Genome-Wide Screening for Genes Associated with FK506 Sensitivity in Fission Yeast

We have been studying calcineurin signal transduction pathway in fission yeast Schizosaccharomyces pombe (S. pombe) by developing a genetic screen for mutants that show hypersensitivity to the immunosuppressive calcineurin inhibitor FK506 (tacrolimus). In the present study, to identify nonessential genes that are functionally related to the calcineurin signaling pathway, we performed a genome-wide screen of 3004 haploid deletion strains and confirmed 72 deletion strains to be FK506 sensitive. These 72 genes are classified into nine functional groups to include membrane trafficking (16 genes), signal transduction (10 genes), ubiquitination (8 genes), chromatin remodeling (6 genes), cytokinesis (4 genes), ribosomal protein (3 genes), RNA binding protein (3 genes), and a variety of other known functions (17 genes) or still unknown functions (5 genes) in the biological system. In our previous screening of FK506-sensitive mutants we isolated several membrane-trafficking mutants showing defective cell wall integrity. Here, we further examined the vacuolar fusion, the v-SNARE synaptobrevin Syb1 localization, and the sensitivity to the β-glucan synthase inhibitor micafungin in these 72 FK506-sensitive strains. Results showed that 25 deletion strains exhibited abnormal vacuole fusion, 19 deletion strains exhibited Syb1 mislocalization, and 14 deletion strains exhibited both abnormal vacuole fusion and Syb1 mislocalization, while 42 deletion strains showed both normal vacuole fusion and Syb1 localization. Likewise, 16 deletion strains showed sensitivity to micafungin. Altogether, our present study indicates that calcineurin mediates a plethora of physiological processes in fission yeast, and that calcineurin is extensively involved in cross-talk between signaling pathways.


Introduction
Calcineurin is a highly conserved calcium-dependent serine/ threonine protein phosphatase. As a key signal transduction molecule, calcineurin mediates the Ca 2+ -dependent signaling to a wide variety of cellular responses. In mammals, calcineurin regulates a variety of physiological processes including T-cell activation, cardiac muscle development, skeletal muscle fiber-type switching, apoptosis, long-term potentiation in learning and memory, neuronal plasticity and oxidative stress [1][2][3][4][5]. Calcineurin is also involved in some pathophysiological processes such as cardiac hypertrophy, ischemia-reperfusion injury, and chronic allograft nephropathy [6][7][8].
In budding yeast Saccharomyces cerevisiae (S. cerevisiae), calcineurin is required for response to environmental stress [9,10], endoplasmic reticulum (ER) stress [11] and cell wall damage [12]. Transcription factor Crz1 functions as a key downstream target of calcineurin [13]. In addition to transcription, calcineurin modifies the distribution of Hph1p within ER and is required for pH stress responses [14]. Slm1p and Slm2p are reported to be two novel calcineurin substrates that are required for heat stress-induced endocytosis of uracil permease [15].
In fission yeast Schizosaccharomyces pombe (S. pombe), calcineurin knockout cells showed defects in cell polarity, mating and spindle pole body positioning [16]. Calcineurin also acts antagonistically with the Pmk1 MAP1 kinase in chloride ion homeostasis [17][18][19]. In addition, we demonstrated that calcineurin activates at least two distinct signaling branches, i.e. the transcription factor Prz1dependent branch that regulates the expression of the Pmc1 Ca 2+ pump and an unknown pathway that functions antagonistically with the Pmk1 MAP kinase pathway [20].
Studies on the biological function of calcineurin have been greatly facilitated by the use of the immunosuppressive drugs FK506 (tacrolimus) and cyclosporin A, which have been widely used to prevent graft rejection after organ transplantation and have been increasingly used in the management of autoimmune diseases. FK506 is a known calcineurin inhibitor. We have been studying calcineurin signal transduction pathway in S. pombe by analyzing FK506-sensitive mutants. We demonstrated that calcineurin is implicated in cytokinesis, septation initiation network, membrane trafficking, and ion transport [21][22][23][24][25]. In the present study, we performed a genome-wide screen of fission yeast nonessential deletion strains to search for the genes in which gene deletion causes sensitivity to FK506, and we found that these genes encode proteins that are involved in various physiological functions such as membrane trafficking, signal transduction, ubiquitination, chromatin remodeling, cytokinesis, ribosomal protein, RNA binding, and a variety of other well-known functions or still unknown functions in the biological system. Altogether, our present study provides valuable insights on a complex signaling network in which calcineurin cross talks with other signaling pathways.

S. pombe Nonessential Gene Knockout Library
Heterozygous diploid deletion strains were constructed by Bioneer Corporation and Korea Research Institute of Biotechnology and Bioscience (http://pombe.bioneer.co.kr/). These deletion strains were generated with a genetic background of h + /h + ade6-M210/ade6-M216 leu1-32/leu1-32 ura4-D18/ura4-D18 using PCR-based deletion method [26]. The haploid deletion library used in this study consists of 3004 nonessential genes, each of which carries a defined deletion of a characterized or a putative nonessential open reading frame (ORF) replaced with the kanMX4 cassette. Deletion of the target ORF was screened by G418 antibiotic selection.

Media, Genetic and Molecular Biology methods
The complete medium YPD (yeast extract-peptone-dextrose) and the minimal medium EMM (Edinburgh minimal medium) have been described previously [27]. YPD plates are supplemented with 225 mg/l adenine to produce YPDA (yeast peptone dextrose adenine) plates. YE plates (0.5% yeast extract, 3% glucose, 2% agar) are supplemented with 225 mg/l adenine, histidine, leucine, uracil and lysine to produce YES (yeast extract with supplements) plates. Gene disruptions are abbreviated by the gene preceded by D (for example, Dryh1). Proteins are denoted by Roman letters and only the first letter is capitalized (for example, Ryh1).

Genome-wide Screen for FK506-Sensitive Deletion Mutants
The deletion mutant library was frozen at 280uC in 96-well microtitre plates in 30% glycerol in liquid YES medium. Prior to performing the experiment, the library was transferred to YPDA plates at 27uC. The log-phase cells were streaked onto YPDA plates with or without 0.5 mg/ml FK506 (gift from Astellas Pharma Inc. Japan) and incubated at 27uC for 5 days.

Localization of GFP-Syb1
The plasmid containing GFP-tagged Syb1, the synaptobrevin equivalent in fission yeast [28], was constructed under the thiamine-repressible nmt1 promoter as described previously [29]. Expression was repressed by the addition of 4 mm thiamine to EMM plus adenine and uracil. The log-phase cells harboring pREP1-GFP-Syb1 were observed as described previously [29].

Vacuole Fusion
The cells were grown to exponential phase in liquid YES medium at 27uC, harvested and washed twice with distilled water. Then the cells were resuspended in 1 ml distilled water and incubated at 27uC for 60 min. The distilled water-treated cells were harvested by centrifugation at 150006 g for 1 min at 4uC to remove excess distilled water, then placed on ice, and immediately examined under a microscope.

Miscellaneous Methods
The differential interference contrast (DIC) and fluorescence microscopy images were recorded digitally on a Zeiss Axiophot microscope equipped with a SPOT-2 camera, and were processed with CorelDRAW software. Calcineurin-dependent response element (CDRE)-reporter activity was measured as described previously [30]. The extracellular and total acid phosphatase activities were determined as described previously [25], and the acid phosphatase secretion index was calculated as the ratio between the extracellular and the total activity and thus represents digits without units. Database searches were performed using the Sanger Center S. pombe data base search service (www.sanger.ac. uk).

Results and Discussion
Identification of genes required for growth upon FK506 exposure To identify nonessential genes associated with increased sensitivity to FK506, we performed a genome-wide screen. First, we assessed the growth of Dryh1 and Dric1 cell, which showed FK506 sensitivity in our previous study [31], on YES plus FK506 or YPDA plus FK506 plates. The results showed that Dryh1 and Dric1 cells completely failed to grow on YPDA plus FK506 plates, whereas the cells partially grew on YES plus FK506 plates (data not shown). Therefore, we used YPDA plus FK506 plates as test plates. Next, we compared the growth of the wild-type cells with the 3004 deletion cells on YPDA and YPDA plus FK506 plates.
This screen resulted in the isolation of 72 deletion strains that displayed varying levels of sensitivity to FK506. The severity of growth inhibition by FK506 was scored as follows: severe (+++) indicating that the cells completely failed to growth on the FK506 containing plates ( Figure 1A, upper panel), moderate (++) indicating that tiny colonies were observed to grow on the FK506-containing plates ( Figure 1A, middle panel), and mild (+) indicating that colonies were observed to grow on FK506containing plates, however, the size of the colonies were significantly smaller than that of the wild-type cell ( Figure 1A, lower panel). Among the 72 FK506-sensitive mutants, 20 mutants were severely sensitive (+++), 24 were moderately sensitive (++), and 28 were mildly sensitive (+).
The 72 genes were grouped according to their functions ( Table  S1). The largest group consisted of genes involved in membrane trafficking (16/72 = 22.2%), the second and third largest groups consisted of genes involved in signal transduction (10/72 = 13.8%) and ubiquitination (8/72 = 11.1%), respectively. Other groups consisted of genes involved in chromatin remodeling, cytokinesis, ribosomal protein, RNA binding, and there were also a variety of genes with other known or unknown functions in the biological system. For each gene listed in Table S1, the systematic name, common name of the gene from S. pombe (if available), along with a brief description of the function of each gene product were also indicated. For convenience, we named the genes after their S. cerevisiae counterparts as the common name of the gene from S. pombe is not available.
To further examine the role of these 16 genes in membrane trafficking, we investigated the vacuolar morphology and the intracellular localization of v-SNARE synaptobrevin Syb1, a Snc1p homolog [23]. On vacuolar morphology, as shown in Figure 2A, the wild-type cells had evidently large vacuoles that resulted from vacuole fusion, whereas 10 of the 16 deletion strains showed significant defects in vacuole fusion. Only 6 of the 16 genes (Aps3, Apl6, Vps1302, Vps71, Erv15 and Myo1) showed normal vacuole fusion. A summary of the characteristic features are listed in Table 1.
In group I deletion strains, similar to wild-type cells, GFP-Syb1 was visibly seen on the plasma membrane with more prominence at the growing ends of the cells (data not shown). Of the seven strains categorized in group I, two strains (Derv15 and Derp2) showed defects in the transport from ER to the Golgi transport, three strains (Daps3, Dapl6, and Dgga1) exhibited defects in the transport from the Golgi to vacuole, and the other two strains (Dvps71 and Dvps1302) showed defects in vacuole protein sorting (Table 1). In group II deletion strains, GFP-Syb1 accumulated as large dot-like structures in the cytoplasm as exemplified by Dapm1 and Dapl4 ( Figure 2B). These two strains exhibited defects in the anterograde transport from the Golgi to the endosome or from the endosome to the plasma membrane, or they showed defects in the retrograde transport from endosome to the Golgi. In group III deletion strains, GFP-Syb1 localized almost evenly all around the cell surface, and the punctate structures in the cytoplasm were barely observed ( Figure 2B, Dend4, Dmyo1, and Dgyp1). The Dend4 and Dmyo1 strains were characterized to have defects in endocytosis [33][34][35][36]. Interestingly, knockout of the gyp1 + gene, which encodes Rab/Ypt GTPase activating protein (GAP), also showed the accumulation of Syb1 on the plasma membrane ( Figure 2B). We speculate that the Dgyp1 mutants have defects in endocytosis, so that Syb1 failed to be endocytosed, thus accumulating on the plasma membrane. In agreement with our hypothesis, in budding yeast it is reported that in vitro Gyp1p acts on Ypt51p, that encodes a Rab/Ypt GTPase required for endocytosis [37,38]. Likewise, in fission yeast, Gyp1 may play an important role in endocytosis via the GAP activity. In group IV deletion strains, GFP-Syb1 failed to localize to the membrane and the fluorescence was weak and hazy ( Figure 2B). Of the four strains categorized in group IV, two strains (Dvps45 and Dsft2) showed defects in the transport from the Golgi to endosomes, one strain (Dryh1) exhibited defects in the recycling from the endosome to Golgi, and still another strain (Dgos1) showed defects in both ER to the Golgi and intra-Golgi transport (Table 1).
To further investigate the genetic/functional interaction between these membrane trafficking mutants and calcineurin, we examined the effect of C-terminal-deleted Ppb1 (Ppb1DC), the constitutively active form of Ppb1, on vacuole fusion and on Syb1 localization. The results showed that Ppb1DC overexpression failed to suppress the defective vacuole fusion of the membrane trafficking mutants listed in Table 1 (data not shown). Also, Ppb1DC overexpression failed to suppress defective Syb1 localization in the Dapl4, Dend4, and Dgos1 cells (data not shown). Altogether, our results suggest that the increased calcineurin activity failed to compensate the defects in membrane trafficking, and that calcineurin and these membrane trafficking genes function in a parallel pathway to control cell wall integrity. Ubiquitination mutants that showed sensitivity to FK506 Ubiquitination is an extremely versatile control mechanism that regulates almost all aspects of cell life, including cell cycle, DNA transcription and repair, differentiation and development, apoptosis, signaling transduction and membrane trafficking. Ubiquitination plays important roles not only in the internalization of plasma membrane proteins, but also in their sorting to the different cellular destinations, namely the trans-Golgi network (TGN), endosome, lysosomes/vacuoles or for degradation by the proteosome [39]. Majority of ubiquitin-tagged molecules usually are degraded by the 26S proteasome, and a minority of the ubiquitin-tagged membrane proteins are endocytosed and degraded in the vacuole. To investigate whether the FK506 sensitivity of these ubiquitination-related gene deletion strains is mediated by membrane trafficking, we observed the Syb1 localization and the vacuole fusion. As shown in Figure 3A, in Dubr1, Dsst2 (SPAC19B12.10), Dmub1 (SPBC31F10.10c), Dhus5/ubc9 (SPAC30 D11.13), and DSPAC31G5.18c cells, GFP-Syb1 failed to localize to the plasma membrane, instead GFP-Syb1 distributed throughout the intracellular space. Syb1 showed normal localization in the mutant cells of SPAC31G5.18c (encoding ubiquitin-protein ligase E3) and in the mutant cells of csn1 + and csn2 + (encoding COP9/ signalosome complex subunit) (data not shown). In Dsst2, Dmub1, and Dhus5/ubc9 cells, the vacuoles remained numerous and small, suggesting a defect in vacuole fusion. Table 2 showed a summary of the vacuole fusion and Syb1 localization in ubiquitination defective mutants. Taken together, we speculate that the ubiquitination-related gene deletion strains displayed FK506 sensitivity mostly via a defect in membrane trafficking ( Figure 3B).

Membrane trafficking and ubiquitination mutants showed defective acid phosphatase secretion
Our previous results showed that Dryh1 and Dapm1 mutants showed defective acid phosphatase secretion [23,29]. This prompted us to investigate whether acid phosphatase secretion is defective in these membrane trafficking and ubiquitination mutants. The results showed that in Dapm1, Dvps1302 and Dgos1 cells the acid phosphatase secretion index was dramatically decreased to approximately 50% of that in wild-type cells (Figure 4). In Dcsn1 and Dcsn2 cells, the acid phosphatase secretion was moderately decreased to approximately 75% of that in wildtype cells (Figure 4). Notably, the deletion of the two subunits of the AP-3 complex, in particular Daps3 and Dapl6, resulted in an approximately two-fold increase in the acid phosphatase secretion index (Figure 4).

Cell wall integrity mutants that showed sensitivity to FK506
In S. cerevisiae and S. pombe, calcineurin is required for response to cell wall damage [12,40]. To investigate whether the 72 FK506- The wild-type (wt) cells and membrane trafficking defective mutants were grown to early log-phase in YES media at 27uC, harvested and washed twice with distilled water. Then the cells were resuspended in 1 ml distilled water and incubated at 27uC for 60 min. The distilled water-treated cells were harvested by centrifugation at 150006 g for 1 min at 4uC to remove excess distilled water, then placed on ice, and immediately examined under a fluorescence microscope. Bar, 10 mm. (B) The Syb1 localization of membrane trafficking defective mutants. The wildtype (wt) cells and membrane trafficking defects mutants were transformed with pKB4160 (pREP1-GFP-Syb1) respectively, and the cells were spread onto EMM plus 225 mg/l adenine, 225 mg/l uracil and 4 mM thiamine. Three single colonies were individually picked up, further subcultured and grown to early log-phase on the EMM plates as described above. GFP-Syb1 was examined under the fluorescence microscope. Bar, 10 mm. doi:10.1371/journal.pone.0023422.g002 sensitive mutants show defects in cell wall integrity, we streaked these cells onto YPDA plate containing micafungin, an inhibitor of (1, 3)-b-D-glucan synthase. Fifteen deletion strains were confirmed to display varying levels of hypersensitivity to micafungin ( Table 3). The severity of growth inhibition on YPDA plate plus 1.5 mg/ml micafungin was scored as follows: severe (+++) indicating that the cells completely failed to grow on YPDA plus 1.5 mg/ml micafungin, moderate (++) indicating that tiny colonies were observed on the micafungin-containing plates, or mild (+) indicating that colonies were observed on the micafungin-containing plates, however, the size of the colonies were significantly smaller than that of the wild-type cells (Table 3). Three mutants (Daps3, Dsst2, and Dcyb5) were strongly sensitive (+++), seven mutants (Dgyp1, Dlsg1, Dryh1, Dapm1, Dvps45, Dpal1, and Dvps1302) were moderately sensitive (++), and six mutants (Dapl6, Dste20, Dubr1, Drga7, Dmug132, and Duge1) were mildly sensitive (+). Previously, we speculated that the FK506 sensitivity of the membrane trafficking defective mutant may result as a consequence of the impaired cell wall integrity. However, among the 15 FK506-sensitive mutants, 8 deletion strains did not show micafungin sensitivity. Therefore, we conclude that in addition to cell wall integrity, other mechanism is involved in FK506 sensitivity. In budding yeast, it is reported that FKS1 and FKS2 encode the catalytic subunits of (1, 3)-b-D-glucan synthase, and the expression of FKS2 is induced by the addition of Ca 2+ to the growth medium in a calcineurin-dependent manner [41,42]. This prompted us to observe whether the overexpression of the constitutively active form of Ppb1 (Ppb1DC) could complement the micafungin sensitivitity. The results showed that Ppb1DC overexpression failed to suppress the micafungin sensitivity, suggesting that the increased calcineurin activity failed to compensate the cell damage caused by micafungin.

Signaling transduction mutants that showed sensitivity to FK506
As shown in Table S1, ten signaling transduction mutants were identified to display FK506 sensitivity. Three genes (Ckb1, Pit1, and Wis1) encode protein kinases (PKs), another three genes (Par1, Clp1, and Pmp1) encode protein phosphatases (PPs), and another gene Rad24 encodes a 14-3-3 family protein, which interacts with more than 200 phosphorylated proteins in mammals. Protein phosphorylation plays a crucial role in biological functions and controls nearly all cellular processes. These processes depend on the highly regulated and opposing actions of PKs and PPs, through changes in the phosphorylation of key proteins. We speculate that some PKs or PPs have common substrates with calcineurin, and that this substrate may be involved in an essential function.
Recently, it is reported that the Ste20-dependent phosphorylation modulates the dynamic localization and endocytic function of Myo1 in fission yeast [35]. This prompted us to investigate whether the localization of Syb1 in Dste20 mutants is similar to that in Dmyo1 mutants. Contrary to our expectations, in Dste20 mutants GFP-Syb1 did not accumulate on the cell surface, and instead was distributed throughout the intracellular space (data not shown). We also investigated the localization of GFP-Syb1 in the other three deletion strains (Dsin1, Dpop3 and Dbit61) of the TORC2 complex components, and results showed that Dste20 mutants, but not the other three TORC2 complex components,  displayed Syb1 mislocalization (data not shown). Taken together, our study suggests that Ste20 plays a role in membrane trafficking in addition to its role as a member of the TORC2 complex.

Signaling transduction mutants, ubiquitination mutants and unknown function mutants showed abnormal CDREreporter activity
We previously established an in vivo real-time monitoring system of calcineurin activity utilizing the reporter harboring the calcineurin-dependent response element (CDRE)-fused to luciferase, and showed that high extracellular CaCl 2 and NaCl concentration caused an increase in the CDRE-reporter activity in fission yeast [30]. To investigate whether the calcineurin activity was affected in these FK506-sensitive mutants, we monitored the CDRE reporter activity in the signaling transduction mutants, ubiquitination mutants and other mutants with unknown function.
The result showed that all of these mutants except for Dckb1 showed responses that cannot be distinguished from wild-type cells to elevated extracellular CaCl 2 . In wild-type cells, elevated extracellular CaCl 2 caused an increase in the calcineurin activity as exhibited by a peak rise and then a slow decay ( Figure 5A, wt). However, in Dckb1 mutants, elevated extracellular CaCl 2 caused an increase in the calcineurin activity that exhibited a peak rise and then approached a constant high level ( Figure 5A, Dckb1). Notably, the basal calcineurin activity in Dckb1 cells was markedly higher than that in wild-type cells ( Figure 5A), suggesting that Ckb1 is involved in the regulation of Ca 2+ /calcineurin/Prz1 pathway.
The deletion strains were classified into three groups (Table 4) according to their response to elevated extracellular NaCl. Gene deletion strains belonging to Group A showed responses that cannot be distinguished from wild-type cells (Table 4 and Figure 5B, wt). Gene deletion strains belonging to Group B showed high basal RLU with normal response to NaCl (Table 4 and Figure 5B, Dmug132). Gene deletion strains belonging to Group C showed high basal RLU with low response to NaCl (Table 4 and Figure 5B, Dpar1 and Dckb1). Of the seven mutants that showed high basal transcription activity (Table 4), 4 mutants (Dsst2, Dste20, Dubr1 and Dmug132) also showed micafungin sensitivity, suggesting that cell wall integrity defect may result in a high transcriptional activity to compensate for the defects. Our results also suggest that Ckb1 and protein phosphatase regulatory subunit Par1 may crosstalk with Ca 2+ / calcineurin/Prz1 pathway.

Perspective on the side effects of FK506
FK506 has well-documented side effects such as infection, cardiac damage, hypertension, nephrotoxicity and various neuropsychiatric problems [43][44][45]. FK506 also increases the risk of cancer [46,47]. However, the molecular basis for these adverse effects is not yet fully understood. In model organisms such as fission yeast, synthetic lethality has been extensively used to characterize the interactions between genes that are likely to be involved in similar processes and to uncover new molecular interactions. In cases when both mutations are null, the interpretation is that the two genes are required in parallel pathways toward a common function, and the loss of gene function is lethal. In the present study, we systematically screened FK506-sensitive mutants to identify the possible molecular component or signaling pathway underlying the therapeutic action and the unwanted effects of FK506, and we identified 72 mutations that display synthetic lethal interactions with calcineurin. It is particularly interesting to note that most of the FK506-sensitive deletion strains identified in this study were defective in various steps in membrane trafficking. Thus, normal membrane trafficking is required for FK506 resistance, and membrane trafficking mutants require calcineurin activity for their growth.

Supporting Information
Table S1 Summary of the gene name and products of FK506sensitive mutants.
(DOCX) The wild-type and deletion cells respectively were transformed with the plasmid containing the 3xCDRE::luc(R2.2) reporter gene, and the reporter activity was monitored as described in Materials and Methods. doi:10.1371/journal.pone.0023422.t004