Rsp5 is an essential and multi-functional E3 ubiquitin ligase in Saccharomyces cerevisiae. The Ala401Glu rsp5 mutant, which is hypersensitive to various stresses, was isolated previously. To understand the function of Rsp5 in stress responses, suppressor genes whose overexpression allows rsp5A401E cells to grow at a high temperature were screened. The KIN28 and POG1 genes, encoding a subunit of the transcription factor TFIIH and a putative transcriptional activator, respectively, were identified as multicopy suppressors of not only high temperature but also LiCl stresses. The overexpression of Kin28 and Pog1 in rsp5A401E cells led to an increase in the transcriptional level of some stress proteins when exposed to a temperature up-shift. Based on DNA microarray analysis under LiCl stress, it appears that the transcriptional level of some proteasome components is slightly increased in rsp5A401E cells overexpressing Kin28 or Pog1. These results suggest that the overexpression of Kin28 and Pog1 enhances the protein refolding and degradation pathways in rsp5A401E cells.
Stresses induce protein denaturation, generate abnormal proteins, and lead to growth inhibition or cell death. The stress-tolerant mechanism in yeast Saccharomyces cerevisiae was analyzed (Hoshikawa et al., 2003; Morita et al., 2003; Nomura & Takagi, 2004). An S. cerevisiae mutant was isolated previously that shows hypersensitivity to various stresses, such as toxic amino acid analogues, high growth temperature in a rich medium, ethanol, LiCl, and H2O2. This mutant carried a single amino acid change of Ala401 by Glu in the allele of RSP5 encoding an essential E3 ubiquitin ligase (Hoshikawa et al., 2003). Rsp5 is an essential E3 ubiquitin ligase containing the HECT domain (homologous to E6-AP carboxyl terminus) involved in ubiquitin-mediated protein degradation. It has been shown that Rsp5 participates in many biological processes, such as the endocytosis of plasma membrane permeases and receptors (Springael & André, 1998; Dunn & Hicke, 2001), degradation of the large subunit of the RNA polymerase II (Huibregtse et al., 1997), mitochondrial inheritance (Fisk & Yaffe, 1999), biosynthesis of unsaturated fatty acids (Hoppe et al., 2000), the endoplasmic reticulum-associated degradation (Haynes et al., 2002), and proper nuclear export of mRNA (Rodriguez et al., 2003) through ubiquitination of the target proteins.
In S. cerevisiae, two major transcription factors, Hsf1 and Msn2/4, appear to be responsible for the stress-induced gene expression (Hashikawa & Sakurai, 2004; Ferguson et al., 2005). Hsf1 binds to heat shock element (HSEs) and Msn2/4 binds to stress response elements (STREs) found in the promoters of many genes encoding stress proteins that can refold stress-induced abnormal proteins, such as HSP12, HSP42, and DDR2. Recently, HSE-mediated gene expression was defective in the rsp5-101 mutant under high-temperature conditions (Kaida et al., 2003). It was also found that the transcription of stress protein genes in rsp5A401E cells was significantly lower than that in the wild-type strain when exposed to a temperature up-shift, ethanol, or sorbitol (Haitani et al., 2006). The protein levels of Hsf1 and Msn2/4 were also remarkably defective in rsp5A401E cells. Interestingly, the mRNAs of HSF1 and MSN2/4 were accumulated in the nucleus of rsp5A401E cells after exposure to stresses, and even under nonstress conditions (Y. Haitani and H. Takagi, unpublished results). Based on the fact that Rsp5 is required for proper nuclear export of RNA molecules (Rodriguez et al., 2003), it is believed that the decreased amounts of synthesized Hsf1 and Msn2/4 cause the reduced induction of stress protein genes in rsp5A401E cells. These results suggest that Rsp5 mediates the expression of stress proteins, primarily by regulating Hsf1 and Msn2/4.
For a better understanding of the mechanism of the Rsp5-involved stress response, a genome-wide screening was carried out for multicopy suppressor genes, which may function as positive regulators of the downstream or parallel pathways. Here, it is reported that the overexpression of two transcriptional factors, Kin28 and Pog1, subunit of the transcription factor TFIIH and a putative transcriptional activator, respectively, could rescue the sensitivities of rsp5A401E cells to high growth temperature and LiCl stresses. Transcriptional analyses revealed that the level of genes encoding some stress proteins or proteasome components was increased in rsp5A401E cells overexpressing Kin28 or Pog1.
Materials and methods
Strains, plasmids, and media
The S. cerevisiae strains used in this study were the wild-type CKY8 (MATαura3-52 leu2-3,112) and the Ala401Glu rsp5 mutant CHT81 (MATαura3-52 leu2-3,112 rsp5A401E) (Hoshikawa et al., 2003). Escherichia coli strain DH5α [F−λ−Φ80lacZΔM15Δ(lacZYA argF)]U169 deoR recA1 endA1 hsdR17(rk−mk+) supE44 thi-1 gyrA96] was used to subclone the yeast gene and construct plasmids.
Yeast episomal plasmids pAD4 and pUV2 containing LEU2 and URA3, respectively, were used for expressing the yeast genes under control of the ADH1 promoter. Plasmid pAD-RSP5 (Haitani et al., 2006) was used to express the RSP5 gene. Plasmid pHS5 (Ogawa et al., 1995) derived from YCp50 harboring CEN4 and URA3, which contains the coding region of lacZ, was used for reporter gene assay. Plasmid YEp24 (Carlson & Botstein, 1982) harboring URA3 and two plasmids pUC18 and pKF3 (Takara Bio, Ohtsu, Japan) harboring the bacterial ampicillin- and chloramphenicol-resistant gene, respectively, were used to subclone the yeast gene.
The media used for S. cerevisiae were YPD (2% glucose, 1% yeast extract, 2% peptone) and SC (Rose et al., 1990) consisting of 2% glucose, 0.67% bacto yeast nitrogen base with ammonium sulfate (Difco Laboratories, Detroit, MI), and drop-out mix lacking leucine (SC-Leu), lacking uracil (SC-Ura), or lacking leucine and uracil (SC-Leu-Ura). The E. coli strains were grown in Luria–Bertani medium (Sambrook & Russell, 2001).
Isolation and subcloning of multicopy suppressors
A 2 µm- and LEU2-based genomic library containing more than 30 000 independent E. coli clones (Nikawa et al., 1987) was introduced into CHT81, and transformants were selected on SC-Leu at 37 °C to identify plasmids, allowing the growth of rsp5A401E cells at a high temperature in a rich medium. Plasmids prepared from the colonies were then shuttled into E. coli DH5α and back into CHT81 to retest the growth on SC-Leu at 37 °C. The plasmids were sequenced to define the ends of the insert DNA using primers 5′-CCA ATG TGA GAT TTT GGG C-3′ and 5′-GTT TTC CCA GTC ACG AC-3′. The sequence obtained was compared with the yeast genome using the blast program (http://seq.yeastgenome.org/cgi-bin/nph-blast2sgd).
Candidate genes were subcloned into pAD4 and retested for the ability to suppress the stress sensitivity. The DNA fragments of KIN28 and POG1 were amplified by PCR performed with genomic DNA of CKY8 and oligonucleotide primers 5′-CCA AAA GAG CTC GTC AAT AAC ACA GAT TCT AC-3′ and 5′-AAA ACT GCA GGA GCG AAC ATA TGA AAG TGA-3′ (the underlined sequences indicate the positions of the SacI and PstI restriction sites, respectively), and 5′-ACC CAA GCT TGC CAT ATT ATG AAG GAG GAG-3′, and 5′-ACC CCT GCA GAC TGC ACG TCA TGT GGG A-3′ (the underlined sequences indicate the positions of the HindIII and PstI restriction sites, respectively). The unique amplified bands of 1220 and 1060 bp corresponding to KIN28 and POG1, respectively, were digested with SacI and PstI, and HindIII and PstI, and then ligated into the SacI–PstI sites and the HindIII–PstI sites of pAD4 to construct pAD-KIN28 and pAD-POG1, respectively.
Disruption of the POG1 gene
The DNA fragment of POG1 was prepared by PCR with genomic DNA of CKY8 and oligonucleotide primers. The forward primer was 5′-CGG GGT ACC TCG AAC ACA TCT CAC CCC CT-3′, and the reverse primer was 5′-CCC CAT TAT CTA GAG ATT GC-3′. The underlined sequences indicate the positions of the KpnI and XbaI restriction sites, respectively. A unique amplified band of 2.5 kb corresponding to POG1 was digested with KpnI and XbaI, and then ligated into the KpnI–XbaI sites of pUC18 to construct pUC-POG1. The 1.2-kb HindIII fragment containing URA3 of plasmid YEp24 was ligated to the HindIII site of pKF3 to construct pKF3-U. Plasmid pUCDpog1-U was then constructed by deleting the 1.1-kb EcoRV–NaeI fragment in POG1 from pUC-POG1, and inserting the 1.2-kb FspI–SmaI fragment containing URA3 of plasmid pKF3-U by blunt-end ligation. The 2.6-kb KpnI–XbaI fragment containing pog1::URA3 of pUCDpog1-U was integrated into the POG1 locus in CKY8, to construct strains CHTDpog1 by transformation. The Ura+ phenotype was selected, and the correct disruption was verified by chromosomal PCR analysis.
Reporter gene assay
The DNA fragments of 500 bp containing the 5′-upstream region of HSP12, HSP42, and DDR2 were PCR-amplified using genomic DNA of CKY8 and oligonucleotide primers 5′-GGC CCA AGC TTG GAC TAG AAG CCA AAA GCC AG-3′, 5′-TTG GGG GAT CCA TTG TTG TAT TTA GTT TTT TTT GTT TTG-3′, 5′-GGC CC A AGC TTC TGG GGT TGG GTA ACA AGT GA-3′, 5′-TTG GGG GAT CCA TTG CTT CGG CTT GGT ATG AT-3′, 5′-CCG CTC AAG CTT TTA TTT CCT CTG ATG TAA-3′, and 5′-TTG GGG GAT CCA TGT TTA AAT CGA TAT TAA ATT AGC GTG-3′. The underlined sequences indicate the positions of the HindIII and BamHI restriction sites, respectively. The PCR products were digested by HindIII and BamHI and inserted into the HindIII–BamHI sites proximal to the coding region of lacZ in pHS5 to generate pHS-HSP12, pHS-HSP42, and pHS-DDR2, respectively; note that in the resultant fusion proteins, the first eight amino acids of β-galactosidase were replaced with the three amino acids Met–Asp–Pro. The nucleotide sequences of the cloned fragments were confirmed by DNA sequencing and each plasmid was used to transform CKY8 and CHT81. After incubation of the resultant transformants in SC-Ura or SC-Leu-Ura medium at 25 °C to the exponential growth phase, the temperature was up-shifted to 37 °C or 200 mM LiCl was added. At the intervals (0–120 min), the cells were harvested and lysed. Then, the β-galactosidase activity was measured by means of the hydrolysis of o-nitrophenyl-β-d-galactopyranoside to produce o-nitrophenol (ONP) and galactose. β-Galactosidase activity (in units) was calculated using the formula OD420 nm of ONP × 1000 min mL−1 OD600 nm−1 of assayed culture.
Northern blot analysis
Northern blot analysis was carried out using an AlkPhos Direct Labelling Reagents and Detection system (Amersham Biosciences, Piscataway, NJ). Total RNA from S. cerevisiae was isolated by the method of Köhrer & Domdey (1991). As a DNA probe, the DNA fragments of HSP12 and ACT1 were prepared by PCR with oligonucleotide primers 5′-CCA AAA CTG CAG CAA AAC AAA AAA AAC TAA ATA CAA CA-3′, 5′-AAC CCC GAG CTC TAC AAA GAG TTC CGA AAG AT-3′, 5′-CCA AAA CTG CAG GGT TGC TGC TTT GGT TAT TG-3′ and 5′-AAC CCC GAG CTC GAA ACA CTT GTG GTG AAC GA-3′. The underlined sequences indicate the positions of the PstI and SacI restriction sites, respectively. The unique amplified bands of 0.3 and 1.4 kb corresponding to HSP12 and ACT1, respectively, were digested with PstI and SacI, and then ligated into the PstI–SacI sites of pAD4. The nucleotide sequence of the cloned HSP12 and ACT1 was confirmed by DNA sequencing. The DNA fragments of HSP12 and ACT1 were digested with PstI and SacI, purified from agarose gel, denatured, and labeled according to the supplier's protocol.
Construction of expression plasmids
The DNA fragments of HSP12, HSP42, and DDR2 were amplified by PCR performed with genomic DNA of CKY8 and oligonucleotide primers 5′-CCA AAA CTG CAG CAA AAC AAA AAA AAC TAA ATA CAA CA-3′, 5′-AAC CCC GAG CTC TAC AAA GAG TTC CGA AAG AT-3′, 5′-CCA AAA CTG CAG GAT CAT ACC AAG CCG AAG CA-3′, 5′-AAC CCC GAG CTC TGT GTG TAT AAA CAG ATA CG-3′, 5′-CCA AAA CTG CAG CAC GCT AAT TTA ATA TCG ATT TAA AC-3′, and 5′-AAC CCC GAG CTC CAG TAA GCG GCG TTT TTC-3′, respectively. The underlined sequences indicate the positions of the PstI and SacI restriction sites, respectively. The unique amplified bands of 376 bp corresponding to HSP12, 1171 bp corresponding to HSP42, and 230 bp corresponding to DDR2 were digested with PstI and SacI, respectively, and then ligated into the PstI–SacI sites of pAD4 to construct pAD-HSP12, pAD-HSP42, or pAD-DDR2, respectively.
The DNA fragments of RPT2, PRE8, and PRE9 were amplified by PCR performed with genomic DNA of CKY8 and oligonucleotide primers 5′-AAA ACC CGG GGA TGC GTA TCA GTG TCA AGC-3′, 5′-AAA AGA GCT CTC GCG AAG GCC CCG ACG CCC-3′, 5′-TCC CCC CGG GGA GAG TGG AAT AGG TGA AGC-3′, 5′-AAG GGA GCT CGG TGA TTG GCG GGG ATA ATT-3′, 5′-AAA ACC CGG GAG CGA AGA GAA CAG ACT GCT-3′, and 5′-AAA AGA GCT CCT TCC CCC GAA AGG GGA ATC-3′. The underlined sequences indicate the positions of the XmaI and SacI restriction sites, respectively. The unique amplified bands of 1314, 868, and 777 bp corresponding to RPT2, PRE8, and PRE9, respectively, were digested with XmaI and SacI, and then ligated into the XmaI–SacI sites of pAD4 to construct pAD-RPT2, pAD-PRE8, or pAD-PRE9, respectively. The DNA fragment containing the ORF of PRE8 or PRE9 and the ADH1 promoter and terminator was amplified by PCR with plasmid DNA of pAD-PRE8 or pAD-PRE9, respectively, and oligonucleotide primers 5′-TCT TTC CTC TAG GCC TTC TAG CTC CCT AAC-3′ and 5′-GCT AGG CTG TAG GCC TGT GTG GAA GAA CGA-3′. The underlined sequences indicate the positions of the StuI restriction sites. The unique amplified bands of 2.4 and 2.3 kb corresponding to RPE8 and RPE9, respectively, were digested with StuI, and then ligated into the SmaI and preblunted XbaI sites of pUV2 to construct pUV-PRE8 and pUV-PRE9, respectively. The nucleotide sequences of the cloned genes were confirmed by DNA sequencing.
DNA microarray analysis
Poly A mRNA was enriched from total RNA by the Oligotex dT30 mRNA purification kit (Takara Bio). The Affimetrix yeast genome S98 arrays (YGS98 GeneChip, Affymetrix, Santa Clara, CA) were used as DNA microarray in this study. The biotinyated cRNA (15 µg) probe was hybridized to DNA microarray at 45 °C for 18 h according to Affymetrix user's manual. The washing and staining of arrays were performed using the GeneChip Fluidics Station 400. The scanning of arrays was carried out using the GeneArray scanner (Agilent technologies, Palto Alto, CA).
The fluorescence intensities of array images were quantified using the affymetrix software package (Microarray Suite version 4.0.1). The statistical analysis after data acquisition and normalization of expression data was carried out using the genespring 7.0 software (Silicon Genetics, Redwood City, CA). The genes flagged as either present or marginal were used for further analyses. Per chip global normalization (Sahara et al., 2002) was performed by the genespring 7.0 software. DNA microarray analyses were repeated twice using total RNA samples from two independent experiments. The genes with an expression ratio (60 min/0 min) >2.0- or <0.5-fold were identified as the induced genes or the repressed genes, respectively. To find the functions for stress tolerance, those induced genes in each strain were statistically evaluated with respect to the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways (http://www.genome.jp/kegg/). The probability that genes were involved in each metabolic pathway by chance alone (a P-value) was calculated. The metabolic pathways with a P-value <0.001 were considered to be an important function for stress tolerance.
Microarray data from the present study have been deposited in the Gene Expression Omnibus (GEO) repository at the National Center for Biotechnology Information (NCBI) under series accession number GSE8729.
The KIN28 and POG1 genes were isolated as multicopy suppressors of the rsp5A401E mutant at a high growth temperature
It should be noted that the temperature sensitivity of the rsp5A401E mutant was observed on a rich complex YPD medium, not on a synthetic minimal SD medium (Hoshikawa et al., 2003). It was also found that this mutant cannot grow at 37 °C even on a synthetic complete SC medium (data not shown). Therefore, the rsp5A401E mutant CHT81 was transformed with a yeast genomic library cloned into the 2 µm-based multicopy vector, and the transformants were screened for growth on SC-Leu solid medium at 37 °C for several days. Three transformants showed plasmid-dependent suppression of the growth defect caused by rsp5A401E, suggesting that the suppressor genes resided in the plasmids.
In addition to the expected wild-type RSP5 gene, the KIN28 and POG1 genes were identified as suppressors (Fig. 1a). An essential gene KIN28 encodes a subunit of the general transcription factor TFIIH, which has cyclin-dependent Ser/Thr protein kinase activity necessary for transcription of a vast majority of the genes in yeast (Cismowski et al., 1995). The POG1 encode a putative transcriptional activator that promotes growth in the presence of a mating pheromone (Leza & Elion, 1999).
Effect of multicopy suppressor genes on yeast growth phenotypes. (a) Wild-type cells (CKY8) harboring empty vector (pAD4) and rsp5A401E cells (CHT81) harboring empty vector (pAD4), RSP5 (pAD-RSP5), KIN28 (pAD-KIN28), or POG1 (pAD-POG1). (b) Wild-type cells (CKY8) harboring empty vector (pAD4), RSP5 (pAD-RSP5), KIN28 (pAD-KIN28), or POG1 (pAD-POG1). Approximately 106 cells of each strain and serial dilutions of 10−1–10−5 (from left to right) were spotted and incubated onto SC-Leu, or YPD medium, SC-Leu or YPD medium containing LiCl (50 or 100 mM) at 25°C for 3 days, or onto SC-Leu or YPD medium at 37°C for 4 days.
Overexpression of Kin28 and Pog1 also suppresses the LiCl sensitivity of the rsp5A401E mutant
It was examined whether the multicopy suppressors isolated could restore various stress sensitivities of rsp5A401E cells. The overexpression of Kin28 and Pog1 complements the ionic stress (50 mM LiCl) sensitivity of rsp5A401E cells (Fig. 1a). However, the suppressive effects of these genes to other stresses such as ethanol, NaCl, and H2O2 were not detected (data not shown).
To investigate the biological roles of the suppressors, the nonessential gene, POG1 was disrupted, on the chromosome by one-step gene replacement. However, the wild-type (CKY8) and the POG1-disrupted (CHTDpog1) strains grew well on YPD agar plates at 37 °C or containing 100 mM LiCl at 25 °C (data not shown). The growth phenotypes of wild-type cells overexpressing each suppressor were also tested. Interestingly, yeast cells overexpressing Pog1 were more tolerant to 100 mM LiCl than cells harboring the vector only (Fig. 1b). The overexpression of Kin28 did not affect any stress resistance of wild-type cells. These results showed that the KIN28 and POG1 genes, when present in multicopy, were able to suppress the high temperature on a rich medium and the LiCl sensitivities of the rsp5A401E mutation.
Overexpression of Kin28 and Pog1 increases expression of stress proteins after temperature up-shift
Recently, it was reported that the transcription of stress protein genes, HSP12, HSP42, and DDR2, was defective in the rsp5A401E mutant under various stress conditions (Haitani et al., 2006). The microarray data showed that increases after the addition of ethanol in HSP12, HSP42, and DDR2 mRNAs were considerably lower in rsp5A401E cells than in wild-type cells. Moreover, similar results were obtained from other Hsf1- or Msn2/4-dependent genes (Haitani et al., 2006). These findings suggest that Rsp5 participates in the expression of multiple stress proteins. Because both KIN28 and POG1 encode the transcriptional factor, it was expected that the overexpression of these proteins might induce stress protein genes. To analyze the effect of multicopy suppressors on the expression of stress proteins (Hsp12, Hsp42, and Ddr2), the reporter gene assay was carried out. Wild-type cells, rsp5A401E cells, and rsp5A401E cells overexpressing Kin28 or Pog1, each of which harbors pHS-HSP12, pHS-HSP42, or pHS-DDR2 were grown at 25 °C and exposed to an elevated temperature up to 37 °C. The β-galactosidase activities in wild-type cells were significantly increased after the shift to 37 °C (Fig. 2a). However, the levels of β-galactosidase activity in rsp5A401E cells were significantly lower than those in wild-type cells. In addition, so long as yeast cells were grown at 25 °C, the activity was not elevated in either strain (data not shown).
Reporter gene assay of stress protein gene transcription. Wild-type cells (•), rsp5A401E cells (○), and rsp5A401E cells overexpressing Kin28 (▵) or Pog1 (□), each of which carries PHSP12-lacZ (pHS-HSP12), PHSP42-lacZ (pHS-HSP42), or PDDR2-lacZ (pHS-DDR2), were grown to the logarithmic phase on SC-Ura or SC-Leu-Ura medium at 25°C and subjected to (a) temperature up-shift (from 25 to 37°C) and (b) LiCl (200 mM). Cell extracts from each strain were prepared from samples collected before (time=0) and at several times after (a) temperature up-shift and (b) addition of LiCl, and then the β-galactosidase activities were measured. Values are means of results from three independent experiments. The SDs for these values were <10% of the value of the point.
When Kin28 was overexpressed in rsp5A401E cells, the transformant harboring pHS-HSP12 clearly showed approximately a two- to five-fold increase in β-galactosidase activity. A similar tendency was observed in plasmids pHS-HSP42 and pHS-DDR2. The overexpression of Pog1 in rsp5A401E cells also resulted in about a 30% increase in β-galactosidase activity derived from pHS-HSP12. However, the activities in the strains harboring pHS-HSP42 and pHS-DDR2 were virtually unchanged from those of the strains carrying only the vector. The up-regulation of HSP12 by the overexpression of Kin28 was confirmed by Northern blot analysis (Fig. 3). The expression of HSP12 in CKY8 was strongly induced at 30 min after the shift to 37 °C, but an rsp5A401E mutation attenuated the gene induction. As expected, a dramatic increase of the HSP12 expression occurred in CHT81 overexpressing Kin28 or Pog1 even at 60 min after the shift to 37 °C. These results showed that the overexpression of Kin28 and Pog1 increases the expression of stress proteins.
Northern blot analysis of HSP12 transcription. Wild-type cells (CKY8) harboring empty vector (pAD4) and rsp5A401E cells (CHT81) harboring empty vector (pAD4), KIN28 (pAD-KIN28), or POG1 (pAD-POG1) were grown to the logarithmic phase on SC-Leu medium at 25°C and subjected to temperature up-shift (from 25 to 37°C). Total RNA (10 µg) from each strain was prepared from samples collected before (time=0) and at several times after temperature up-shift, and was hybridized to HSP12. As an internal control, the level of ACT1 encoding actin was also measured in the same blot. At the bottom of the panel, the ratios of the amount of HSP12 mRNA to ACT1 mRNA are shown.
To demonstrate the hypothesis, the HSP12, HSP42, and DDR2 genes were constitutively overexpressed under control of the ADH1 promoter in rsp5A401E cells. When cells were exposed to 37 °C in SC-Leu medium, transformants were capable of growing to the level of wild-type cells (Fig. 4a). Thus, overexpression of stress proteins complemented the high-temperature sensitivity of rsp5A401E cells. The authors think that the increased sensitivity of rsp5A401E cells to high growth temperature mainly results from the reduced induction of stress proteins.
Overexpression of stress protein or proteasome component genes in rsp5A401E cells. (a) Wild-type cells (CKY8) harboring empty vector (pAD4) and rsp5A401E cells (CHT81) harboring empty vector (pAD4), RSP5 (pAD-RSP5), HSP12 (pAD-HSP12), HSP42 (pAD-HSP42), or DDR2 (pAD-DDR2). (b) Wild-type cells (CKY8) harboring empty vectors (pAD4) (pUV2) and rsp5A401E cells (CHT81) harboring empty vectors (pAD4) (pUV2), RSP5 (pAD-RSP5) (pUV2), RPT2 and PRE8 (pAD-PRT2) (pUV-PE8), or RPT2 and PRE9 (pAD-RPT2) (pUV-PRE9). Approximately 106 cells of each strain and serial dilutions of 10−1–10−5 (from left to right) were spotted and incubated onto SC-Leu for 3 days (a) or SC-Leu-Ura medium for 4 days (b).
Overexpression of Pog1 and Kin28 may enhance a proteasome pathway under LiCl stress
The β-galactosidase activity of the transformants exposed to 200 mM LiCl was further measured. As shown in Fig. 2b, the basal levels of the β-galactosidase activity were significantly lower than those after the temperature up-shift. There was no remarkable increase in the activities in CKY8 and CHT81, although the β-galactosidase activities in CKY8 were higher than those of CHT81. Neither Kin28 nor Pog1 caused an increase of β-galactosidase activity in the presence of LiCl. These results suggest that the stress proteins tested here are not primarily involved in the LiCl stress response.
To obtain insights into the response mechanism of cells overexpressing Kin28 and Pog1, DNA microarray analysis of the LiCl treatment was performed. The total RNA samples were extracted from rsp5A401E cells carrying the empty vector, KIN28, or POG1 growing in SC-Leu medium in the absence or in the presence of 200 mM LiCl for 60 min. The expression ratio (60 min/0 min) was calculated in CHT81 (pAD4), CHT81 (pAD-KIN28), and CHT81 (pAD-POG1), and the genes with significant changes were selected in each strain. For the characterization of each strain, the genes in CHT81 (pAD4) were subtracted from the genes in CHT81 (pAD-KIN28) or in CHT81 (pAD-POG1). The induction of 288 genes and the repression of 164 genes were observed in CHT81 (pAD-KIN28) (data not shown). Similarly, 221 induced genes and 264 repressed genes were identified in CHT81 (pAD-POG1) (data not shown).
As shown in Table 1, it appears that the genes involved in the pathway of protein degradation (ubiquitin-mediated proteolysis and proteasome) are slightly induced by overexpression of Kin28 or Pog1. Table 2 lists four genes relative to proteasome (PRE1, PRE9, RPT2, and RPN7) with expression ratio (pAD-KIN28/pAD4 or pAD-POG1/pAD4) more than 1.5-fold in CHT81 (Ferrell et al., 2000). Hence, each gene was constitutively overexpressed under control of the ADH1 promoter in rsp5A401E cells. However, the single gene dosage effect had no influence on the growth of rsp5A401E cells under LiCl stress (data not shown). This is probably because many gene products constitute the subunits or components of proteasome or form the complex together. The growth phenotype of rsp5A401E cells co-overexpressing two genes involved in proteasome with LiCl stress was next tested. It is noteworthy that co-overexpression of RPT2 (ATPase of the 26S proteasome) and PRE8 (proteasome component Y7) or PRE9 (proteasome component Y13) complemented LiCl stress-hypersensitivity of rsp5A401E cells (Fig. 4b). These results suggest that the abnormal proteins generated by LiCl stress were degraded in proteasome in rsp5A401E cells overexpressing Kin28 or Pog1.
List of genes involved in proteosome induced by the overexpression of Kin28 or Pog1
22.6 kDa proteasome subunit
Proteasome component Y13
ATPase of the 26S proteasome
Subunit of the regulatory particle of the proteasome
Gene with expression ration (pAD-kin28/pAD4 or pAD-POG1/pAD4) more than 1.5-fold in CHT81 are indicated by bold letters and underlines.
The accumulation of abnormal proteins in cells under stress is a serious problem. To overcome this, the following two strategies can be considered: (1) degrading the proteins in proteasome or (2) refolding the proteins by molecular chaperones. The results suggest that Kin28 and Pog1 as the multicopy suppressors of an rsp5A401E mutation are involved in the refolding and degradation of the stress-induced abnormal proteins.
In S. cerevisiae, the carboxy-terminal domain of the largest subunit of RNA polymerase II is phosphorylated by Kin28 during early steps in the gene transcription cycle (Cismowski et al., 1995). Previous studies showed that many genes require Kin28 for transcription (Valay et al., 1995). However, the transcription of stress protein genes, such as SSA4, is highly induced even when cells lack Kin28 function (Lee & Lis, 1998). Activation of the heat shock genes involves the essential DNA-bound activator Hsf1; therefore, it seems that Hsf1 has the ability to activate the transcription independently of Kin28 (Sakurai & Fukasawa, 1999). Interestingly, it has been shown recently that the induction of stress protein genes, such as HSP12, HSP42, and DDR2, was defective in rsp5A401E cells under various stresses (Haitani et al., 2006). The Hsf1 and Msn2/4 levels were decreased in rsp5A401E cells, probably due to the accumulation of the mRNAs of HSF1 and MSN2/4 in the nucleus of rsp5A401E cells (Y. Haitani and H. Takagi, unpublished results). It is believed that the decreased amounts of synthesized Hsf1 and Msn2/4 cause the reduced induction of stress protein genes in rsp5A401E cells. It is therefore probable that Kin28, when overexpressed, can induce some stress protein genes in rsp5A401E cells that have been severely depleted of Hsf1 activity.
Pog1 may regulate genes involved in the pheromone response pathway as a transcription factor (Leza & Elion, 1999). The overexpression of Pog1 confers α-factor resistance by inhibiting α-factor-induced G1 arrest and translational repression of the CLN1 and CLN2 genes (Leza & Elion, 1999). By chromatin immunoprecipitation, Pog1 was predicted to bind the promoter of 96 putative genes that function in cell cycle regulation, cytoskeletal organization, and spindle assembly, including GIF1, SLA1, SPC105, and BAR1 (Horak et al., 2002). There was no obvious enrichment for proteolytic genes, but the proteasome subunit RPN7, one of the up-regulated genes in rsp5A401E cells overexpressing Pog1, is a target of Pog1 (Horak et al., 2002). Although the targets of Pog1 were not enriched for genes involved in the stress response, Pog1 overexpression could induce HSP12 after a temperature up-shift. In addition, Kin28 or Pog1 overexpression might induce some of the genes involved in proteasome under LiCl stress. These may explain how Kin28 or Pog1 overexpression confers resistance to a high growth temperature and LiCl in rsp5A401E cells.
It is also unclear whether Rsp5 directly regulates the KIN28 and POG1 genes. A multicopy-suppressor strategy can be a good way to identify new players in a process of interest but it is also possible that the players identified represent artifacts, which are not really involved in the process. It has not been reported that Kin28 and Pog1 do not play a role in transcription of the stress protein genes. In addition, DNA microarray analysis showed virtually similar amounts of the KIN28 and POG1 transcripts in strains CKY8 and CHT81 (data not shown). These results suggest that Kin28 and Pog1 are not on a specific pathway between Rsp5 and the candidate genes but happen to impact the process through the Rsp5-independent pathway when overexpressed. To clarify the relationship between Rsp5 and these multicopy suppressors under various stresses, the interaction of these molecules must further be analyzed. In terms of application, the overexpression of proteins such as Kin28 and Pog1, which can enhance the degradation or refolding of the stress-induced abnormal proteins, would be a useful method for breeding novel stress-resistant yeast strains.
M.D. and Y.M. contributed equally to this study.
The authors thank Drs Chris A. Kaiser (Massachusetts Institute of Technology), Satoshi Harashima (Osaka University, Japan) and Jun-ichi Nikawa (Kyushu Institute of Technology, Japan) for providing the yeast materials and Hiroyuki Hiraishi of our laboratory for helpful discussion. This work was supported by a grant to H.T. and J.S. from the Program for Promotion of Basic Research Activities for Innovative Biosciences (PROBRAIN).