.E. coli transformation was performed using a XL1-Blue Electroporation Competent Cell (Agilent Technologies).DNA sequencing was performed using an ABI 3730xl Analyzer.

Fluorescence Imaging
Images of yeast culture plates with GFP fluorescence were obtained using a LAS4000 lumino-image analyzer (Fujifilm) and were processed using Photoshop CS2 software (Adobe).
Preparation of DNA Fragments for GRC in S. pombe Preparation of the plasmid vector with restriction enzyme digestion: About 10-mg pDblet prepared with DNA miniprep from E. coli was digested with the mentioned restriction enzymes in a 100-mL solution.A total of 2-mL (,0.2 mg) digested plasmid solution was used directly for yeast transformation.Preparation of DNA fragments by PCR: DNA fragments were amplified using KOD polymerase (Toyobo) according to the manufacturer's protocol in 50-mL solution using the primers listed in the Tables S1 along with the template DNA described previously.A total of 45 mL of the amplified DNA solution (containing about 10-mg DNA) was used directly for yeast transformation.The molar ratio of the digested vector and PCR products was about 1:200,300.

Supporting Information
Tables S1 This is a PDF file containing lists of primers used in this study.Found at: doi:10.1371/journal.pone.0009652.s001(0.07 MB PDF) Text S1 This is a text file containing the nucleotide sequence of pKH001.

Figure 1 .
Figure 1.Basic gap-repair cloning procedure.DNA fragment(s) from a linearized plasmid vector are prepared by restriction enzyme digestion or by PCR.DNA fragment(s) from the insert are amplified by PCR, ensuring that the sequence at the end of the fragment is homologous with that of the plasmid.Both DNA fragments are then simultaneously introduced into the cell.Transformants are selected by identifying the plasmid vector marker.The fragments are connected by the homologous recombination activity that occurs within the cell.The constructed plasmid is recovered from the transformed cell into E. coli to amplify and check the structure of the plasmid with restriction enzyme digestion and DNA sequencing.Cells with plasmids that have the desired structure are used for functional analysis.doi:10.1371/journal.pone.0009652.g001 InsertNo. of positive clones/tested Positive clone (%) .1371/journal.pone.0009652.t001

Figure 3 .
Figure 3. Construction of a plasmid containing nmt1 promoter: EGFP using GRC in S. pombe.(A) Blueprint of the plasmid construction.DNA fragments containing nmt1 promoter and EGFP were amplified by PCR using the indicated primer sets with pHM004_nmt1-1 and pKT128 as templates, respectively.Homologous regions between the vector plasmid (pDblet) and the insert fragments are shown in the same colors.(B) Result of transformation.Transformants of Sp286h+ were selected on -uracil plates (EMM with leucine and adenine).The fluorescent GFP image was superimposed on a bright-field image.(C) Independent colonies from plate B were streaked on -uracil plates.The resulting image is shown in B. doi:10.1371/journal.pone.0009652.g003

Figure 4 .Figure 5 .
Figure 4. Construction of a plasmid containing nmt1 promoter: EGFP using GRC in the lig4D mutant S. pombe.GRC was performed as in Figure3except FY4102 was used as a host strain.Transformants were randomly selected and streaked on -uracil +thiamine plate, then the transformants were replica-plated on +thiamine (A) and -thiamine (B) plates.Negative controls (transformants with the vector) were also streaked.Arrowheads indicate transformant with weak GFP fluorescence, in which GFP was integrated into genome.doi:10.1371/journal.pone.0009652.g004