Identification of new players in cell division, DNA damage response, and morphogenesis through construction of Schizosaccharomyces pombe deletion strains.

Many fundamental biological processes are studied using the fission yeast, Schizosaccharomyces pombe. Here we report the construction of a set of 281 haploid gene deletion strains covering many previously uncharacterized genes. This collection of strains was tested for growth under a variety of different stress conditions. We identified new genes involved in DNA metabolism, completion of the cell cycle, and morphogenesis. This subset of nonessential gene deletions will add to the toolkits available for the study of biological processes in S. pombe.

Large-scale genetic screening is a particularly effective means to establish a wiring diagram of functional relationships for genes of unknown or redundant function (Costanzo ). This approach was pioneered in the budding yeast, Saccharomyces cerevisiae (Dixon ), and has also been used successfully in the fission yeast, Schizosaccharomyces pombe, to identify genes involved in cell cycle control (Navarro and Nurse 2012; Hayles ), cell size (Kelly and Nurse 2011), sporulation (Ucisik-Akkaya ), TORC1 inhibition (Rallis ), valproic acid sensitivity (Zhang ), cobalt homeostasis (Ryuko ), resistance to TORC1 inhibition (Rallis ), cadmium tolerance (Kennedy ), micafungin sensitivity (Zhou ), the DNA damage response (Pan ), chromosome cohesion (Chen ), and other processes.

The majority of the aforementioned S. pombe genetic screens used a haploid strain collection containing 3000–3300 protein coding gene deletions constructed by and obtained from Bioneer Corporation (Republic of Korea). This valuable collection represents a significant portion of the known 3576 nonessential protein-coding genes (Kim ). At the time this ambitious deletion project began, it was estimated that 4914 protein-coding genes existed in the S. pombe genome. Since then, additional protein-coding genes have been identified or predicted (Hayles ; Bitton ), and there are currently 4981 protein-coding genes annotated for S. pombe (http://www.pombase.org/) (Wood ). It follows, then, that large-scale genetic screens performed thus far did not include all possible viable S. pombe gene deletions.

To fill this gap in available gene deletion strains, we constructed 281 haploid gene deletion mutants that were not a part of the approximately 3000 gene deletion strains available as version 3 from Bioneer Corporation. We conducted growth assays for all these deletions under different environmental stresses and report new players in the processes of DNA metabolism, cell division, and morphogenesis.

Materials and Methods

Yeast strains, media, and materials

The parent strain used for constructing the gene deletions was ade6-M210 ura4-D18 leu1-32 h+ (KGY247). Deletions containing rad22-GFP::kan were made by crossing the deletion strains with a lab stock strain rad22-GFP::kan- (KLG16715) and performing tetrad analysis. All chemicals were from Sigma-Aldrich (St. Louis, MO) unless otherwise noted.

Gene deletions

Gene deletions were accomplished as previously published with some modifications as described below (Wach ). Primers used to amplify the flanking sequences were designed as shown in Figure 1 and listed in Supporting Information, Table S1 and Table S2. Initially the 5F, 5R, 3F, and 3R 24-bp sequences were extracted from the following positions: 5F, −324 to −300 upstream of the ORF ATG; 5R, −24 to −1 upstream of the ORF ATG (starting with the kan start 23 bp in addition); 3F, +1 of the STOP codon to +24 (starting the Kan end 23 bp in addition); and 3R, +324 to + 300 from the ORF STOP codon. Checking primers 5′chk and 3′chk were designed within 500 bp on either side of the ORF, and ORF chk oligo within 100 bp of the ORF. Next, the GC percentages of all primers were examined. When they were found to be lower than 30%, the 5F and 3R sequences and the three checking sequences were optimized with the software Primer3 (Primer3web version 4.0.0; http://primer3.ut.ee/) (Untergasser ; Koressaar and Remm 2007) by re-selecting a sequence (between 24 and 35 bp) further upstream or downstream of the original site with at least 30% GC content. In these cases, the length of the flanking sequences changed from the targeted 300 bp to range from 200 to 500 bp and the checking oligonucleotide was also re-designed accordingly to make sure it was outside of the deletion construct. Because the position of the 5R and 3F sequences cannot be changed, oligonucleotide sequences with low GC % were extended up to a length of 35 bp to increase the melting temperature.

Figure 1

Schematic of gene deletion strategy. (A) The diagram shows the position of the gene specific oligonucleotides used in PCR reactions to make and check the deletions. 5F/5R and 3F/3R were used for amplifying the flanking sequences of the ORF to be deleted. 5′ chk, 3′ chk, and ORF chk were used to validate the successful insertion of the kanMX6 cassette and deletion of the ORF. (B) Schematic of the process of making the DNA fragment used for deletion of the ORF by homologous recombination. The first round of PCRs produced the 5′ and 3′ flanking sequences of the ORF. The second PCR reactions were performed using the first round PCR products as primers and the kanMX6 cassette as template to amplify the cassette with gene-specific flanks. The resulting PCR product was used for cell transformation and homologous recombination to complete gene deletion. The positions of the common checking oligonucleotides within the kanMX6 cassette are shown.

Oligonucleotides in 96-well plates were re-suspended in 1 mM Tris, pH 7.4, 0.1 mM EDTA at a final concentration of 100 µM. They were then diluted to 4 µM with nuclease free water in a new 96 well plate; 5 µl each of F and R oligonucleotides were mixed with 25 µl GoTaq Green Master Mix (Promega, Fitchburg, WI) and 225 ng S. pombe genomic DNA (KGY246; ade6-M210 ura4-D18 leu1-32 h and nuclease free water to a final volume of 50 µl. The PCR was performed using a T3000 Thermocycler (Biometra) with the following program: 95°, 2 min; 12 cycles of (95° for 1 min, 60° for 1 min, and 72° for 1 min); and 25 cycles of (95° for 1 min, 50° for 1 min, and 72° for 1 min); 72° for 10 min. Then, 5 µl of the PCR product was examined on 0.8% agarose gel. If the product was weak or undetectable, then the PCR was re-run with the annealing temperatures reduced to 53° and 48° in the above two-step program.

PCR of the kanMX6 cassette

The second-stage PCR reactions included the following in a total volume of 50 µl: 1 µl each of the first-stage 5′ and 3′ flanking sequence PCR products, 100 ng pFA6-kanMX6 (Bahler ), 5 µl 10× Easy A buffer, 4 µl dNTPs (stock is 2.5 mM each), and 0.5 µl (2.5 U) Easy A Hi-Fi polymerase (Agilent Technologies, Santa Clara, CA). The following PCR program was used: 94°, 1 min; 30 cycles of (94° for 1 min, 45° for 5 min, and 72° for 3 min); 72° for 10 min. Then, 5 µl of PCR product was checked on 0.8% agarose gel prior to transformation.

Yeast transformations

S. pombe strain leu1-32 ura4-D18 ade6-M210 h was grown in YE media at 32° overnight to mid-log phase. Ten OD of the culture was used for each transformation. Cells were pelleted by centrifugation and washed with 10 ml of TE/LiAc (10 mM Tris, pH 7.6, 1 mM EDTA, 100 mM lithium acetate) twice and then re-suspended in TE/LiAc. For each transformation, 100 µl cells, corresponding to the original 10 OD of culture, were mixed and incubated with 20 µl transforming DNA (the second PCR product) and 40 µg boiled salmon sperm DNA. After incubation at room temperature for 10 min, 260 µl of TE/LiAc/PEG (40% PEG in TE/LiAc) was added, and the cells were vortexed gently and incubated at 32° for 30–45 min; 43 µl of DMSO was then added to each tube, followed by incubation at 42° for 6 min. Cells were centrifuged and washed with 0.5 ml sterile H2O before being re-suspended in 5 ml of YE media and incubated at 29° overnight. Then, 500 µl of cells were plated on YE-G418 plates and incubated at 29° for 3–5 d. Each plate was replica plated three times to fresh YE-G418 plates before colonies were checked by PCR.

Whole cell PCR checking of yeast transformants

Whole cell PCR was used to check and select a colony with correct deletion. For each PCR reaction, 5 µl of 4 µM of both forward and reverse oligos were mixed with 10 µl GoTaq green. The following PCR program was used: 95°, 1 min; 12 cycles of (95° for 45 sec, 60° for 1 min, and 72° for 1 min); 25 cycles of (95° for 45 sec, 52° for 45 sec, and 72° for 1 min). Then, 10 µl of each PCR product was checked by running it on an 0.8% agarose gel.

Sensitivity/resistance assay

Wild-type and deletion strains were grown in YE in 96-well plates to saturation. Ten-fold serial dilutions of each strain were made and spotted onto minimal medium agar plates (supplemented with adenine, uracil, and leucine) or YE agar plates in the absence or presence of the following additions: brefeldin A, 25 µM; bleomycin, 1 µg/ml; calcofluor white, 0.5 mg/ml; cycloheximide, 10 µg/ml; EGTA, 5 mM; hydroxyurea, 7.5 mM; KCl, 1 M; latrunculin A (Cayman Chemical, Ann Arbor, Michigan), 0.25 µM; methyl benzimidazol-2-yl-carbamate (MBC), 10 µg/ml; methyl methanesulfonate (MMS), 0.01%; sodium dodecyl sulfate (SDS), 0.005%; sorbitol, 1.2 M; and thiabendazole (TBZ), 12.5 µg/ml. Plates were incubated at 29° or other temperatures as indicated for 2–8 d and colonies were imaged with a scanner. To assay strain sensitivity, growth of each strain on rich YE media at 29° was compared with other conditions. If deletion cells grew significantly less when exposed to certain reagents or conditions compared with that strain’s growth on YE alone, the gene deletion strain was scored as sensitive to the reagent. Growth sensitivities were confirmed and classified as “very sensitive” if they did not grow relative to wild-type cells, “sensitive” if deletion cells were growing at approximately two or more dilution factors lower than that of wild-type cells, “resistant” if deletion cells grew better than wild-type cells, or “not sensitive” if deletion cells grew similarly or at one dilution factor of wild-type cells. An example is shown in Figure S1.

Cluster analysis

Cluster analysis was performed with the open source clustering software Cluster3.0 and visualized using Java Treeview. Both software packages were downloaded from http://bonsai.hgc.jp/~mdehoon/software/cluster/software.htm; 2, 1, 0, and −1 were used to represent “very sensitive,” “sensitive,” “not sensitive,” and “resistant,” respectively. Hierarchical clustering was performed with the complete linkage method.

Microscopy methods

For protein localization studies, live cells were imaged. For visualization of DNA, cells were fixed with ice-cold 70% ethanol. After washing with PBS, the fixed cells were incubated with PBS containing 5 mg/ml methyl blue (MB) or 5 µg/ml calcofluor white for 15 min on ice. MB-stained cells were then pelleted by centrifugation and re-suspended in PBS before imaging. Calcofluor white-stained cells were washed four times with PBS before re-suspension in PBS prior to imaging. Cells were mixed with DAPI before imaging. One of the two following microscopes were used to collect images: a spinning-disk confocal microscope (Ultraview LCI; PerkinElmer) with a 100× NA 1.40 Plan-Apochromat oil immersion objective and a 488-nm argon ion laser (GFP), or a 594-nm helium neon laser (mCherry and RFP) or HBO 100 mercury lamp (DAPI, methyl blue, and calcofluor), or a personal microscope system (DeltaVision; Applied Precision) including a microscope (IX71; Olympus), 60× NA 1.42 PlanApo and 100× NA 1.40 UPlanSApo objectives, fixed-cell and live-cell filter wheels, a camera (CoolSnap HQ2; Photometrics), and softWoRx imaging software. Images on the confocal were captured on a charge-coupled device camera (Orca-ER; Hamamatsu Photonics) and processed using MetaMorph software (version 7.1; Molecular Devices). Images were analyzed and quantified in Image J and imported into Adobe Illustrator for construction of figures. Bar graphs were created in Microsoft Excel and p values were calculated using two-tailed unequal variance Student’s t-tests.

Results

Gene deletion strategy

To make S. pombe gene deletions, two sequential PCRs were optimized to produce a single DNA fragment that was then used to replace each gene with the kanMX6 cassette (Wach ). This deletion strategy is schematized in Figure 1. First, primers were designed to amplify from S. pombe genomic DNA approximately 300 bp of 5′ and 3′ sequences flanking each open reading frame (ORF) targeted for deletion (see Materials and Methods for primer design and Table S1, for primer sequences). Then, 23 nucleotides complementary to the kanMX6 cassette were added to the reverse primer for the 5′ flank reaction and to the forward primer for the 3′ flank reaction. Usually, amplification of the flanking sequences was robust and a sufficient quantity of the PCR product was obtained (see, for example, Figure S2). A second PCR was then performed to amplify the kanMX6 module within pFA6-kanMX6 (Bahler ) sandwiched by the 5′ and 3′ flanking sequences by using the PCR products from the first round of PCR as primers. Each final PCR product (∼2 kb in length) was transformed into the haploid S. pombe strain ade6-M210 leu1-32 ura4-D18 h because we were interested in obtaining deletions of only nonessential genes for genome-wide screens; integration of the marker was selected by growth on YE-G418 plates at 29°.

To test whether ORFs had been correctly replaced with kanMX6, three PCR reactions were performed on G418-resistant colonies (1–2 per targeted gene). The “left” check PCR reaction was performed with a forward primer corresponding to sequences upstream of the 5′ flanking sequence (5′ chk; Figure 1A) and a reverse primer corresponding to sequences 20 bp downstream of the start codon of the kan ORF (kanR chk, Figure 1B; Table S1 and Table S2). The “right” check reaction was performed with a forward primer corresponding to sequences 113 bp upstream of the 3′ end of the kanMX6 cassette (kanF chk, Figure 1B), and a reverse primer corresponding to sequences downstream of the reverse primer used for 3′ flanking sequence amplification (3′ chk; Figure 1A; Table S1 and Table S2). To make sure the ORF had been removed, a third PCR (ORF check) was performed. This reaction used the same forward primer as that used for “left check” (e.g., 5′ chk) and a reverse primer corresponding to sequences inside the ORF (ORF chk; Figure 1A; Table S1 and Table S2). The absence of a PCR product from this reaction strongly suggested that the ORF had been deleted; to rule out the possibility that a technical issue prevented amplification of a product in the ORF check reaction, a fourth PCR reaction was performed using wild-type S. pombe cells and the same set of primers. Using this approach, 281 deletions were successfully constructed (Table S1) from ∼400 attempted. Some of the genes that were unsuccessfully deleted from the haploid strain may be essential. This deletion set of 281 strains is available through the Yeast Genetic Resource Center (http://yeast.lab.nig.ac.jp/nig/index_en.html).

Stress sensitivities of the deletion strains

Although a large percentage of the deleted genes have characterized or predicted functions, several have no functional assignment (Wood ). To help determine which biological processes the uncharacterized genes participate in, and to explore additional functions for characterized genes, the growth of each of the 281 deletion strains was assayed in 16 different stress conditions. Specifically, the strains were tested for growth at 29° on rich YE agar containing one of the following 13 compounds: the fungal microtubule destabilizing drugs thiabendazole (TBZ) or methyl-2-benzimidazole-carbamate (MBC) (Yamamoto 1980), the actin cytoskeleton destabilizing drug latrunculin A (Lat A) (Ayscough ), the DNA replication inhibitor hydroxyurea (HU) (Yarbro 1992; Slater 1973), the DNA damage-inducing drugs bleomycin (Povirk 1996) and methyl methanesulfonate (MMS) (Beranek 1990), the secretory pathway inhibitor brefeldin A (BFA) (Turi ), the protein synthesis inhibitor cycloheximide, the cell wall or membrane structure compromising agents calcofluor white or sodium dodecyl sulfate (SDS), high salt (KCl), the osmolyte sorbitol, or the calcium chelator EGTA. In addition, cells were tested for growth on rich medium at low (19°) or high (36°) temperatures and at 29° on minimal medium (EMM2) agar plates supplemented with adenine, uracil, and leucine. The sensitivities of the 77 deletion strains whose growth was affected by at least one stress are summarized in Table S3 and the various levels of sensitivity are illustrated in Figure S1.

Approximately 12% (35) of the deletion strains showed sensitivity to three or more stresses (Table S4). A survey of the GO functional annotations for the corresponding genes (Wood ) linked most to intracellular transport, biosynthesis, cellular metabolism, stress response, and/or transcription (Figure 2). Group 1 genes are involved in intracellular transport, transcriptional regulation, and DNA damage response/repair. Loss of group 1 genes results in pleiotropic effects (i.e., multiple sensitivities), likely because these genes are involved in core cellular functions. Deletions of group 2 genes were characterized by high sensitivity to low nutrient condition (minimal media), calcofluor white, and resistance to HU treatment. Interestingly, several members of this group are involved in biosynthesis and mitochondrial respiration, indicating a strong linkage between respiration and nutrient availability.

Figure 2

Phenotypic clustering of the deletions that showed sensitivity or resistance to at least three stress conditions. Hierarchical clustering of the sensitivity and resistance patterns of the deletions was performed. GO biological processes of the ORFs involved are shown on the right. Growth conditions are as defined in Materials and Methods.

New effectors of cell division

Among the 35 strains that exhibited sensitivity to three or more stresses (Figure 2), deletion of two uncharacterized ORFs showed cell division defects that have not been previously reported. SPAC1F12.07∆ cells were slow-growing at low temperature and sensitive to TBZ, calcofluor white, and SDS, congruent with defects in cell morphogenesis. Microscopic imaging of SPAC1F12.07∆ cells showed that they were elongated at both 32° and 19° (Figure 3, A and B), and approximately 50% of the cells formed a persistent septum that failed to cleave the daughter cells (Figure 3, C and D); a low percentage of SPAC1F12.07∆ cells formed no septa at all. SPAC1F12.07 is predicted to be a phosphoserine aminotransferase involved in L-serine biosynthesis (Wood ) and peaks in abundance during mitosis (Carpy ).

Figure 3

SPAC1F12.07Δ and efr3Δ cells exhibit cell division defects. (A) DIC images of wild-type (wt) and SPAC1F12.07Δ cells. (B) Cell length of wt and SPAC1F12.07Δ cells at 32° and 19°. ∆, SPAC1F12.07Δ. Standard error of the means (SEM) are shown. ***p < 0.001. (C) DAPI and methyl blue (MB) staining of wt and SPAC1F12.07Δ cells. (D) Quantitation of persistent septal material as shown in (C). (E) DAPI and MB staining of wt and efr3Δ cells. (F) Quantitation of off-center septum in wt and efr3Δ cells. SEM is shown. **p < 0.01. (G) DAPI and MB staining of efr3Δ cells showing “cut” phenotype. Scale bars, 5 µm.

Deletion of efr3 also led to multiple growth sensitivities and defects in cell division (Figure 3, E–G). Staining of the nuclei and cell wall showed that 23% of septating efr3Δ cells had a septum misplaced toward one cell tip compared with only 8% of septating wild-type cells, as observed previously (Jourdain ) (Figure 3, E and F). As a result, some efr3Δ cells mis-segregated their chromosomes, displaying a “cut” phenotype, or had variable cell sizes because of asymmetric divisions (Figure 3G). In S. cerevisiae and humans, Efr3 is a component of the plasma membrane phosphoinositide (PI) 4-kinase complex that produces PI4P (Wu ; Baird ). By analogy, deletion of efr3 in S. pombe likely disrupts PI4P production at the plasma membrane and, therefore, the localization of proteins involved in contractile ring placement and ultimately septum targeting and deposition.

New players in DNA metabolism

The 10 deletion mutants sensitive to one or more drugs that interfere with DNA metabolism are listed in Table S5. Because the three drugs used differ in their modes of interfering with DNA metabolism, sensitivities also varied. Four gene deletion strains not previously linked with DNA metabolism were analyzed further.

As shown previously (Walworth ; Bimbo ), deletion of the checkpoint kinase gene, chk1, resulted in high HU sensitivity, which served as a control for our experiments (Figure 4A). Two other deletions in our set, upf2Δ and rpa34∆, were also very sensitive to HU, suggesting an involvement in DNA metabolism (Figure 4A). Rpa34 encodes a subunit of the RNA polymerase I and S. cerevisiae Rpa34 interacts with DNA topoisomerase I (Top1) (Beckouet ). S. cerevisiae rpa34∆ cells are uniquely sensitive to the overexpression of TOP1 and are also sensitive to the DNA damaging agent camptothecin (Reid ). Thus, Rpa34 regulation of rDNA transcription may play a conserved role in DNA repair. S. pombe Upf2 has not been well-characterized, but it and its human ortholog (UPF2) form complexes with Upf1/UPF1 and Upf3/UPF3 and participate in nonsense-mediated mRNA decay (Chang ; Matia-Gonzalez ). Interestingly, human UPF1, an RNA-dependent and DNA-dependent 5′–3′ helicase, is also involved in DNA synthesis during replication and repair and is required for genome stability (Azzalin and Lingner 2006). The S. pombe Upf complex is additionally required for the proper response to oxidative stress (Matia-Gonzalez ). upf2Δ cells are elongated relative to wild-type (Figure 4B), possibly because they contain damaged DNA, which triggers a modest G2/M checkpoint delay.

Figure 4

Characterization of gene deletion strains sensitive to DNA damaging agents. (A) The 10-fold dilutions of wild-type, upf2Δ, rpa34∆, and chk1∆ cells were spotted on YE agar in the absence or presence of 7.5 mM hydroxyurea (HU). Plates were incubated at 29° and 19° for 3–8 d. (B) The indicated strains were grown at 29° or 19° in liquid YE medium to log phase and the lengths of septating cells were measured with microscopy. SEM is shown. ***p < 0.001. (C) Spotting growth assay of the indicated strains on YE agar in the absence or presence of 1 µg/ml bleomycin (Bleo) or 7.5 mM HU. (D) Images of DIC and DAPI/calcofluor white staining of wild-type and SPAC607.02cΔ cells treated (lower panels) or untreated (upper panels) with 1.25 µg/ml bleomycin for 4 hr. (E) DAPI/calcofluor white staining images of ptc1Δ cells treated or untreated with 1.25 µg/ml bleomycin for 4 hr. (F) DAPI/calcofluor white staining images of new13Δ cells treated or untreated with 1.25 µg/ml bleomycin for 4 hr. (G) Cell length of septated cells of the indicated strains grown to log phase at 32° measured with microscopy. SEMs are shown. **p < 0.01; ***p < 0.001. (H) Quantitation of Rad22-GFP foci in the indicated strains. Scale bars, 5 µm.

In contrast to upf2∆ and rpa34∆, deletions of ptc1 and two uncharacterized genes, SPAC607.02c and new13, were not overtly sensitive to HU but were very sensitive to bleomycin (Figure 4C). The gene product encoded by SPAC607.02c localizes to the nucleolus (Matsuyama ) and is conserved in fungi but has no apparent S. cerevisiae ortholog (Wood ). A small percentage (2%) of SPAC607.02cΔ cells displayed cell separation defects and contained multiple septa (Figure 4D). On bleomycin treatment, SPAC607.02cΔ cells arrested in the cell cycle, similar to wild-type cells, indicating an intact checkpoint response. However, unlike in the wild-type, the nuclei of SPAC607.02cΔ cells did not remain centered (74% off-center vs. 10% in wild-type), and the cells did not become as elongated (Figure 4D).

In deletions of ptc1, which encodes a protein phosphatase 2C, 24% of cells exhibited septation and/or chromosome segregation defects exemplified by off-center septa and unequal distribution of the daughter nuclei (Figure 4E). On bleomycin treatment, ptc1Δ cells elongated like wild-type cells, indicating an intact DNA damage checkpoint, but the DNA became indistinct or lost in 10% of cells relative to 0% in wild-type (Figure 4E). Although a role for Ptc1 in DNA damage response has not previously been reported, the stress kinase Sty1(Spc1) acts upstream of Ptc1 (Gaits ), and Sty1 is important for the DNA damage response to oxidative stress (Alao and Sunnerhagen 2008). Thus, Ptc1 might act in this same pathway.

The new13 ORF overlaps with that of dsc4, which encodes a component of a Golgi E3 ligase complex, and new13 and dsc4 transcripts converge (Wood ). Therefore, deleting the entire new13 ORF removed sequences encoding the C-terminus of Dsc4 and vice versa. In this regard, it is interesting that dsc4∆ cells are sensitive to the DNA damaging agent, 4-nitroquinoline N-oxide (Deshpande ), which is similar to the effects of bleomycin. This sensitivity is perhaps due to the concomitant deletion of new13 rather than loss of Dsc4-dependent Golgi E3 ligase activity. It will be necessary to make smaller deletions in the new13 and dsc4 ORFs to dissect their unique roles. new13∆ cells treated with bleomycin arrested growth like wild-type; therefore, their sensitivity to this agent is not due to a bypass of the DNA damage checkpoint. However, we note that DNA staining became diffuse in 13% of cells, indicating defects in chromatin organization (Figure 4F).

All three of these bleomycin-sensitive deletion strains were elongated relative to wild-type (Figure 4G), a phenotype that could reflect a mild DNA damage-induced checkpoint delay of entry into mitosis. Therefore, we tested whether higher levels of DNA damage exist in ptc1∆, new13∆, or SPAC607.02cΔ cells relative to wild-type using Rad22-GFP as a marker for the presence of DNA damage repair foci (Noguchi ). Although there were more Rad22-GFP foci in new13∆, there was a reduced number in ptc1∆ and SPAC607.02cΔ cells compared with wild-type (Figure 4H), suggesting that these latter genes might play a role in assembly or maintenance of DNA repair foci during a normal S phase.

Morphogenesis factors

Ten gene deletions showed increased sensitivity to Lat A, which inhibits F-actin and blocks actomyosin ring formation and cytokinesis, or Lat A plus one additional factor (Figure 5A and Table S6). Fab1 is a 1-phosphatidylinositol-3-phosphate 5-kinase responsible for producing PI(3,5)P2 (McEwen ). In S. cerevisiae, Fab1 is required for proper intracellular protein trafficking and vacuolar organization (Gary ). Similarly, S. pombe fab1∆ cells are short, stubby, and have large vacuoles (Figure 5B), as observed previously for point mutations of fab1 (previously called ste12 due to their sterility) (Morishita ). Thus, it is likely that actin filament disruption exacerbated the decreased intracellular trafficking in fab1Δ cells, resulting in cell death.

Figure 5

Characterization of gene deletion strains that are sensitive to actin cytoskeleton perturbation. (A) The 10-fold dilutions of cells from the indicated strains were spotted on YE agar in the absence or presence of 0.25 µM latrunculin A (Lat A). Plates were incubated at 29° or 19° for 3–8 d. (B) Wild-type (wt) and fab1Δ cells were grown at 32° in liquid YE to log phase. DIC images are shown. Scale bar, 5 µm. (C) DIC images of the indicated strains treated or untreated with Lat A for 4 hr at 32° are shown. Scale bar, 5 µm.

Deletions of three other genes not previously linked to morphogenesis (SPAC20G4.09∆, rps401∆, and new20∆) also caused high sensitivity to Lat A (Figure 5A). SPAC20G4.09∆ cells are obviously misshapen and become more so when treated with Lat A (Figure 5C). rps401∆ cells vary slightly from normal morphology but do not show obvious morphological changes on Lat A treatment (Figure 5C). Despite severe Lat A sensitivity, new20∆ cells appeared morphologically wild-type with or without Lat A treatment (Figure 5C). F-actin staining in these cells was consistent with their morphology in that F-actin patches were confined to areas of cell growth (Figure S3). More studies will be required to determine why these cells are so sensitive to actin disruption.

New temperature-sensitive deletion strains

The gon7Δ and new19Δ deletions were unable to survive when plated at 36° (Figure 6A and Table S7). Although gon7Δ cells had normal morphology at 25°, after growth at 36° overnight (16 hr) the cells became highly elongated and multi-septated (Figure 6, B and C). In S. cerevisiae, Gon7 is a component of the KEOPS complex that is required for t6A tRNA modification (Srinivasan ) and promotes telomere uncapping and elongation (Downey ), transcription (Kisseleva-Romanova ), and chromosome segregation (Ben-Aroya ). Although it is unclear how Gon7 impacts septation, loss of GON7 in S. cerevisiae resulted in defects in N-linked oligosaccharide biosynthesis, and this may affect cell wall metabolism (Corbacho ). The new19Δ cells are morphologically normal at 25°, but at 36° a large percentage divide without completely degrading their septum and septum remnants persist at new cell tips (Figures 6, B and D). New19 is the apparent ortholog of the S. cerevisiae signal peptidase complex component Spc1 (Wood ); therefore, new19 deletion is expected to disrupt protein targeting to the ER. The new19 deletion may impair the targeting of primary septum degrading enzymes to the cell division site.

Figure 6

Characterization of temperature sensitive gene deletion strains. (A) The 10-fold dilutions of cells from the indicated strains were spotted on YE agar and plates were incubated at 29° or 36° for 2–3 d. (B) DAPI/calcofluor white staining images of wild-type (wt), gon7Δ, and new19Δ cells grown at 25° and 36° overnight. (C) Quantitation of the septum number of indicated strains grown at 36° overnight. (D) Quantitation of nonseptated cells with septum at the cell tips. Cells were grown at 36° overnight. Mean±SD from two separate experiments is shown. Scale bar, 5 µm.

Discussion

To expand the toolkits available for S. pombe genetic screens (Kim ; Hayles ; Gregan ; Spirek ; Rumpf ), we used the kanMX6 marker (Wach ) to construct individual deletion strains of 281 genes. We assayed the growth of this cohort of strains under a battery of stresses, identifying new players in cell division, DNA damage response, and morphogenesis.

In some cases in which sensitivities of deletion strains have been reported, for example, with chk1 and hhp1 (Bimbo ; Dhillon and Hoekstra 1994; Walworth ), our data largely match the published literature, validating our methodology and providing confidence that the growth tests revealed new and useful information. Interestingly, there are some differences between sensitivities of various gene deletions between fission and budding yeast. For instance, deletion of the poly-ubiquitin gene, ubi4, did not significantly affect cell growth under any of the stress conditions tested; this is in contrast to loss of the orthologous gene in S. cerevisiae, which exhibits multiple phenotypes and stress sensitivities (Finley ). One possible explanation for this difference is the presence of a fifth ubiquitin fusion gene in fission yeast, compared with only four in budding yeast, which could compensate for poly-ubiquitin loss.

During construction of the gene deletion strains, we unexpectedly found that a significant percentage of the G418-resistant colonies arising from transformation with the linear DNA fragment containing the kanMX6 cassette contained a mixed population. In other words, PCR products indicative of both the correctly integrated kanMX6 cassette and the ORF that had been targeted for deletion were detected. This perplexing result may be explained if the linear DNA fragment used for transformation was retained outside of the genome, with recombination and replacement of the ORF occurring only in a sub-population of the colony’s cells. We found that replicating the transformants to fresh YE-G418 plates at least three times and/or streaking to single colonies led to the elimination of cells containing the ORF in most cases. This observation may explain why certain deletion strains have been reported to retain the ORF that had been targeted for deletion, and yet be G418 resistant. Our careful analysis with three independent PCR reactions ensured the fidelity of this set of 281 strains.

Many of the genes targeted for deletion via this approach have been previously deleted with a similar approach (Hayles ) or with other markers such as ura4. For example, nup124, btn1, and chk1 were deleted using other markers years ago (Chen ; Gachet ; Walworth ). However, this deletion series contains many newly generated kanMX6 deletions, expanding the list of nonessential S. pombe genes and genome-wide tools available for large-scale screens.