Background Although no potential homologues of multicellular apoptotic genes (e.g. Bax, Bak, Bcl-2, caspases and p53) have been identified in a unicellular eukaryote, previous reports contain several implications of the apoptotic behaviour of yeasts (i.e. Saccharomyces cerevisiae and Schizosaccharomyces pombe). Therefore, whether or not yeast undergoes apoptosis has been a topic of some debate. hCCG1, which is the largest subunit of TFIID and a histone acetyltransferase, appears to be involved in the regulation of apoptosis. The factor hCIA interacts with hCCG1 and functions as a histone chaperone in mammalian cells; its homologue in yeast is Asf1p/Cia1p. Therefore, we anticipated that a yeast mutant in Asf1p/Cia1p would be a valuable tool for studying apoptosis in yeast.
Results We established a strain of S. cerevisiae lacking the histone chaperone ASF1/CIA1. This disruptant, asf1/cia1, arrested preferentially at the G2/M-phase and died. We systematically analysed the phenotype associated with the death of this mutant yeast and identified many changes, such as fragmentation of the nuclei, condensation and fragmentation of chromatin, reduction of the mitochondrial membrane-potential, dysfunction of the mitochondrial proton pump, and a discernible release of cytochrome c to cytoplasm that resembles those in apoptotic multicellular organisms. Other changes potentially associated with the death in our mutant included a reduction in the vacuolar membrane potential, dysfunction of the vacuolar proton pump, reduction of endocytosis, and the presence of many autophagic bodies. However, these mutant yeast cells also showed cellular enlargement, which is characteristic of necrosis.
Conclusions Cell death in S. cerevisiae occurs with a phenotype that largely resembles apoptosis in multicellular organisms, but that has some features of necrosis. Therefore, we indicate that yeast undergoes a ‘prototypal active cell death’ that retains some characteristics of passive cell death (necrosis). In addition, we think that active cell death is ubiquitously the essential attribute of life. Although such an active cell death system in yeast remains open to confirmation, we speculate that deletion of the histone chaperone Asf1p/Cia1p inhibits the normal assembly/disassembly of nucleosomes in yeast and thereby initiates the active cell death system.
Cell death in multicellular organisms proceeds through two distinct pathways—apoptosis and necrosis. Apoptosis is an ‘active’ cell death that functions to focally eliminate useless and unfavourable cells during embryonic development and differentiation and leads to cell turnover in many dysfunctional and healthy adult tissues. Necrosis is an accidental, ‘passive’ cell death that is caused by chemical or physical stresses. Apoptosis is associated with a set of characteristic changes (called apoptotic remarks) such as the shrinking of cells, fragmentation of cells, fragmentation of nuclei, condensation and fragmentation of chromatin, fragmentation of DNA, reduction of the mitochondrial membrane potential (Matsuyama et al. 1998), release of cytochrome c, increased caspase activity, and focal phagocytic elimination of the cell by surrounding cells and macrophages (Kerr et al. 1994; Wyllie et al. 1980). All these changes are phenotypic manifestations of genetic regulation. In contrast, necrosis presents changes such as the swelling of cells, disintegration of nuclei, loss of chromatin, swelling and breakage of mitochondria, leakage of cytoplasm, and inflammation of neighbouring cells (Wyllie et al. 1980); these changes are independent of genetic regulation.
Although many think that unicellular organisms do not manifest genetically regulated cell death, debate on whether the unicellular eukaryote yeast undergoes apoptosis is now underway (Fraser & James 1998). Because no potential homologues to apoptotic genes in multicellular organisms (e.g. Bax, Bak, Bcl-2, caspases and p53) have been identified in either Saccharomyces cerevisiae or Schizosaccharomyces pombe, yeast has been thought not to undergo apoptosis. However, three recent results have suggested that yeast might have a mechanism for apoptosis. The first result which was suggestive of apoptosis in yeast was that expression of human Bax or Bak led to cell death in both S. cerevisiae and S. pombe (Sato et al. 1994; Zha et al. 1996; Ink et al. 1997; Ligr et al. 1998) and was inhibited by the co-expression of human Bcl-2 (Ligr et al. 1998; Xu & Reed 1998). Likewise, the introduction of ced-4 induced cell death in S. pombe (James et al. 1997) and S. cerevisiae (Tao et al. 1999). The second finding was that mutation of the cell division cycle gene CDC48 prompted the appearance of several apoptotic remarks in S. cerevisiae (Madeo et al. 1997). The third result was the identification of a homologous gene of the mammalian inhibitor of apoptosis (IAP) in both S. cerevisiae and S. pombe (Dereaux & Reed 1999).
To investigate whether yeast cells undergo apoptosis, we systematically analysed their morphological and functional phenotype by focusing on CCG1-interacting factor A (CIA1), for the following reasons. Human cell cycle gene 1 (hCCG1) is the largest subunit of TFIID (Hisatake et al. 1993; Ruppert et al. 1993), which is a key factor for transcriptional events in mammalian cells (Hai et al. 1988; Horikoshi et al. 1988a,b). hCCG1 itself regulates the mammalian cell cycle (Sekiguchi et al. 1991), and this factor appears to be a histone acetyltransferase (HAT; Mizzen et al. 1996). Moreover, a human cell line mutant in hCCG1 underwent spontaneous apoptosis; therefore, hCCG1 is involved in the regulation of apoptosis in mammalian cells (Sekiguchi et al. 1995).
We isolated a human factor CIA (hCIA) that interacts with hCCG1 and identified it as a histone chaperone (Munakata et al. 2000). hCIA is a human homologue of the S. cerevisiae anti-silencing factor Asf1p, which was identified previously (Le et al. 1997). To study the morphological and functional phenotype of cell death in yeast, we used the S. cerevisiae ASF1/CIA1 disruptant, asf1/cia1 and identified a cell-death phenotype that was largely characterized by classic features of apoptosis but that had some features indicative of necrosis. Therefore, it is necessary to create a new, comprehensive concept of cell death that encompasses both active and passive cell death systems so that cell-death processes in unicellular organisms can be linked with those in multicellular ones.
The asf1/cia1 yeast arrests during the cell cycle and dies preferentially at G2/M-phase
When cultured, the S. cerevisiae disruptant asf1/cia1 showed apparent slow growth. To identify the phase of the cell cycle that might direct this growth, we used flow cytometry with propidium iodide (PI; Haugland 1998) to analyse asf1/cia1 cells by nuclear staining. The flow cytometry profile of the wild-type cells showed a prominent peak at 1C, which corresponded to G1/S-phase cells, and a broad peak at 2C, which corresponded to G2/M-phase cells (Fig. 1). In comparison, the profile of the asf1/cia1 cells revealed a smaller population of G1/S-phase cells at 1C and a larger population of G2/M-phase cells at 2C (Fig. 1). Therefore, the asf1/cia1 cells arrested predominantly at the G2/M-phase.
To further study this phenomenon, we completed fluorescent viability staining with FUN-1 and Calcofluor White M2R (Haugland 1998) in order to observe the cells using confocal laser scanning microscopy (LSM). In the asf1/cia1 samples, we observed many dead cells, most of which were at the G2/M-phase. Typically these dead cells appeared to be enlarged, budding or dumbbell-shaped (Fig. 2). The viability of the asf1/cia1 cells decreased markedly over time, dropping to 83% after 12 h of culture and to 73% after a 72 h culture period. In comparison, the viability of wild-type yeast cells remained constant at ≈90% between 6 h and 72 h of culture. When the yeasts were subcultured every 72 h for 2 months, viability fell to ≈60% for asf1/cia1 cells, whereas that for the wild-type organism remained near ≈90%. Therefore, cell death, which occurred spontaneously in wild-type S. cerevisiae, may have been accelerated in the asf1/cia1 cells.
The asf1/cia1 cells were enlarged
We then questioned whether the cell death of the asf1/cia1 yeast cells was apoptotic or necrotic in nature. We felt that it was important to confirm the potential enlargement of the asf1/cia1 cells (Fig. 2) because it might be similar to a necrotic indicator in multicellular organisms (Wyllie et al. 1980). To this end, we studied the yeast cells using light microscopy in the differential interference phase contrast mode (DIC-mode), which enables visualization of an organelle, the vacuole, without staining. We confirmed that the asf1/cia1 cells were irregular in shape and size (i.e. these cells were enlarged to approximately 8 µm × 13 µm), whereas the wild-type cells were uniform in shape and size (approximately 5 µm × 8 µm; Fig. 3). From the micrographs, we also noted that the number and size of the vacuoles in the asf1/cia1 cells were increased in comparison with those of the wild-type yeast.
The dead asf1/cia1 cells have fragmented nuclei and condensed, fragmented chromatin
The enlargement of the asf1/cia1 cells suggested that they underwent a necrosis-like cell death. Because nuclear morphology can be used to discriminate between apoptosis and necrosis in multicellular organisms, we specifically stained the DNA of our yeast strains by using 4′,6-diamidino-2-phenylindole (DAPI; Haugland 1998). Subsequent fluorescent light microscopy (FLM) revealed an apparent fragmentation of the nuclei in the dead asf1/cia1 cells but not in the living wild-type cells, resembling a characteristic of apoptosis in multicellular organisms (Fig. 4).
Because transmission electron microscopy (TEM) reveals definitive ultrastructural features of the chromatin in apoptotic multicellular organisms (Kerr et al. 1994), we used this method to study our asf1/cia1 and wild-type yeasts. The dead asf1/cia1 cells manifested several changes which were characteristic of apoptosis, such as chromatin, condensed as small, dispersed particles (Fig. 5A–C), chromatin condensed as aggregates along the nuclear membrane (Figs 5A,C,D), and the nuclei fragmented (Fig. 5E–G). These nuclear changes did not occur in the living wild-type cells (Fig. 5H). We were unable to ascertain whether the fragmentation of the chromatin was shown in DNA ladders in either the asf1/cia1 or wild-type yeasts (data not shown).
The nuclei of apoptotic cells of multicellular organisms show various apoptotic remarks as the process proceeds. First, the chromatin condenses along the nuclear membrane and then the chromatin fragments, after which the nuclei fragment. The changes we observed in the asf1/cia1 yeast cells (Fig. 5A–G) were similar to the characteristics described above for multicellular organisms. Perhaps more importantly, we often saw these changes in dead wild-type yeast cells. Therefore, these nuclear changes suggest that yeasts undergo an apoptosis-like cell death.
The vacuoles and mitochondria in asf1/cia1 yeast cells display functional changes
Using DIC-mode microscopy, we revealed the increased number and enlarged size of the vacuoles in asf1/cia1 yeast cells (Figs 3A, B). Vacuoles are normally acidic and are the final destination of biological materials, which are taken there by endocytosis. Therefore, we used quinacrine mustard to assess the acidity of the vacuoles in asf1/cia1 cells and Lucifer Yellow (Haugland 1998) to evaluate their endocytotic activity. We found that the acidity of the vacuoles in asf1/cia1 cells was decreased, whereas that of the cytoplasm had increased (Figs 6A, B). Furthermore, although the uptake of Lucifer Yellow into the cytoplasm of the asf1/cia1 cells was comparable to that of the wild-type yeast, the uptake of the dye from the cytoplasm into vacuoles was reduced dramatically in asf1/cia1 cells (Figs 6C, D). Therefore, the vacuoles in asf1/cia1 showed both morphological and functional changes.
The remarkable change in the acidity balance between the vacuoles and cytoplasm of asf1/cia1(Figs 6A, B) suggested a reduction of the vacuolar membrane-potential due to dysfunction in the vacuolar proton pump. A reduction in mitochondrial membrane-potential is a characteristic of apoptosis in multicellular organisms (Matsuyama et al. 1998). To study the vacuolar and mitochondrial membrane-potentials in asf1/cia1, we used 3,3′-dihexyloxacarbocyanine iodide (DiOC6; Haugland 1998) to stain the membranes of these organelles. In the asf1/cia1 cells, vacuoles were seen as circles with indistinct, green-coloured borders, and the mitochondria appeared as dimly green-coloured dots (Fig. 6E). In contrast, the vacuoles of the wild-type cells were seen as circles with distinct green borders, and the mitochondria were bright green dots (Fig. 6F). Therefore, the membrane-potentials of the vacuoles and mitochondria of the asf1/cia1 cells were reduced, which suggests that their death was apoptosis-like.
Release of cytochrome c to cytoplasm
The reduction in mitochondrial membrane-potential and the release of cytochrome c are associated with the early stages of multicellular apoptosis. The release of cytochrome c to the cytoplasm is an apoptotic trait, which may otherwise not be ubiquitous (Green & Reed 1998). Therefore, we examined if the release was discernible in asf1/cia1 yeast cells. We prepared spheroplasts of both asf1/cia1 and wild-type cells and isolated mitochondria from the spheroplasts according to previously reported methods (Law et al. 1995). We then spectrophotometrically determined the cytochrome c that remained in the mitochondria and leaked from the mitochondria to the medium in the presence of a membrane-destroying agent nystatin, as previously reported (Manon et al. 1997). The quantity of cytochrome c in the mitochondria was reduced and that in the medium was increased for the asf1/cia1 cells in comparison with the wild-type cells (Fig. 7, left). The sum of the cytochrome c in the mitochondria and in the media was nearly the same between the asf1/cia1 cells and the wild-type cells. Based on these data, we estimated a release ratio of cytochrome c from the inner mitochondrial membrane to be about 65% and 20% for the asf1/cia1 and the wild-type cells, respectively (Fig. 7, right (insert)). Thus, the release of cytochrome c was significant in asf1/cia1 cells.
TEM reveals autophagic bodies in the vacuoles of non-starved asf1/cia1 cells
Autophagic bodies are often seen in the vacuoles of starved yeast cells (Takeshige et al. 1992). Vacuoles in yeast have a similar function to that of lysosomes in multicellular organisms. In addition, autophagolysosomes, which are lysosomes having autophagic bodies in them, of multicellular organisms often appear when the cells undergo starvation or apoptosis (Clark 1990). TEM revealed discernible autophagic bodies in the vacuoles, along with the chromatin condensed in the nuclei, of asf1/cia1 cells that were grown in a eutrophic culture medium (Fig. 8) but not in wild-type cells grown under similar conditions. Therefore, the presence of autophagic bodies in the vacuoles of asf1/cia1 cells is another indicator of apoptosis-like cell death, although multicellular animals lack vacuoles.
Altered expression of vacuolar and mitochondrial proton pump related genes in asf1/cia1 cells
In light of the results we obtained, the vacuoles and mitochondria in asf1/cia1 were dysfunctional. We used Gene Chip systems to analyse the expression of genes associated with the proton pumps of these organelles. The expression of several genes associated with the vacuolar proton pump, including VMA7, VMA8, VMA10, VMA21 and VMA22 (Stevens 1997) was decreased in asf1/cia1 cells. Note that VMA21 and VMA22 form the integral membrane sector of the ATPase in the endoplasmic reticulum (Fig. 9). In addition, we identified an apparently enhanced expression of mitochondrial proton pump related genes in asf1/cia1 cells, including ATP1 and ATP5 (data not shown).
Yeast undergoes a peculiar cell death, being largely apoptotic but slightly necrotic
To answer the question of whether yeast undergoes apoptosis, we systematically analysed the phenotype of yeast cells lacking CCG1-interacting factor A (asf1/cia1). Firstly, both the mutant and wild-type strains of yeast died spontaneously, and the death was accelerated remarkably in the asf1/cia1 cells, with a preferential arrest at the G2/M-phase (Figs 1 and 2). Secondly, the phenotype of these dead mutant cells comprised a set of changes that was largely similar to that manifested by apoptotic cells of multicellular organisms but that also included some features of necrosis (Table 1). The enlargement of the dead cells (Fig. 3) is possibly a necrotic characteristic. Our preliminary results on osmotic resistance suggested that the cell walls of the dead asf1/cia1 cells were about three times as fragile as those of the living wild-type cells (data not shown). The fragility of the cell wall in these yeast cells might lead to a reduced osmotic resistance, as is seen in necrotic mammalian cells. The fragmentation of nuclei (Fig. 4) is a characteristic of multicellular apoptosis. Shrinkage of the nuclei, which often accompanies fragmentation in apoptosis, was not obvious in the asf1/cia1 cells, and we suspect this to be due to a difference in structure between multicellular and yeast nuclei. The condensation and fragmentation of chromatin (Fig. 5) is also a characteristic of apoptosis. However, DNA ladders were absent.
Table 1. Comparison of changes seen in cell death between yeast and multicellular organisms.
+, present; –, absent; ?, unknown; nd, not determined.
Many authors have thought that the reduction in mitochondrial membrane-potential (Figs 6E, F) is a trait of apoptosis (Matsuyama et al. 1998), although a recent report has indicated that such a reduction can also occur in necrotic mammalian cells (Matsumura et al. 2000). In conjunction with the altered expression of mitochondrial proton pump related genes, this reduced potential indicates a dysfunction in the mitochondrial proton pump—another apoptotic trait. In addition, the extensive dysfunction of vacuoles, as reflected by their altered size and number (Figs 6D and 3), acidity balance (Figs 6A, B), endocytosis (Figs 6C, D), membrane-potential (Figs 6E, F), and gene expression (Fig. 9) in the asf1/cia1 cells, leads to a crucial perturbation of intracellular homeostasis. In multicellular organisms, a perturbation of this magnitude and scope triggers apoptosis (Kerr et al. 1994). The release of cytochrome c(Fig. 7) is a highly positive apoptotic trait (Green & Reed 1998). Furthermore, the autophagic bodies seen in the vacuoles of asf1/cia1(Fig. 8) are analogous to those in the autophagolysosomes of apoptotic cells of multicellular organisms (Clark 1990). Because yeast is unicellular, the presence of autophagic bodies in vacuoles suggests that this organelle has an important phagocytic role during cell death. In short, changes similar to apoptotic indicators predominate in asf1/cia1 cells (Table 1). Moreover, there is current consensus over the absence of caspase activity and DNA ladders in yeast (Fröhlich & Madeo 2000).
As described in the Introduction, expression of Bax or Bak or ced-4 in yeast led to cell death (Sato et al. 1994; Zha et al. 1996; Ink et al. 1997; James et al. 1997; Ligr et al. 1998; Tao et al. 1999). All these studies similarly provided phenotypes in yeast that were consistent with the fragmentation of nuclei, condensation and fragmentation of chromatin (these are apoptotic), and the absence of both DNA ladder and caspases (these are necrotic). Thus the phenotype from the deletion of ASF1/CIA1 is consistent with that which results from over-expression of the apoptotic gene in yeast (see Table 1). To reinforce this statement, however, other changes when ced-4 is introduced in yeast should be investigated in the near future. In addition, a reactive oxygen species induces active cell death process in another unicellular organism, Trypanosoma (Ridgley et al. 1999). These findings suggest that many unicellular organisms have a regular system for apoptosis. In the light of our findings and together with this suggestion, we infer that yeast undergoes an active cell death which retains some features of passive cell death. We speculate that in yeast, active cell death is not fully differentiated from passive cell death, and we therefore call the type of death manifested in the asf1/cia1 cells ‘prototypal active cell death’.
A proposed comprehensive concept that links the cell death of multicellular organisms with that of unicellular ones
Our surprising inferral—that yeast undergoes a cell death with predominant apoptotic features—inevitably calls for a new, comprehensive concept of cell death that encompasses active cell death (i.e. apoptosis) and passive cell death (i.e. necrosis), thereby linking these processes in multicellular organisms with those in unicellular ones. Many authors have thought that the original process of cell death was purely passive and that active cell death developed and was refined whilst evolution from unicellular to multicellular organisms took place. However, this popular concept may be somewhat limited, because cell death in yeast very likely retains some of the characteristics of the original, active mode of cell death. Instead, we propose that the original type of cell death was both active and passive in character and that active cell death differentiated from this original mode as evolution progressed. Therefore, we suggest that active cell death, which is under genetic regulation, is the essential attribute of the overall death process and ubiquitously functions to preserve sound individuals in all species for all organisms. To test this hypothesis, an extensive study on the cell death mechanisms of prokaryotes, including eubacteria and archaebacteria, as well as of unicellular eukaryotes, is needed.
Interestingly, the principal signalling factors for apoptosis, such as TNF (Vercammen et al. 1998) and Fas (Matsumura et al. 2000), also induce necrosis in multicellular organisms. Because these observations suggest an intricate interplay between apoptosis and necrosis mechanisms in multicellular cells, we think that it is possible that an active cell death system works together with a passive cell death system in yeast. Yeast may have evolved a prototypal active cell death system in which the vacuole is responsible for phagocytic elimination, and this system may have been refined to govern the sophisticated events of apoptosis during development, differentiation and cell turnover in multicellular organisms.
Abnormal assembly/disassembly of nucleosomes may cause cell death in yeast
Our study focuses on analysing the phenotype of death in yeast; therefore it is too early to propose a completely plausible model of the mechanism by which active cell death occurs in yeast. However, we speculate that because deletion of the histone chaperone ASF1/CIA1 apparently stimulated the active cell-death mechanism naturally occurring in yeast, this chaperone may (to a greater or lesser degree) affect the overall system by which cell death occurs. Therefore, we suggest that deletion of ASF1/CIA1 inhibits the normal assembly/disassembly of the nucleosome, which relies on Asf1p/Cia1p; consequently, deletion of this factor stimulates the cell-death system. In addition, hCIA probably interacts with hCCG1 in mammalian cells (Munakata et al. 2000). hCCG1 is a regulator of transcriptional events in mammalian cells (Hisatake et al. 1993; Ruppert et al. 1993) and is involved in the regulation of apoptosis (Sekiguchi et al. 1995). Furthermore, a mutation in S. cerevisiae CCG1 decreased the expression of vacuolar proton pump related genes (Holstege et al. 1998). Therefore, the deletion of ASF1/CIA1 modifies these functions of CCG1 and, consequently, initiates the cell death mechanism. CCG1 is a HAT (Mizzen et al. 1996), and our recent study showed that Tip60, another member of the HAT family (Yamamoto & Horikoshi 1996), is a regulator of apoptosis in mammalian cells (Ikura et al. 2000). We therefore suggest that factors responsible for regulating the assembly/disassembly of nucleosomes (e.g. histone chaperone and HAT) may be intimately involved in the cell death mechanism of yeast.
From Sigma Chemical Co. (St. Louis, MO, USA), we purchased propidium iodide (PI), 4′,6-diamidino-2-phenylindole (DAPI), Lucifer Yellow, and RNase A (bovine pancreas). From Molecular Probes Inc. (Eugene, OR, USA), we purchased FUN-1, Calcofluor White M2R dye, 3,3′-dihexyloxacarbocyanine iodide (DiOC6), and quinacrine mustard. From Nakalai Chemical Co. Ltd. (Tokyo, Japan), we purchased Lowicryl k4M resin, acrolein and glutaraldehyde.
Yeast strains and culture
ASF1/CIA1 was cloned as an approximately 1.3 kb PCR fragment using W303 genomic DNA as a template. The primers used in the reaction were ASF1/CIA1-1 (5′-ATCGGCTTTGTGCCACACCTAACC-3′) and ASF1/CIA1-2 (5′-CTGGGGCCCTAACCCGACTTTATCTTATGTC-3′). This 1.3 kb fragment was subcloned into pBluescript II SK which had been digested with EcoRV; the resulting construct then was cut with HincII and PstI. The HincII-PstI fragment was ligated into pBluescript II SK that had been digested with PvuII, and the resulting plasmid was named pASF1/CIA1.
To construct pDeltaASF1/CIA1U, a 1.1 kb URA3-containing HindIII-SmaI fragment from pJJ244 was used to replace the 0.7 kb NdeI-AccI fragment of pASF1/CIA. To construct the ASF1/CIA1 gene disruptant, pDeltaASF1/CIA1U was digested with EcoRI and HindIII, and the resulting 1.1 kb fragment was used to transform the MATa/α diploid W303, and Ura+ colonies were selected. Correct integrants were identified by genomic PCR analysis. Mutant haploid cells were isolated from the sporulated diploids, and the genotypes of these mutant cells were confirmed by verifying their Ura+ phenotype and the results of the genomic PCR analysis.
The disruptant and wild-type yeasts were cultured on YPDA plates (containing 1% yeast extract, 2% Bacto-peptone, 4% dextrose, 4% glucose and 0.02% adenine) until the cells reached a stationary phase after 48–72 h of fermentation at 30 °C. To obtain a 2-month culture, the cells were subcultured every 72 h.
Flow cytometry analysis of the asf1/cia1 and wild-type yeast cells used the PI nuclear staining method (Haugland 1998) at 25 °C on a FACS Vantage model (Becton-Dickinson, San Jose, CA, USA) that has a 575/26 bandpath filter. Yeast isolated from a stock strain (kept at −80 °C in glycerol) was cultured in liquid YPDA medium at 30 °C until the 600 nm optical density of the medium reached ≈0.5. The yeast cells then were rinsed five times with sterile water and collected by centrifugation (13 000 r.p.m. for 5 min). The washed cells were resuspended in 300 µL of sterile water and were then dehydrated and fixed by the addition of 700 µL of 100% ethyl alcohol to the suspension. After incubation at room temperature for 30 min, the cells were collected by centrifugation, resuspended in 500 µL of RNase A solution (0.1 mg RNase A/mL in 50 mm sodium citrate [pH 7.0]), and placed at 37 °C for 3 h to degrade the RNA. The resulting RNA-free cells were stained with PI (final concentration, 10 µm; Haugland 1998).
Light microscopy involved an Axioplan microscope (Karl-Zeiss, Jena, Germany) used in the differential interference phase contrast mode on unfixed samples. Yeast cells were cultured on YPDA plates at 30 °C. A colony at the stationary growth stage (i.e. after 72 h of culture) was isolated and resuspended in 1 mL of 50 mm PBS (pH 7.2). The yeast cells then were collected by centrifugation, and the pelleted cells were resuspended in 1 mL of 50 mm PBS. 2 µL of the suspension was placed on a glass slide, covered with a coverslip, and examined by light microscopy.
Fluorescent and confocal laser scanning microscopy
Fluorescent light microscopy (FLM) involved an Axioplan microscope (Karl-Zeiss) with a filter that had bandpath (340–380 nm) and longpath (> 425 nm) filtering capability. Confocal laser scanning microscopy (LSM) involved an LSM510 model microscope (Karl-Zeiss) and a TCS SP microscope (Leica, Heerbrugg, Switzerland) that were equipped with He-Ne lasers.
Cell viability staining
A yeast colony at the stationary growth stage (cultured on a YPDA plate at 30 °C for 72 h) was resuspended in 1 mL of 10 mm Na-HEPES buffer (pH 7.4) containing 2% glucose. Pelleted cells obtained by centrifugation were resuspended in 1 mL of the same buffer. The cell suspension was first subjected to cell-viability staining by the addition of 1 µL of FUN-1 (Molecular Probes Inc.) and 5 µL of Calcofluor White M2R (Molecular Probes Inc.) to the suspension, which then was cultured at 30 °C for 15–30 min The cells were rinsed once with 10 mm Na-HEPES buffer (pH 7.4) containing 2% glucose and then resuspended in 50 µL of the same buffer. 2 µL of the cell suspension was placed on a glass slide, covered with a coverslip, and immediately examined using LSM (TCS SP model, Leica).
Yeast cells obtained from a colony at the stationary growth stage (cultured on a YPDA plate at 30 °C for 72 h) were rinsed with 1 mL of 50 mm PBS. The resulting pelleted cells were resuspended and then fixed with 3.7% formaldehyde by incubating at room temperature for 1–2 h. The cells were rinsed three times with 1 mL of distilled water and then resuspended in 1 mL of distilled water. 1 µL of the cell suspension was placed on a glass slide, stained with 1 µL of DAPI (1 mg/mL), covered with a coverslip, and examined by FLM (Axioplan model, Karl-Zeiss).
Transmission electron microscopy
Transmission electron microscopy (TEM) involved a JEM-1010 model (JEOL, Tokyo, Japan) that had an accelerating voltage of 80 kV (magnification: ×10 000–30 000). TEM micrographs were taken with an exposure time of 1.5 s. Yeast cells at the stationary growth stage were fixed at 4 °C for 2 h using 2% acrolein containing 0.25% glutaraldehyde. The cells were dehydrated using a graded alcohol series (from 50% to 100%) and were then embedded in Lowicryl k4M resin at −45 °C for 48 h. The cells were cut into ultra-thin sections (thickness ≈ 70 nm) using an Ultra S microtome (Leica); the sections were mounted on a nickel grid, stained with 2% uranyl acetate and 0.4% lead citrate, and examined by using TEM.
Assessing the membrane potential of vacuoles and mitochondria
For the membrane potential staining (Haugland 1998), plated yeast cells at the stationary growth stage were recultured in liquid YPDA at 30 °C for 15–30 min in the dark in the presence of DiOC6 (final concentration, 100 ng/mL). 2 µL of the cell suspension was placed on a glass slide, covered with a coverslip, and examined using an LSM (LSM510 model, Karl-Zeiss).
Plated yeast cells at the stationary growth stage were recultured in liquid YPDA containing quinacrine mustard (final concentration, 25 mg/mL) for 30 min at 30 °C. 2 µL of the cell suspension was placed on a glass slide, covered with a coverslip, and examined using an LSM (LSM510 model, Karl-Zeiss).
Plated yeast cells that underwent secondary liquid culture in the presence of Lucifer Yellow (5 mg/mL) at 30 °C for 30 min were used to assess endocytosis activity. 2 µL of the cell suspension was placed on a glass slide, covered with a coverslip, and examined using an LSM (LSM510 model, Karl-Zeiss).
Assay for release of cytochrome c
We isolated mitochondria from the yeast cells according to previously reported methods (Law et al. 1995). Briefly, we used yeast cells that had been cultured in YP liquid media containing 1% yeast extract, 0.5% gelatin hydrolysate, 0.1% potassium phosphate, 0.12% ammonium sulphate, and 2% d-lactose (pH 5.0). The cells were harvested by centrifuge at 4500 r.p.m. for 5 min, and twice washed with an ice-cold 0.1% galactose-water solution. We made spheroplasts of these cells by using cytohelicase in a medium that contained 0.01 m citrate, 0.01 m disodium phosphate, 1.35 m sorbitol, and 1 mm EGTA (pH 5.8), collected them by centrifuge in a solution containing 0.01 m Tris-maleate, 0.75 m sorbitol, 0.4 m mannitol, 2 mm EGTA, and 0.1% BSA (pH 6.8). We isolated mitochondria from the spheroplasts by combinations of homogenization with a low-speed (2000 r.p.m)/high-speed (12 500 r.p.m) centrifuge.
After adjusting the concentrations of mitochondria thus isolated, we determined the cytochrome c remaining in the mitochondria and was leaked from the mitochondria to the medium by a membrane-destroying agent, nystatin, according to the previously reported methods (Manon et al. 1997). We determined quantities of cytochrome c by measuring absorbance at the Soret band (namely, γ peak) with a Beckman DU640 spectrophotometer.
Gene expression analysis
The expression levels of various vacuolar and mitochondrial proton pump-related genes in the asf1/cia1 and wild-type S. cerevisiae were determined using the GeneChip analysis method, which involved a set of four oligonucleotide arrays (GeneChip Ye6100 arrays, Affymetrix, Santa Clara, CA, USA) containing probes for 6218 yeast ORFs; the data analysis used the GeneChip basic software (Affymetrix). Yeast cells that had been cultured at 30 °C in 10–20 mL of YPDA medium and had reached a density of 1 × 107 cells/mL were used. RNA was extracted from duplicate samples of these cells by using glass beads (Sigma Chemical Co.) and was affinity-purified by using Oligotex dT30 Super (JSR, Tokyo, Japan). The purified RNA (1–1.5 µg) was reverse-transcribed to its cDNA by using T7-(dT)24 primers and SuperScript II (Gibco BRL, Rockville, MD, USA). The cDNA was purified by using the phenol-chloroform extraction method and Phase-Lock Gel.
The cRNA from a 1 µg aliquot of this cDNA was synthesized and biotin-labelled by using a MEGAscript T7 kit (Ambion, Austin, TX, USA) and incubation at 37 °C for 4–6 h with Bio-11-CTP (no. 42818, Enzo, New York, NY, USA), Bio-16-UTP (no. 42814, Enzo), and NTP substrates (Ambion). The cRNA was purified by using an RNeasy column (no. 74104, Qiagen, Bothell, WA, USA) to yield 40–44 µg and was fragmented by incubation at 94 °C for 35 min in a buffer of 40 mm Tris-acetate (pH 8.1), 100 mm potassium acetate and 30 mm magnesium acetate. These cRNA fragments were heated at 99 °C for 5 min and hybridized to the oligonucleotide arrays, washed, stained and scanned according to previously reported methods (Wodicka et al. 1997). The resulting samples then underwent GeneChip (Affymetrix) analysis according to the manufacturer's recommendations.
We thank all the members of the Horikoshi Gene Selector Project and the Laboratory of Developmental Biology at The University of Tokyo for their fruitful discussions of this research. This work was supported in part by Grants-in-Aid for Science Research from the Ministry of Education, Culture, Sports, Science and Technology of Japan, and the Exploratory Research for Advanced Technology (ERATO) of Japan Science and Technology (JST).