The characteristics of chromatin architecture are important for gene expression and chromosome stability, and they are regulated by two types of protein complex, namely the histone modification complexes and the chromatin remodelling complexes. The histone modification complexes covalently modify histones (reviewed in Roth et al., 2001), while the chromatin remodelling complexes use ATP as energy to remodel the structure of chromatin. The ATP-based chromatin remodelling complexes are grouped into several families, viz. the ISW1, CHD, SWI/SNF and INO80 complexes (reviewed in Gangaraju and Bartholomew, 2007). Except for the presence of a conserved ATPase core subunit, each complex contains various different additional components; these additional components are important when the complex interacts with DNA or the nucleosome. In Saccharomyces cerevisiae, the SWI/SNF family consists of two homologous complexes; these are the SWI/SNF (the switching of mating type/sucrose non-fermenting) complex and the RSC (remodel the structure of chromatin) complex. The RSC complex was identified during a homologous search of the yeast genome using the SWI/SNF complex (Cairns et al., 1996). Many subunits of the RSC complex are homologous to their related SWI/SNF subunits. However, the functions of the SWI/SNF and RSC complexes are not the same. In yeast cells, the RSC complex is more abundant than the SWI/SNF complex, and the vast majority of subunits making up the RSC complex are essential for cell viability.
The RSC complex contains about 17 subunits, of which STH1 is the core subunit and contains the ATPase domain. The RSC complex can be assembled into a variety of different subcomplexes (Cairns et al., 1996, 1999). These include the Rsc1 and Rsc2 subcomplexes, which respectively contain or do not contain the Rsc3/30 dimer. Mutants of the RSC complex exhibit multiple cellular defects, including ones in transcriptional regulation (Govind et al., 2005; Inai et al., 2007; Moreira and Holmberg, 1999), sister chromatids cohesion (Baetz et al., 2004; Hsu et al., 2003), DNA breakage repair (Chai et al., 2005; Shim et al., 2005), ploidy maintenance (Angus-Hill et al., 2001; Campsteijn et al., 2007; Lanzuolo et al., 2001) and cell wall integrity (Angus-Hill et al., 2001; Chai et al., 2002; Wilson et al., 2006). On exploring the regulation of cell wall integrity, several rsc mutants were found to be hypersensitive to the agents that cause cell wall stress. The cell wall defect of rsc3 is suppressed by over-expression of PKC1 and its upstream components (Angus-Hill et al., 2001); the PKC1 gene is important for the cell wall biogenesis and stress response (reviewed in Levin, 2005). In addition, a sth1 mutant displays synthetic weakness with respect to a ∆slg1 mutant, a PKC1 upstream component (Chai et al., 2002). During suppression of an rsc mutant, the suppression does not work with respect to downstream factors of the PKC1 pathway, which suggests that the RSC complex is functionally coordinated with an alternative PKC1 pathway. Moreover, transcriptional expression of several cell wall proteins is changed in rsc mutants (Angus-Hill et al., 2001; Conde et al., 2007).
The HTL1 (high temperature lethal) gene encodes a non-essential subunit of the RSC complexes. Physical and genetic interactions between HTL1 and the RSC complexes have been demonstrated (Florio et al., 2007; Lu et al., 2003; Romeo et al., 2002; Wilson et al., 2006). These reports suggested that HTL1 is associated with a subset of RSC complexes and mediates the assembly of Rsc8p into the RSC complexes; however, the relationship between structural integrity and functionality of the RSC complex, when it is assembled by HTL1, remains unclear. Yeast cells that are ∆htl1 mutants show lethality at the non-permissive temperature (37°C). Furthermore, several defects similar to those in other rsc mutants are found in these ∆htl1 cells; these include ploidy increase and mitotic arrest at the non-permissive temperature (Lanzuolo et al., 2001; Romeo et al., 2002). The cellular function of HTL1 also has been linked to cell wall integrity. Multiple-copy plasmid expressing HTL1 suppresses the growth defect present in a conditional PKC1 mutation; moreover, an additive growth defect is present in double mutants containing knockouts of HTL1 and BCK1, the downstream factor of PKC1 (Romeo et al., 2002). Nevertheless, a defect in cell wall integrity has not been reported in ∆htl1 single mutants. HTL1 has been reported to be important in three individual genome-wide studies, including one on ionizing radiation sensitivity (Bennett et al., 2001), one on abnormal telemetric length (Askree et al., 2004) and one on small nucleoli RNA processing (Qiu et al., 2008). Altogether, it would seem that HTL1 is not only functionally associated with the RSC complexes and the rearrangement of nucleosome, but may also be involved in a much wider range of important cellular functions.
In this study, we investigated the cellular functions of HTL1. We report that phenotypes of the ∆htl1 mutant at high temperature are based on a functional defect rather than a structural defect of the RSC complex. In addition, HTL1 may play a role different to other RSC components in regard to cell wall integrity and the G2–M transition. Furthermore, we show that the defects in cell wall integrity and the G2–M transition of the ∆htl1 mutant would seem to be interconnected.
Materials and methods
Yeast strains and growth
The S. cerevisiae strains used in this study are listed in Table 1 (see also Supporting information, Table S1). RSC components under the control of the GAL1 promoter (GAL1–HTL1, GAL1–STH1 and GAL1–RSC8) were constructed in the BY4741 background. The endogenous promoters of the individual gene were replaced by a GAL1 promoter using a one-step, PCR-based strategy (Longtine et al., 1998). The depletion of these genes in glucose medium was examined by immunoblotting using individual antibodies (see Supporting information, Figure S1). The deletion of MAD1, BUB2 and SLT2 was carried out by one-step PCR-based recombination; these genes were introduced into BY4741 or other yeast strains as indicated. HTL1-degron was generated as previously described and introduced into the YKL200 background (Kanemaki et al., 2003). We designed a PCR product that did not contain the myc epitope. For the SCC1–3HA strains, the gene was tagged with three copies of haemagglutinin (HA) in BY4741 by one-step PCR-based recombination (Knop et al., 1999). The tagging was verified by PCR and immunodetection with anti-HA antibody (Covence, MMS 101R 16B12). Several ∆htl1 clones were used in this study. BY5751 was generated from wild-type BY4741 (Winzeler et al., 1999), which was purchased from Open Biosystems. Deletion of HTL1 from the other strains in this study was carried out as previously reported (Lu et al., 2003). Deletion of RSC7 was carried out using the same strategy as deletion of HTL1 in BY5751. The TAP-tagged strains used in this study were also purchased from the Open Biosystems (Ghaemmaghami et al., 2003).
Yeast strains were grown in YPAD (1% yeast extract, 2% peptone, 2% glucose and 20 mg/ml adenine) or YPAGal (1% yeast extract, 2% peptone, 2% galactose and 20 mg/ml adenine). The growth temperature was routinely 30°C, except as indicated.
The spot assay in this study was carried out as follows. Cells were incubated with YPAD or YPAGal at 25°C or 30°C to saturation and then refreshed to mid-log phase. For the strains with the HTL1-degron background, the cells were incubated with YPARaffinose containing copper ions (100 mm CuSO4) to saturation and refreshed to mid-log phase. Cultures were 10-fold serially diluted and then dropped onto the indicated plates. The plates were then incubated at the indicated temperature for 3 days.
In order to examine DNA content, the cells were refreshed as described and then flow-cytometry analysis was carried out as follows. Briefly, cells were collected and resuspended in the FACS solution (40% EtOH, 0.1 m sorbitol, 5 mm EDTA, pH 8.0, and 5 mm NaN3) to fix them and then stored at −80°C until use. Before analysis, the cells were resuspended in PBS containing Triton X-100 (0.5%) and mixed well by vortexing. The cells were then collected by centrifugation and resuspended in 500 µl Tris–HCl buffer (50 mm, pH 8.0) containing 100 µg/ml RNase A. Next, the cells were incubated at 37°C for > 1 h and then 300 µl sodium citrate (38 mm) containing 25 µm Systox Green (Molecular Probes, S7020) was added to the sample with mixing. After sonication, 400 µl sample was transferred to a tube containing 2.1 ml phosphate-buffered saline (PBS) and then the sample was ready for the assay.
Protein extraction and immunoblotting
Total proteins were extracted with lysis buffer (50 mm Tris–HCl, pH 8.0, 150 mm NaCl, 1 mm dithiothreitol (DTT), 1 mm phenylmethylsulphonyl fluoride (PMSF), 1 mm NaVO3, 5 mm NaP2O7, 1 mm NaF, 0.1% Triton X-100 and protease inhibitor cocktail tablets (Roche, 1 873 580), using glass beads (Sigma, G-8772) at 4°C. The lysate supernatant was collected and mixed with an equal volume of 2× SDS sample buffer (125 mm Tris–HCl, pH 6.8, 4% SDS, 10% glycerol, 0.06% bromophenol blue and 200 mm DTT). The sample was then immediately boiled at 100°C. SDS–gel electrophoresis was performed following the usual procedure. To separate Scc1–3HA, 6.5% SDS–PAGE was used. For other proteins, 7.5–9% SDS–PAGE analysis was used. To detect Htl1p, 12% Tris–tricine gel electrophoresis was used (Schagger and von Jagow, 1987). Expression of Slt2p was examined using anti-Slt2p antibody (Santa Cruz, Sc-6803).
Yeast strains were incubated with 50 ml YPAD to log phase and collected. Total proteins were extracted with 400 µl lysis buffer, with or without 1 m sorbitol. Then 100 µl supernatant was transferred to a 1.5 ml tube containing 300 µl lysis buffer. Anti-Sth1 or anti-Rsc8 antiserum was added to the lysate and the mixture incubated on a nutator (Becton Dickinson, Cat. No. 421105) at 4°C for 3 h. After incubation, the mixture was transferred to a new 1.5 ml tube containing 50 µl protein A beads (Amersham Biosciences, 17-5280-01). The mixture was incubated again for 1.5 h. The beads were washed three times with 1 ml lysis buffer. Finally, the beads were resuspended in SDS sample buffer and boiled.
The phenotypes of the ∆htl1 mutant are rescued by 1 m sorbitol
A previous report showed that multiple-copy HTL1 rescued the growth defect of the pkc1 mutant, and that there was additional growth defects in the ∆htl1 and pkc1 double mutant (Romeo et al., 2002); however, the characteristics of the cell wall defect have not been explored in ∆htl1 single mutants. In the present study, we asked whether one or more cell wall defects were present in ∆htl1 cells. In terms of cell wall defect, ∆htl1 cells in the BY4741 background were hypersensitive to agents that cause cell stress, such as calcofluor white (CFW), caffeine and 1 m NaCl (Figure 1A) at 30°C. This study shows that the CFW sensitivity of ∆htl1 yeast cells can be rescued by an episomally expressed HTL1 gene (Figure 1A).
We examined several phenotypes of ∆htl1 yeast cells in the presence of 1 m sorbitol, since the cell wall defect of the sth1 and rsc3 mutants has been shown to be suppressed by 1 m sorbitol (Angus-Hill et al., 2001; Chai et al., 2002). 1 m sorbitol is approximately equal to the internal osmolarity of yeast cell growing in YPD (Arnold and Lacy, 1977). The presence of 1 m sorbitol in the medium stabilizes the yeast cells against osmotic stress (Levin and Bartlett-Heubusch, 1992). As shown in Figure 1B, C, the temperature sensitive and CFW-sensitive phenotypes of ∆htl1 were completely or partially rescued in the presence of 1 m sorbitol. The sorbitol rescued the temperature sensitivity of the ∆htl1 mutant at 34°C but not at 37°C (data not shown). ∆htl1-1 and ∆htl1-2 were derived from BY4741 and SEY6211, respectively. The difference in temperature sensitivity of the two ∆htl1 strains shown in Figure 1B may be due to distinct genetic backgrounds.
We further examined whether the ploidy increase of ∆htl1 was also suppressed by 1 m sorbitol (Lanzuolo et al., 2001). The ploidy was increased in the GAL1–HTL1 strain when Htl1p was conditionally depleted in YPAD medium at 30°C, but the ploidy increase of the GAL1–HTL1 was suppressed in the presence of 1 m sorbitol (Figure 1D). Similar results were also observed in the HTL1-degron strain when the cells were incubated in YPAGal medium at 32°C, the maximal permissive temperature for the HTL1-degron, in the presence of 1 m sorbitol (Figure 1D).
Sorbitol does not restore the disassociation of Rsc8p from RSC complexes in the ∆htl1 mutant
HTL1 is considered a component of the RSC complex and may assemble into both the Rsc1 and Rsc2 sub-complexes (see also Supporting information, Figure S2). Furthermore, a previous report showed that Htl1p is not only associated with Rsc8p but also mediated the assembly of Rsc8p into the RSC complex (Florio et al., 2007). In the present study, we have found that the CFW sensitivity and ploidy increase of the ∆htl1 mutant are temperature-dependent, and the phenotypes of ∆htl1 cells are rescued by isosmotic concentration of sorbitol (Figures 1, 2). Therefore, we asked whether the structural defect of the RSC complex in ∆htl1 cells was consistent with the extent of the phenotypes. In addition, we also asked whether sorbitol stabilized the structural defect of the RSC complex that does not contain Htl1p (Hohmann, 2002). To answer these questions, we examined the association between Rsc8p and RSC complexes in ∆htl1 cells by co-immunoprecipitation (Co-IP) analysis. As shown in Figure 3A, the amount of co-immunoprecipiated Rsc8p was reduced in ∆htl1 cells when the Co-IP analysis was carried out with anti-Sth1 antiserum; however, this reduction was not increased when the incubation temperature was elevated (Figure 3A). Moreover, the disassociation of Rsc8p from RSC complexes in ∆htl1 cells was not rescued when cells were incubated in YPAD containing 1 m sorbitol at 30°C (Figure 3B). Taken together, these results suggest that HTL1 plays a role in assisting the assembly of Rsc8p into RSC complexes, and that the disassembly of Rsc8p from RSC complexes in the ∆htl1 mutant is not consistent with the increasing severity of the phenotypes as the temperature increases. In addition, complementation of the phenotypes of ∆htl1 cells by an isosmotic concentration of sorbitol is not due to an effect involving the stabilization of the RSC complex lacking Htl1p.
The accumulation of phosphorylated Scc1p in ∆htl1 cells is relieved by deletion of MAD1
Previous reports have demonstrated that G2–M arrests of the sth1 and sfh1 mutants are the consequence of activation of the spindle checkpoint (Hsu et al., 2003; Tsuchiya et al., 1998). Deletion of MAD1, a component of spindle checkpoint, releases mutant cells from G2–M phase. Similarly, G2–M arrest in ∆htl1 yeast cells at the non-permissive temperature has also been described (Romeo et al., 2002). In the present study, we asked whether the accumulation at G2–M phase in ∆htl1 yeast cells at the non-permissive temperature was also MAD1-dependent. As shown in Figure S3 (see Supporting information), deletion of MAD1 did relieve the accumulation of 2n DNA content when Sth1p or Rsc8p were depleted. When the Htl1p was depleted at 37°C in a ∆mad1 background, the 2n DNA peak remained prominent; however, an additional 4n DNA peak appeared (see Supporting information, Figure S3; GAL1–HTL1, HTL1-degron). This 4n DNA peak may have been caused by the release of G2–M arrest when the Htl1p was depleted at restrictive temperature in the ∆mad1 background.
Next we examined the expression and the phosphorylation status of cohesion protein Scc1p. SCC1 was tagged with three copies of the HA epitope (3HA) in GAL1–STH1, GAL1–RSC8 and GAL1–HTL1 backgrounds. In yeast cells, sister chromatids are connected by the cohesion complex. Expression of the cohesion protein Scc1p is cell cycle-regulated and peaks during S phase. Scc1p is hyperphosphorylated during G2–M phase, and phosphorylation of Scc1p is required for its cleavage by Esp1p (Alexandru et al., 2001; Uhlmann et al., 2000). In this experiment, cells were incubated with YPAD to deplete the RSC subunits and expression of Scc1–3HA was examined. The results showed that the migration of Scc1–3HA during electrophoresis was much slower in cells where Sth1p or Rsc8p had been depleted (Figure 4A, B). As a phosphorylation control for Scc1–3HA, the wild-type SCC1–3HA cells were treated with the microtubule-disruptor nocodazole and this showed a similar electrophoretic pattern. Next, MAD1 was deleted from SCC1–3HA/GAL1–STH1 cells and phosphorylation of Scc1–3HA was examined. The result showed that the accumulation of phosphorylated Scc1–3HA on the depletion of Sth1p was reduced in the ∆mad1 background (Figure 4B).
When the SCC1–3HA/GAL1–HTL1 cells were examined, accumulation of phosphorylated Scc1–3HA occurred on the depletion of Htl1p at 30°C and 37°C (Figure 4A, B), and it was greater at 37°C. In addition, this accumulation was reduced by the deletion of MAD1 (Figure 4B). The expression of Scc1–3HA was also examined in the ∆htl1 background, and the result was similar to that of SCC1–3HA/GAL1–HTL1 cells. Therefore, these results suggest that the mitotic arrest of the ∆htl1 mutant is also MAD1-dependent.
Expression of MAP kinase SLT2 in the ∆htl1 mutant is upregulated
A previous report has shown that overexpression of PKC1 rescued the organization of the cellular cytoskeleton of a sth1 mutant and the cell wall defect of a rsc3 mutant (Angus-Hill et al., 2001; Chai et al., 2002). In the present study, we asked whether the cell wall defect of the ∆htl1 mutant might also be rescued by overexpression of PKC1 gene, since the ∆htl1 mutant is synthetically lethal when combined with the pkc1 mutant (Romeo et al., 2002). The results showed that the sensitivity of ∆htl1 cells to temperature and CFW was not rescued by a dominantly active PKC1 (Helliwell et al., 1998) (see Supporting information, Figure S4).
Previous reports have suggested that the RSC complex is coordinated by an alternative PKC1 pathway, since overexpression of factors downstream of PKC1 do not rescue the growth and cell wall defects in the sth1 mutant (Chai et al., 2002). Overexpression of PKC1 did not rescue cell wall defect of the ∆htl1 mutant, which suggests that HTL1 may regulate these functions by a mechanism different to that of other RSC components. Next, we asked whether the activation of MAP kinase pathway was able to influence the ∆htl1 mutant. The MAP kinase pathway is a signalling cascade downstream of PKC1 and other cell wall integrity (CWI) pathways (Hohmann, 2002; Levin, 2005). We examined the expression of Slt2p/Mpk1p. In cell wall mutants or in the presence of cell wall stress, expression of Slt2p is upregulated (Lagorce et al., 2003). The results showed that expression of Slt2p was increased in the wild-type strain when cells were treated with CFW or a 39°C heat shock (Figure 5). A similar response was also demonstrated in the ∆htl1 mutant. The results also showed that the expression level of Slt2p in ∆htl1 cells was higher than that of wild-type cells, no matter whether or not CFW or heat shock was present. Therefore, the MAP kinase pathway is functional and overactivated in the ∆htl1 mutant. We also examined the expression level of Slt2p in a ∆rsc2 mutant and also when Sth1p was depleted as a comparison. The results showed thatin both the ∆rsc2 mutant and Sth1p-depleted cells, the expression level of Slt2p was also higher than that of wild-type cells in the absence of CFW and heat shock; however, the increase of Slt2p that occurred on treating cells with CFW or shifting to 39°C was limited (Figure 5). RSC7 encodes another non-essential subunit of the RSC complex. Since the ∆rsc7 mutant and the ∆htl1 mutant share similar phenotypes (Wilson et al., 2006) (data not shown), the expression of Slt2p was also examined in the ∆rsc7 mutant. Similar to the situation in ∆htl1 cells, the expression level of Slt2p in ∆rsc7 cells was higher than that of wild-type cells under all conditions examined (Figure 5). Taking these results together, it would seem that, in the absence of CFW and heat shock, the CWI pathway is activated in the rsc mutants examined in this study, and the response to the cell wall stresses in the various rsc mutants seems to be different.
Mitotic arrest of the ∆htl1 mutant is rescued by 1 m sorbitol and deletion of SLT2
Finally, we asked whether the defect in cell wall integrity of the ∆htl1 mutant also affected G2–M transition in ∆htl1 mutant. We examined the expression of Scc1p in the ∆htl1 mutant in the presence of 1 m sorbitol. The wild-type and ∆htl1 cells in the SCC1–3HA background were incubated at 30°C to log phase in YAPD alone or in YPAD containing 1 m sorbitol. Total proteins were extracted and expression of Scc1–3HA was examined. The results showed that the accumulation of phosphorylated Scc1–3HA in ∆htl1 cells was moderated in the presence of 1 m sorbitol (Figure 6A).
The above results show that the expression level of Slt2p was upregulated in the ∆htl1 mutant. Therefore, we examined whether the accumulation of phosphorylated Scc1p in ∆htl1 cells could be suppressed by deletion of SLT2, since an interconnection between cell cycle progression and cell wall integrity in yeast cells has been reported (Clotet et al., 2006; Yaakov et al., 2009). The results showed that the accumulation of phosphorylated Scc1–3HA was partially rescued in the ∆slt2 background when Htl1p was depleted (Figure 6B); furthermore, the accumulation of 2n DNA content on the depletion of Htl1p was also suppressed by ∆slt2 when the cells were examined by flow cytometry. These results suggest that the same signalling pathway mediates both cell wall integrity and G2–M transition in ∆htl1 mutants.
Although the HTL1 gene was reported to have genetic interaction with genes encoding the PKC1 pathway, the cell wall defect has not been reported in ∆htl1 single mutants (Romeo et al., 2002). In the present study, we have shown that ∆htl1 cells display characteristic defects in cell wall integrity, e.g. hypersensitivity to CFW, caffeine and 1 m NaCl. We have also found that the cell size of the ∆htl1 mutant is enlarged (data not shown). Moreover, the CFW sensitivity and enlarged cell size in the ∆htl1 mutant are rescued by 1 m sorbitol (Figure 1 and data not shown), an osmotic stabilizer. This suggests that the HTL1 gene is involved in cell wall integrity in addition to the G2–M transition.
This study found that the assembly of Rsc8p into RSC complexes was reduced in the ∆htl1 mutant, indicating that Htl1p is required for the assembly of Rsc8p into RSC complexes. This is consistent with a previous report (Florio et al., 2007). Using Co-IP analysis, it was found that the disassociation of Rsc8p from RSC complexes in the ∆htl1 mutant was not increased when the growth temperature was elevated, which indicates that defects in the ∆htl1 mutant at the elevated temperature are due to the deficient functioning of RSC complexes, but not to an aberrant RSC complex structure. However, we still are unable to exclude the possibility that other components are unstable or may disassemble from the RSC complexes in the ∆htl1 mutant at the non-permissive temperature. We have also demonstrated that the disassembly of Rsc8p from RSC complexes in the ∆htl1 mutant is not restored by incubating cells with 1 m sorbitol, suggesting that the rescue of ∆htl1 cells by 1 m sorbitol is not due to a restoration of RSC integrity. This also indicates that the CWI defect is critical to the phenotypes of the ∆htl1 mutant. The association between the Rsc8p and RSC complexes was also examined in the ∆rsc7 mutant in this study. The results showed that the assembly of Rsc8p into the RSC complex was also reduced in the ∆rsc7 mutant, and this disassembly of Rsc8p was not increased or restored when ∆rsc7 cells were incubated at non-permissive temperature or in the presence of 1 m sorbitol (data not shown). These suggest that RSC7 not only regulates a number of cellular functions but also mediates RSC integrity in a way similar to HTL1. However, a previous report has shown that Rsc8p is unstable but is not disassembled from the RSC complex in the ∆rsc7 mutant (Wilson et al., 2006); this difference in results may be a function of the different methodologies used in the two studies.
Our results showed that the expression level of Slt2p in the ∆htl1 mutant was significantly increased under all conditions examined, which indicates that the function of PKC1 pathway is not lost in the ∆htl1 mutant. This result is also important in terms of the analysis of the ∆htl1 mutant, since the cell wall defect and the temperature sensitivity of the ∆htl1 mutant was not rescued by overexpression of PKC1. The upregulation of MAP kinase suggests that the CWI pathway is activated to compensate for the cell wall defect of the ∆htl1 mutant. However, the signalling network needs further investigation in order to understand the mechanism of MAP kinase upregulation in ∆htl1 cells. The HOG1 kinase pathway mediated by osmotic stress is one possibility, since Slt2p has recently been shown to cooperate with the HOG1 pathway in response to cell wall stress (Bermejo et al., 2008). Additionally, the response to the cell wall mutants may contribute to the upregulation of Slt2p in ∆htl1 cells (Lagorce et al., 2003).
The expression level of Slt2p was increased in the ∆rsc2 mutant and the Sth1p-depleted condition at 30°C. However, the increase in Slt2p in response to the CFW or heat shock was limited in both situations. In contrast, in the ∆htl1 mutant, the increase in Slt2p in response to the cell wall stresses was similar to that of the wild-type cell. This result probably reflects the intrinsic difference between the ∆htl1 and other rsc mutants in terms of the growth and cell wall defects. The inability to increase the expression of Slt2p by CFW or heat shock stimulation indicates that the regulation of cell wall stress or heat shock stress is affected in the ∆rsc2 and Sth1p-depleted cells. The expression of Slt2p can be stimulated in the presence of CFW or heat shock, indicating that this cell wall integrity pathway in the ∆htl1 mutant remains intact. Thus, the cell wall defect in the ∆htl1 mutant may due to the defect in different signalling pathway.
Our results show that defects in cell wall integrity and the G2–M transition in the ∆htl1 mutant may be interconnected. This suggestion is supported by two observations: first, that both the CFW and temperature sensitivity of the ∆htl1 mutant are partially rescued by 1 m sorbitol; second, that the mitotic arrest of Htl1p-depleted cells is rescued by ∆slt2. However, we still can not exclude the possibility that sorbitol may rescue these two phenotypes of the ∆htl1 mutant independently; in addition, SLT2 may have a role other than one in the CWI signalling pathway.
We are grateful to T. F. Wang at the Institute of Molecular Biology, Academia Sinica, for supplying the plasmids containing the KanMX6 and His3MX6 modules as the template for PCR amplification in our construction of the yeast promoter replacement strains and the HA-tagged strains. We thank Mei-Yu Chen at the Institute of Biochemistry and Molecular Biology, National Yang-Ming University, for a kind gift of the anti-Slt2 antibody and plasmids containing PKC1 gene. We also thank EUROSCARF for providing yeast strains and plasmids for the construction of the yeast degron-tagged strains. This work was supported by Grant No. NSC96-2311-B-010-007 from the National Science Council, Taiwan, and a grant (‘Aim for the Top University Plan’) from the Ministry of Education, Taiwan.