Fission yeast protein kinase C gene homologues are required for protoplast regeneration: a functional link between cell wall formation and cell shape control

Protein kinase C (PKC) plays a central role in the signal transduction pathways that control various cellular processes (Coussens et al., 1986; Nishizuka, 1986). Many subspecies of PKC have been discovered in various mammalian tissues (Coussens et al., 1986; Nishizuka, 1988; Ono et al., 1988) and other higher eukaryotes (Schaeffer et al., 1989; Otte and Moon, 1992). PKC gene homologues have also been found in lower eukaryotes (Levin et al., 1990; Ravid and Spudich, 1992; Toda et al., 1993), confirming that the PKC family occurs in all eukaryotes. Although members of the PKC family are divided into three major subtypes, depending on the structure of the protein, their enzymatic properties, reaction to ligands, and pattern of expression, the functional distinction between different PKC subspecies in higher and lower eucaryotes remains unclear.

Yeasts are unicellular eukaryotes and have several advantages as experimental systems. Their simple systems make it easier to study complex problems. Recent developments in molecular genetics and the availability of various mutants have reinforced the value of these organisms in the study of fundamental cellular processes. Two yeasts have been used extensively: the budding yeast Saccharomyces cerevisiaeand the fission yeast Schizosaccharomyces pombe. Only one PKC gene homologue, PKC1, has been cloned in S. cerevisiaeeven though different cloning methods have been tried (Levin et al., 1990; Yoshida et al., 1992). Two novel PKC (nPKC) gene homologues, pck1⁺and pck2⁺, were recently isolated from S. pombe(Toda et al., 1993). These two genes share an overlap-ping function essential for cell viability and cells of a single pck2⁺disruptant (Δpck2) display severe defects in cell shape. PCK1in S. cerevisiaeis essential for viability, but the lethality caused by cell lysis is reversed by the inclusion of osmotic stabilizers (Levin and Bartlett-Heubusch, 1992; Paravicini et al., 1992). These results suggest that PKCs in yeasts may play a direct or indirect role in cell wall formation.

As we reported previously, actin is associated with initiation of cell wall formation, the proper deposition of cell wall materials, and maintaining the normal morphology of regenerating protoplasts of S. pombe(Kobori et al., 1989). We are interested in the involvement of PKC in actin organization.

In this study, regenerating protoplasts of S. pombewere used as a simple model to clarify the involvement of PKCs in cell wall formation and actin organization. The protoplasts are devoid of all original wall materials (Nečas and Svoboda, 1985) and S. pombeis one of the few yeasts that can regenerate the cell wall and revert to intact cells in liquid medium. We have already clarified the successive morphological events during the regeneration of S. pombeprotoplasts by several electron microscopic techniques. The process of secretion and formation of β-glucan and α-galactomannan, components of the cell wall of S. pombe, during protoplast regeneration was elucidated by ultra-high resolution and low-voltage scanning electron microscopy (Osumi et al., 1990, 1992). Contrast enhancement by ruthenium tetroxide revealed the interior and exterior ultrastructure of regenerating protoplasts of S. pombewith a transmission electron microscope (Naito et al., 1991; Osumi, 1992). These studies confirmed that the regenerating protoplast of S. pombeis a good model system to clarify the mechanisms of cell wall formation and secretion of polysac-charides of yeasts. Here, the ability to regenerate the cell wall was compared with that of a wild-type strain and single gene disruptants of pck1⁺(Δpck1) and pck2⁺(Δpck2). We show that the two genes of nPKC in S. pombehave different functions in protoplast regeneration. The pck2⁺gene is required for the regeneration while the pck1⁺gene is not. The function of nPKC gene homologues in cell wall formation, actin organization, cell shape control, and growth polarity are also discussed.

Staurosporine, an antibiotic alkaloid derived from Strepto-myces staurospores, is reported to be a potent inhibitor of various protein kinases such as PKCs, cAMP-dependent protein kinases, casein kinase II, and src tyrosine kinase, although PKC is believed to be one of its most potent in vivo targets (Omura et al., 1977; Tamaoki et al., 1986; Nakano et al., 1987). Staurosporine-supersensitive mutants isolated from S. pombeand S. cerevisiaeindicate that staurosporine sensitivity is genetically complex (Toda et al., 1991; Yoshida et al., 1992). We have shown previously that pck2⁺is a molecular target of staurosporine in fission yeast (Toda et al., 1993). We also confirmed that the pck2⁺gene product is an in vivo target of staurosporine in protoplast regeneration.

Rhodamine-conjugated phalloidin was obtained from Molecular Probes (Junction City, OR, USA). 4′-6-diamidino-2-phenylindole (DAPI) was purchased from Sigma Chemical Co. (St Louis, MO, USA). Calcofluor White M2R was purchased from Polysciences, Inc. (Warrington, PA, USA), and NovoZym 234 from Novo Nordisk (Bagsvaerd, Denmark). Staurosporine was purchased from Sigma Chemical Co.

The strains of S. pombeused were the wild-type strain (L972 h−), the Δpck1strain (TP134-3B) (L972 h− leu1 ura4 pck1 :: ura 4⁺), and the Δpck2strain (TP170-2B) (L972 h− leu1 pck2:: LEU2) (Toda et al., 1993). The cells were cultured at 30°C in YPD broth (1% yeast extract, 2% polypeptone and 2% glucose) to early log phase. Plating efficiency was determined by directly counting the yeasts and counting the colonies appearing on YPD agar among the intact cells in early log phase.

Cells were washed once with E buffer (50 mM sodium citrate, 100 mM sodium phosphate buffer, pH 6.0). To form protoplasts, the cells were incubated with 5 mg/ml NovoZym 234 in E buffer containing 1.2 M sorbitol. The frequency of protoplast formation was determined by counting the protoplasts under a phase-contrast microscope. To regenerate protoplasts, the cells that formed them were harvested by centrifugation, inoculated into YPD broth containing 1.2 M sorbitol (regeneration medium) (Kobori et al., 1989), and allowed to regenerate in the regeneration medium at 30°C with aeration for 13 hours. The frequency of protoplast regeneration was determined by harvesting the regenerating protoplasts at 13 hours, diluting them with sterilized water to make them burst and plating them on hypotonic YPD agar. The NovoZym-treated samples were also inoculated into YPD broth and then plated on YPD agar at 0 hours and 13 hours to estimate the number of intact cells present in the samples. The number of colonies formed from intact cells was subtracted from the number of colonies formed from protoplasts to calculate the frequency of regeneration.

The effect of staurosporine on regeneration of protoplasts of the wild-type, Δpck1and Δpck2strains was determined by incubating protoplasts in liquid regeneration medium and liquid YPD medium containing various staurosporine concentrations (0 to 8.0 μg/ml) for 13 hours. Staurosporine was dissolved in dimethyl sulfoxide and the final concentration of the solvent in the medium was adjusted to 0.1% (v/v). The effect of dimethyl sulfoxide was tested separately as a control and the concentration of the solvent used had no detectable effect on the protoplast regeneration or actin configuration.

The 13-hour-regenerating protoplasts were fixed with 3.7% formaldehyde and 50 mM KP buffer (potassium phosphate buffer, pH 6.5) for 1.5 hours, and were washed three times with PBS (50 mM potassium phosphate buffer, pH 7.3, plus 150 mM of NaCl) (Kobori et al., 1989). The cell wall materials and the nuclei were made visible by staining the fixed samples with calcofluor (50 μg/ml) and DAPI (1 μg/ml), respectively. To make the F-actin visible, the samples were permeabilized with 50 mM of β-mercaptoethanol in PBS and 1% Triton X-100 in PBS and then stained with 0.3 μM rhodamine-conjugated phalloidin (Kobori et al., 1989). The samples for fluorescence microscopy were suspended in a drop of mounting medium containing p-phenylene diamine (1 mg/ml) in KP buffer to diminish the quenching of fluorescence (Johnson and Nogueira Araujo, 1981). Slides were viewed and photographed with an Olympus BM-2 microscope equipped with an epifluorescence attachment.

The severe defects in cell shape observed in pck2⁺null mutants of S. pombeprompted us to examine cell wall formation in the wild-type, Δpck1, and Δpck2strains using protoplasts. The frequency of protoplast formation in all three strains of S. pombewas 99.8, 99.6 and 99.9%, respectively, and their plating efficiency was 84, 92 and 93%. When the protoplasts of the wild-type strain (5×10⁶/ml) or those of Δpck1(5×10⁶/ml) were inoculated into an osmotically stabilized liquid regeneration medium and allowed to regenerate for 13 hours, 3.2×10⁶/ml and 1.4×10⁶/ml, respectively, regenerated the cell wall (Table 1). However, none of the 4.8×10⁶protoplasts of Δpck2per ml inoculated into the regeneration medium had regenerated the cell wall by 13 hours (Table 1). This striking difference between the Δpck1 and Δpck2strains indicates that the pck2⁺gene is essential for protoplast regeneration while the pck1⁺gene is not. It is also worth noting that the frequency of protoplast regeneration in Δpck1was half of that in the wild-type strain (Table 1). This implies that the pck1⁺gene has a minor role in protoplast regeneration, although it is not essential for the process.

The morphological characteristics of the 13-hour-regenerat-ing protoplasts of the three strains were examined, and the results of phase-contrast microscopy, of calcofluor staining for newly formed cell wall materials, of actin staining for cytoskeletal changes, and of DAPI staining for nuclei are shown in Fig. 1. Newly formed cell wall materials were seen in the wild-type and the Δpck1strains (Fig. 1A2,B2). The actin was reorganized, that is, actin dots were concentrated where the septa and new cell wall materials were formed (Fig. 1A3,B3). Nuclear division took place and the two nuclei were divided by the septum in most of the regenerated protoplasts (Fig. 1A4,B4). This shows that regeneration in the Δpck1strain is the same as that in the wild-type strain and that the lack of the pck1⁺gene does not affect the morphological events involved in the regeneration of protoplasts of S. pombe. In contrast, in the Δpck2strain there was no noticeable regeneration of protoplasts (Fig. 1C1), fluorescence in the calcofluor-stained specimen (Fig. 1C2) or dynamic change in actin distribution (Fig. 1C3) although the cells increased in size and the nucleus divided (Fig. 1C4). This confirmed that the pck2⁺gene is an essential component for protoplast regeneration in S. pombe.

Staurosporine is reported to inhibit various types of protein kinases. If nPKCs are in vivo targets of staurosporine in S. pombe, staurosporine-treated wild-type cells might show similar defects in both protoplast regeneration and morphology to those observed in the Δpck2strain. We determined the effect of staurosporine on protoplast regeneration. The minimum concentration causing complete arrest of protoplast regeneration in the three strains was 3 μg/ml (Table 2). An apparently identical failure to regenerate was seen in staurosporine-treated protoplasts of wild-type and Δpck1strains, suggesting that the pck2⁺gene product is a major target of staurosporine in vivo. The morphological characteristics of the three strains in the presence of staurosporine were examined to reinforce the assumption that nPKCs are in vivo targets of staurosporine in S. pombe. Photographs of the 13-hour-regenerating protoplasts incubated in the presence of staurosporine (3 μg/ml) are shown in Fig. 2. The morphological characteristics of the wild-type (Fig. 2A1-A4), Δpck1(Fig. 2B1-B4), and Δpck2(Fig. 2C1-C4) cells remained the same: (1) the shapes of regenerating protoplasts did not change; (2) new cell wall materials were not formed; (3) the actin was not reorganized; (4) nuclear division was not arrested; and (5) the volume of the cytoplasm increased. These striking similarities confirmed the assumption that the gene product of pck2⁺is a major in vivo target of stau-rosporine in S. pombe. It is also interesting to note that nuclear division was not arrested in the wild-type strain by this concentration of staurosporine.

The pck2⁺gene is essential for protoplast regeneration. The protoplast regeneration in S. pombeincludes the process of cell wall regeneration as well as the process of reversion of spherical protoplasts to rod-shaped cells. Since the protoplasts of the pck2⁺null mutant remained spherical and no cell wall materials were formed during incubation in the regeneration medium, pck2⁺would be involved in these two processes.

The pck1⁺gene is not essential for protoplast regeneration; however, it would have a minor role because the frequency of protoplast regeneration in the Δpck1strain was half of that in the wild strain as shown in Table 1. The pck1⁺gene would also have a minor role in cell wall formation because scarcely any of the loosely bound cell wall materials were formed in the 13-hour-regenerating protoplasts in the single gene disruptant of the pck2⁺gene. These cell wall materials were seen when non-fixed samples were observed under a phase-contrast microscope (data not shown), although they were easily removed during the procedures used in this experiment to prepare samples, such as fixation and washing of the samples by gentle centrifugation. From these results it is reasonable to conclude that both pck1⁺and pck2⁺genes are involved in the process of protoplast regeneration and cell wall formation, although the role of pck2⁺is much greater than that of pck1⁺. This is consistent with previously reported phenotype differences between Δpck1and Δpck2. Δpck2shows more severe defects in cell shape and staurosporine sensitivity (Toda et al., 1993).

The pck1⁺and pck2⁺genes in S. pombehave been demonstrated to share an overlapping function essential for cell viability (Toda et al., 1993). Cells harbouring a disruption of either pck1⁺or pck2⁺genes are viable and only double disruptants of both pck1⁺and pck2⁺genes are lethal. In this study, we have demonstrated that the two differ in the functions essential for protoplast regeneration. Disruptants only of pck2⁺genes lose their ability to regenerate protoplasts. This is an interesting feature of PKCs, which demonstrates that the two subspecies of PKC homologues in S. pombeshare a function essential for viability, while at the same time differing in the function essential for protoplast regeneration. The two PKCs in S. pombeare therefore multifunctional enzymes. Use of regenerating protoplasts can thus be viewed as a more sensitive assay to elucidate the function of PKCs than the assay of viability or growth in S. pombe.

We cannot exclude the possibility that the functional difference between pck1⁺and pck2⁺genes is merely due to a quantitative difference in the molecular products of the two genes. That is, both genes code functionally identical proteins, but the amount of pck1 produced is less than that of pck2. The amount of pck1 would be enough to support the normal growth of the intact cells of S. pombe, but it would not be enough to regenerate protoplasts. Our results strongly support the assumption that the two genes are qualitatively different from each other. However, a definite answer to the question will be obtained by further study to determine whether overexpression of pck1 sup-presses the cell wall defect of Δpck2.

The spatial coincidence between actin distribution and cell wall growth during the cell cycle is well documented in various yeasts (Marks and Hyams, 1985; Adams and Pringle, 1984; Kilmartin and Adams, 1984; Anderson and Soll, 1986; Kobori et al., 1992). Involvement of actin in the determination of growth polarity was demonstrated in S. pombe(Marks and Hyams, 1985; Marks et al., 1986; Kobori et al., 1989). Involvement of actin in cell shape control and morphology has also been studied in S. pombeusing cytochalasins (Kobori et al., 1989). Treatment of regenerating protoplasts with a high concentration of cytochalasin D (CD) causes actin to disappear as well as the complete inhibition of cell wall formation of S. pombe. A low concentration of CD results in weakly stained unlocalized actin, which induces grossly aberrant cell wall deposition as well as substantial changes in the morphology of the regenerating protoplasts. Electron microscopic observation of the CD-treated regenerating protoplasts revealed that actin is associated with the initiation of cell wall formation and formation of a fibrillar network of β(1→3) glucan on the surface of the regenerating protoplasts of S. pombe(Kobori et al., 1989; Osumi et al., 1989). These findings strongly support the concept that actin controls cell wall formation, growth polarity and morphology of yeasts. We speculate that nPCKs of S. pombecontrol actin organization as the dynamic changes in actin organization were arrested in both the Δpck2and staurosporine-treated wild-type cells. The nPKCs in S. pombewould thus be involved in the proper organization of the actin cytoskeleton, which could contribute to transport of cell wall components to their appropriate destination and to maintenance of normal cell shape and growth polarity.

We thank Dr H. Miwatani, Tokyo Medical and Dental University, for his helpful advice.