RACK1/Asc1p, a Ribosomal Node in Cellular Signaling

RACK1/Asc1p and its essential orthologues in higher eukaryotes, such as RACK1 in metazoa, are involved in several distinct cellular signaling processes. The implications of a total deletion have never been assessed in a comprehensive manner. This study reveals the major cellular processes affected in a Saccharomyces cerevisiae Δasc1 deletion background via de novo proteome and transcriptome analysis, as well as subsequent phenotypical characterizations. The deletion of ASC1 reduces iron uptake and causes nitrosative stress, both known indicators for hypoxia, which manifests in a shift of energy metabolism from respiration to fermentation in the Δasc1 strain. Asc1p further impacts cellular metabolism through its regulative role in the MAP kinase signal transduction pathways of invasive/filamentous growth and cell wall integrity. In the Δasc1 mutant strain, aberrations from the expected cellular response, mediated by these pathways, can be observed and are linked to changes in protein abundances of pathway-targeted transcription factors. Evidence of the translational regulation of such transcription factors suggests that ribosomal Asc1p is involved in signal transduction pathways and controls the biosynthesis of the respective final transcriptional regulators.

The G␤-like WD40-repeat protein Asc1p of Saccharomyces cerevisiae is an integral component of the small 40S ribosomal subunit (1,2). Because of the distinct seven-bladed propeller structure of Asc1p and its exposed localization near the ribosomal mRNA exit tunnel, Asc1p depicts an eminent platform for protein-protein interactions and a nexus to the translational apparatus (1,3).
Genome-wide genetic, biochemical, and interaction studies have defined ASC1/Asc1p as a potential interaction partner for a multitude of various genes/proteins (4 -7). Asc1p specifically interacts with the mRNA-binding protein Scp160 (4,5) and has been shown to be required for the interaction of Scp160p with the ribosome, suggesting a role for Asc1p as an adapter protein, bringing specific mRNAs, translation factors, and the ribosome in proximity (8). An extended Asc1p network (Smy2p, Eap1p, Scp160p, and Asc1p, known as the SESA network) has been identified to specifically inhibit the translation initiation of the POM34-mRNA, encoding an integral membrane protein of the nuclear pore complex (9). The influence of Asc1p on the translational process is further evident in the increased Gcn2p-kinase-dependent phosphorylation of eukaryotic translation initiation factor 2 (eIF2) in a ⌬asc1 background and increased levels of phosphorylated eIF4A, an RNA helicase, when ASC1 is deleted (10). In addition to encoding for Asc1p, the ASC1 locus harbors an intron coding for the small nucleolar RNA (snoRNA) U24. U24 is required for the maturation of the 60S ribosomal subunit via site-specific 2Ј-O-methylation of the 28S rRNA (11). As a result, snoRNA U24 deletion results in reduced levels of mature 60S ribosomal subunits, producing a halfmer phenotype in polysome experiments, which might affect translation efficiency (12).
The particular structure of Asc1p is highly conserved among eukaryotes, including mammalian RACK1, which was originally found as a receptor of activated protein kinase C (PKC) 1 (13). RACK1 has since been described as an anchoring and translocation-mediating protein for several other cellular proteins, including Ras-GAP, dynamin-1, Src, several integrins (14), and the cAMP-specific phosphodiesterase PDE 4 D 5 (15). RACK1 presumably recruits these proteins to their appropriate subcellular sites, thereby integrating their actions into intracellular signaling pathways (14). Accordingly, RACK1 has recently been shown to enable cross-talk between the PKC and cAMP signaling pathways by mediating the PKCdependent activation of PDE 4 D 5 in response to elevated intracellular cAMP levels (16).
Correspondingly, the deletion of ASC1 in S. cerevisiae and its orthologues in other eukaryotic organisms leads to pleiotropic phenotypes based on significant misregulations in signal perception and transduction (10,17). As a result, in S. cerevisiae, developmental differentiation is affected when ASC1 is deleted, resulting in the absence of filamentous growth (haploid invasive or diploid pseudohyphal) and severely compromised cell wall integrity (10,18). Furthermore, Asc1p's involvement in the organization of cellular respiration and fermentation is suggested by its initial characterization as absence of growth suppressor of a hem1 Ϫ cyp1 Ϫ /hap1 Ϫ mutant strain in S. cerevisiae (19). In Schizosaccharomyces pombe, Cpc2p (Asc1p in S. cerevisiae) regulates sexual differentiation in a complex with Moc1p/Sds23p and Moc2p/ Ded1p (20). The Asc1p complement Rak1 of Ustilago maydis interacts with several ribosomal proteins and has been shown to regulate virulence and mating by influencing the mRNA levels of the transcriptional activator Rop1 (21). Also in higher eukaryotes, RACK1 is required for several developmental processes, including seed germination, root formation, leaf production, and flowering in Arabidopsis thaliana (22,23). RACK1 of Drosophila melanogaster is expressed in many tissues with a specific requirement at multiple steps of development (24). Studies of human cell lines revealed that RACK1 influences cellular processes that are directly connected to cell proliferation and cell cycle progression (25). Thus, RACK1 has been repeatedly described in the context of uncontrolled cell division and as a contributing factor in in vivo tumor growth (26,27). It is up-regulated during angiogenesis, as well as in colon carcinoma, non-small cell lung carcinoma (28), and melanomas (29).
Because of the essential function of RACK1 in developmental processes, its deletion is lethal even at early embryonic stages and can therefore not be examined in higher eukaryotes such as plants or metazoa (30). In S. cerevisiae, Asc1p is also involved in cell development, evident by the inability of a ⌬asc1 strain to grow adhesively or form pseudohyphae, but the consequences of an ASC1 deletion in this simple eukaryotic model organism are far less severe. This allows the investigation of a total ASC1 deletion in S. cerevisiae to determine the cellular and molecular function of the highly conserved eukaryotic protein Asc1p.
This study is based on a proteome and transcriptome analysis of an S. cerevisiae ⌬asc1 strain. Together with phenotypical and molecular characterizations, it delivers functional groups of proteins and mRNAs with an altered expression in response to the deletion of ASC1 and determines affected cellular processes. We show that Asc1p post-transcriptionally regulates the abundances of transcription factors involved in the MAPK signaling pathways of invasive/pseudohyphal growth and pheromone response. Furthermore, cell wall integrity, regulated by the Pkc1p-MAPK pathway, as well as iron homeostasis and energy metabolism, is imbalanced in a ⌬asc1 mutant.

EXPERIMENTAL PROCEDURES
Yeast Strains and Growth Conditions-The Saccharomyces cerevisiae strains used in this work were of the ⌺1278b background and are listed in Table I. Strain RH3428 was derived from RH2816 by deleting the ASC1 gene with a URA3-cassette, as described by Gueldener et al. (31). The deletion strains RH3461-RH3464 were obtained via am-  Gueldener et al. (31), was performed with strains RH3463 and RH3464 prior to the deletion of YHB1 as described above, yielding strains RH3497 and RH3498. For the construction of strain RH3510, the loxP-URA3-loxP cassette was amplified from pUG72 with the oligonucleotides 5Ј-GACTTTGATCTCCTG-GAAGTTGACTGGTGACGACCAAAAGTTTGGAGCAGCTGAAGCTT-CGTACGC-3Ј (forward) and 5Ј-TCTTGGACAATGTGACTGTGACC-CTTGAAAGATCTAACTGGGACAGCATAGGCCACTAGTGGAT-CTG-3Ј (reverse), containing sequences identical to regions of exon 1 in the ASC1 locus. The resulting fragment was inserted into the genome via homologous recombination. Subsequently, the URA3 gene was removed by the expression of the Cre recombinase from plasmid pSH63, leaving behind one loxP site in exon 1 (supplemental Fig. S2) (31). Strains RH3429 -RH3436 and strains RH3478, RH3479, RH3503, RH3504, and RH3505 were generated by C-terminally tagging HBN1, YHB1, PHD1, FLO8, and STE12 with 3xmyc according to the procedure described by Janke and colleagues (32). Transformations were carried out according to the lithium acetate method (33). Cultures were grown at 30°C overnight in 10 ml liquid minimal yeast nitrogen base media (YNB: 1,5 g/l yeast nitrogen base lacking amino acids and ammonium sulfate, 5 g/l ammonium sulfate, 2% glucose) containing respective supplements, diluted and cultivated in main cultures to midlog phase before the isolation of protein extracts. Experiment-specific growth conditions are given in the respective paragraphs. Plasmid Construction-All plasmids used in this study are listed in Table II. To construct plasmids pME3751-pME3755, the lengths of the 5ЈUTRs of respective transcription factors were obtained from the European Bioinformatics Institute Web site (34), and the fragments were amplified and inserted in the testing vector pME3680 as described by Rachfall et al. (35) (supplemental Table S1). Plasmids were amplified in the Escherichia coli strain DH5␣ in LB medium with 100 g/ml ampicillin.
De Novo Proteome and Two-dimensional PAGE Analysis-Yeast cultures were grown to midlog phase (OD 600 ϭ 0.8) prior to the addition of 450Ci Met-[ 35 S]-label (Hartmann Analytik, Braunschweig, Germany) and further incubated for an additional hour. Cells were harvested and washed in wash buffer (0.1 M Tris, pH 8.0, with 1% DMSO and 1 mM PMSF). Protein extracts were obtained by lysing cells with Y-PER ® Plus reagent (#78999, Pierce, Rockford, IL) containing complete Protease Inhibitor Mixture (Roche Diagnostics GmbH, Mannheim, Germany) and purifying via methanol-chloroform extraction (36). Protein concentrations were determined using a BCA Protein Assay kit from Pierce (#23227). 100 g of purified protein extracts were used in two-dimensional PAGE analyses. Gels were silver-stained according to the procedure of Blum et al. (37), vacuum dried, and exposed on imaging plates (Fuji, Tokyo, Japan) for 2 weeks. The protein-spots of the resulting autoradiographies were quantified with the image analysis software PDQuest™ (Bio-Rad, Munich, Germany). Protein-spots were submitted to LC-MS/MS upon a regulation by at least 2-fold. The analysis was performed for seven biologically independent replicates.
LC-MS/MS Protein Identification-Excised polyacrylamide gel pieces of stained protein-spots were digested with trypsin according to the procedure presented by Shevchenko et al. (38). Tryptic peptides extracted from each gel piece were injected onto a reversephase liquid chromatographic column (Dionex-NAN75-15-03-C18 PM) utilizing the ultimate HPLC system (Dionex, Idstein, Germany) to further reduce sample complexity prior to mass analyses with an LCQ DecaXP mass spectrometer (Thermo Scientific, San Jose, CA) equipped with a nano-electrospray ion source. Cycles of MS spectra with m/z ratios of peptides and four data-dependent MS2 spectra were recorded via MS. The "peak list" was created with extractms provided with the Xcalibur software package (BioworksBrowser 3.3.1, Thermo Scientific). The MS2 spectra with a total ion current higher than 10,000 were used to search for matches against a yeast genome protein sequence database from the National Center for Biotechnology Information (NCBI) Saccharomyces Genome Database (SGD) (Stanford, CA; March 2005, 6882 sequences, plus 180 sequences of the most commonly appearing contaminants such as keratins and proteases, provided with the BioworksBrowser package) using the TurboSEQUEST algorithm (39) of the Bioworks software. The search parameters based on the TurboSEQUEST software included (i) precursor ion mass tolerance less than 1.4 amu; (ii) fragment ion mass tolerance less than 1.0 amu; (iii) up to three missed tryptic cleavages allowed; and (iv) fixed cysteine modifications by carboxyamidomethylation (plus 57.05 amu), variable modifications by methionine oxidation (plus 15.99 amu), and phosphorylation of serine, threonine, or tyrosine (plus 79.97 amu). At least two matched peptide sequences of identified proteins had to meet the following criteria: (i) the rank by primary score must be 1 for all peptides, (ii) the peptide probability value must be below 0.1, and (iii) the cross-correlation score of matches must be greater than 1.5, 2.0, and 2.5 for peptide ions of charge states 1, 2, and 3, respectively. Each protein included in further analyses had to be identified according to these criteria in at least two independent proteome analyses performed (supplemental Table S2). Peptides of identified proteins were individually blasted against the SGD database (BLASTP against the dataset "Protein Encoding Genes") to ensure their unambiguous assignment to the TurboSEQUEST-specified protein.
Scintillation Counting-Metabolically labeled protein extracts were separated from free Met-[ 35 S]-isotope via methanol-chloroform extraction (36), and the sediment was resuspended in 400 l Y-PER ® Plus reagent. 200 l of the suspension were mixed with 5 ml of ULTIMA GOLD XR LCS-Mixture (Perkin Elmer, Boston, MA) and subjected to scintillation counting. The measured counts per minute were related to the concentration of protein extracts determined using a BCA Protein Assay kit (Pierce). The insoluble fraction was washed in wash buffer and incubated in Y-PER ® Plus reagent overnight. After being washed in H 2 O, the pellet was weighed and resuspended in 1 ml H 2 O. 100 l of this suspension were added to 5 ml of ULTIMA GOLD XR LCS-Mixture, and this sample underwent scintillation counting. The RNA Isolation, cDNA Labeling, and Hybridization-Yeast cultures were grown in 1l minimal medium (YNB) at 30°C to midlog phase. Total RNA was isolated following the protocol described by Schmitt and colleagues (40). For the preparation of fluorescently labeled cDNA, 25 g of total RNA were treated according to the instructions for the CyScribe First-Strand cDNA Labeling Kit (GE Healthcare, Munich, Germany). Primer annealing was carried out with random nonamers and anchored oligo(dT), and for the extension reaction, Cy3 or Cy5 labeled dCTP (GE Healthcare) was used. Labeled cDNA was purified according to the CyScribe GFX Purification Kit (GE Healthcare). Before hybridization, samples were denatured at 98°C for 5 min. The hybridization was carried out in Tom Freeman hybridization buffer (41) for 15 h at 58°C with cDNA containing ϳ60 pmol of Cy3 and Cy5 in an automatic Lucidea slide processor (GE Healthcare). Slides were washed with two consecutive washes with 1x SSC buffer containing 0.2% SDS and with 0.1x SSC. Finally, the hybridization chambers were flushed with isopropanol, and the slides were dried via evaporation. Scanning was done using a GenePix 4000B microarray scanner (Molecular Devices, Sunnyvale, CA) with the GenePix Pro 6.0 software.
Data Analysis-To normalize the data, the arithmetic mean of the ratios was set equal to 1 (GenePix Pro 6.0 software). Only features with fluorescence greater than the local background plus one standard deviation were included in the analysis. To correct for features with irregular spot morphology, we included only those differing by less than 30% in the ratio of medians, the ratio of means, and a regression ratio. The presented transcriptional data are the result of three biologically independent experiments. For each biological replicate, two technical replications and dye-swaps were analyzed. Genes were considered significantly up-or down-regulated if the logarithmic (to the basis of two) expression ratio of medians was Ն0.59 or ՅϪ0.59 for a gene in all three biological replicates, corresponding to a transcriptional change of at least 1.5-fold. Additionally, the statistical error could not exceed 5%. Microarray data were deposited in the NCBI Gene Expression Omnibus database (Accession number GSE 34286, http://www.ncbi.nlm.nih.gov/geo/query/ acc.cgi?token ϭ bjuntwckiqiiyfw&acc ϭ GSE%2034286).
Spot Growth Assays-Yeast cell main cultures were grown to midlog phase, adjusted to OD 600 ϭ 0.1, and serially 10-fold diluted up to 1:10,000. 10 or 20 l of each dilution were spotted onto yeast extract-peptone dextrose (YEPD) or YNB plates, respectively, containing 2% agar and the denoted supplements. YNB plates containing 30 M of the iron (II) chelator bathophenanthrolinedisulfonic acid (BPS) were used to determine the sensitivity to iron limitation. The control plates contained 100 M (NH 4 ) 2 Fe(SO 4 ) 2 . Cell wall integrity was scored on YEPD medium containing 125 g/ml congo red. Aerobic growth was determined on yeast extract peptone medium containing 2% glucose or 2% ethanol (EtOH) as a carbon source.
After 3 days of incubation at 30°C, the plates were washed under a gentle stream of water.
Growth Conditions for Glucose-dependent Colony Morphology Alterations-Yeast cells were grown to midlog phase and diluted to 2 ϫ 10 3 cells/ml. 150 l of the cell suspension were plated onto YEPD medium with 0.4% or 2% glucose, and the plates were incubated at 30°C for 3 days before examination of the colonies.
Assay to Score for Respiratory Activity-Yeast cells were grown to midlog phase and diluted to 2 ϫ 10 3 cells/ml. 150 l of the cell suspension were plated onto minimal medium (YNB) with 0.4% glucose to establish preferential conditions for respiration. The plates were incubated at 30°C for 3 days. To determine the respiratory activity, 10 ml of triphenyltetrazolium chloride (TTC) agar (1.2% agar and 0.5% TTC) were overlaid, and red and white colonies were photographed after 20 min of incubation at 30°C. Red colonies indicate the reduction of TTC, which requires the activity of the respiratory chain, whereas colonies impaired in their respiratory chain remain white (42).
Determination of Glucose Consumption and Ethanol Production-Cell main cultures were grown in YNB medium at 30°C to OD 600 ϭ 0.5. Thereafter, samples of 500 l were taken every hour for 8 h. Additionally, the cell density was determined at the respective time points. Proteins were precipitated with 0.4 ml 0.33 M perchloric acid and chilled. After neutralization with 0.1 ml 1.25 M NaOH, the proteins were sedimented via centrifugation at 12,000 rpm for 5 min. The supernatants were diluted 1:10 with H 2 O prior to the determination of ethanol or glucose concentrations. To determine the ethanol production, 100 l of the diluted supernatant, 100 l standard (2 mM EtOH), and 100 l YNB (blank) were added to 3 ml E-buffer (26.7 g pyrophosphate, 6.7 g semicarbazide, 1.26 g glycine, 800 ml H 2 O), 100 l 13 mM NAD, and 100 l 13.84kU ADH from S. cerevisiae, respectively. The solution was incubated at 37°C for 25 min before the optical density was measured at 363 nm. Standard and blank values served for quantification and normalization, respectively. For the determination of glucose consumption, the glucose concentration in the medium was measured. 50 l of the diluted supernatant, 50 l standard (5 mM glucose), and 50 l H 2 O (blank) were added to 1 ml glucose test solution (Glucose-Hexokinase kit, DiaSys Diagnostic Systems GmbH, Holzheim, Germany). OD 363 was measured after 15 min of incubation at room temperature. Glucose concentrations were determined according to standard and blank values.
␤-galactosidase Assay-Assays were performed with extracts grown in liquid minimal medium. 10 ml pre-cultures were grown overnight at 30°C, and 1 ml was used to inoculate 10 ml main cultures. Cells were harvested after 6 h of growth at 30°C. Protein extracts were obtained via lysing with glass beads in breaking buffer (100 mM Tris pH 8.0, 20% glycerol). Specific ␤-galactosidase activities were normalized to the total protein amount (43) in each extract and calculated in accordance with the work of Rose and Botstein (44) ((A 415 ϫ 0.3)/(0.0045 ϫ protein concentration ϫ extract volume ϫ time)). The values obtained for the inserted 5ЈUTR sequences were related to the values obtained for the empty testing vector for wild type and ⌬asc1 strain, respectively.

De Novo Proteome Analysis: Hypoxic Energy Metabolism in ⌬asc1
Cells-Given Asc1p's potential impact on translation as ribosomal G␤-like protein, the deletion of its encoding gene is expected to result in decisive changes in cellular gene and protein expression. We compared the proteomes of ⌬asc1 and an isogenic wild type S. cerevisiae strain to assess those changes and identify affected proteins and processes. The incubation with Met-[ 35 S]-isotope enables the visualization of the de novo protein biosynthesis in each cell population of 1 h (supplemental Fig. S1). Subsequent cell lysis with Y-PERா Plus reagent resulted in two separate radioactive fractions: the soluble protein extracts, used for two-dimensional PAGE analysis, and the insoluble remains (supplemental Fig. S1). In order to determine qualitative differences in the de novo protein biosynthesis rates of wild type and ⌬asc1 mutant, both fractions were analyzed by means of scintillation counting to detect the respective amounts of incorporated radioactive methionine (Fig. 1). The soluble fraction represents a measure for the de novo biosynthesis of soluble proteins, whereas the insoluble fraction corresponds to newly synthesized insoluble proteins (e.g. cell wall or plasma membrane proteins) (45). Relative to the wild type, the ⌬asc1 mutant showed elevated de novo biosynthesis of soluble proteins. The measured activities for the deletion strain were about 50% higher than for the wild type. For the insoluble fraction, in contrast, activities were reduced by nearly a quarter in the ⌬asc1 strain. This converse effect on the de novo biosynthesis of soluble and insoluble proteins indicates a supporting function of Asc1p in cell wall and/or membrane biogenesis. This finding is affirmed by the results of the de novo proteome analysis. Although insoluble cell wall proteins generally do not appear in two-dimensional proteome analyses of the soluble fraction, some proteins involved in cell wall biogenesis and overall cell morphology were found to be significantly reduced in ⌬asc1 cells ( Fig. 2A). Pdi1p and Mlc1p are essen-tial for viability and appear to be positively controlled by Asc1p. Pdi1p is a protein disulfide isomerase required for the formation of disulfide bonds in secretory and cell surface proteins. Mlc1p localizes to the bud neck and is involved in actin ring formation, promoting targeted membrane deposition at the bud neck during cytokinesis. In contrast, the protein levels of the guanylate kinase Guk1p, required for mannose outer chain elongation of cell wall N-linked glycoproteins, were increased 7-fold in the mutant strain.
By far the largest group, consisting of almost 50% of all proteins found to be regulated in ⌬asc1 cells, is made up of proteins involved in energy metabolism ( Fig. 2A). This comprehensive group is composed of proteins taking part in glycolysis, mitochondrial biogenesis and respiration, oxidative stress, and fermentation. Representative protein-spots for each group are illustrated in Fig. 2B.
The six proteins involved in glycolysis span nearly the entire glycolytic process. Cdc19p, Fba1p, Pgi1p, and Tpi1p were down-regulated in ⌬asc1 cells, but Eno1/2p and Gpm1p were found to be increased. A negative genetic interaction was previously established between ASC1 and GCR2, coding for a transcriptional activator of genes involved in glycolysis (7). Glycolysis is directly connected to the energy restoring and generating processes in the cell via the delivery of substrates for either respiration in the mitochondria or fermentation in the cytosol.
For the four proteins assigned to the group of mitochondria and respiration, a similar differential expression was found. The D-lactate dehydrogenase Dld2p resides in the mitochondrial matrix. Its biosynthesis was reduced in the ⌬asc1 strain and for Rdl1p, a protein of the outer mitochondrial membrane. Ssc1p and Yhb1p showed 4.5-fold and 2.3-fold increases, respectively. Ssc1p is an essential constituent of the import motor component of the translocase of the inner mitochondrial membrane (TIM23 complex). The yeast flavohemoglobin Yhb1p acts The activity for the soluble fraction was normalized against the total protein content of the respective extract as determined via BCA protein assay, whereas the activity of the insoluble fraction was normalized against the weight of the respective pellet.
as nitric oxide (NO) oxidoreductase required for NO detoxification (46). NO is believed to lead to the inhibition of mitochondrial respiration and enhance the production of reactive oxygen species (ROS) (47), which are increasingly produced during compromised cellular respiration and cause oxidative damage within the cell (48). In this analysis, four proteins were assigned to the "oxidative stress" group. The nitroreductase-like protein Hbn1p has been described as important for superoxide dismutase activity (49). Whereas the other three proteins involved in oxidative stress response-namely, Ahp1p, Hor2p/Rhr2p, and Tsa1p-showed reduced expression, Hbn1p biosynthesis was increased 8-fold in the ⌬asc1 strain. FIG. 2. Proteins identified as being regulated in ⌬asc1 cells. A, proteome analysis. Regulated proteins were classified according to their cellular function or localization. The regulation for each protein is given as the average relative spot intensity (spot intensity (⌬asc1)/spot intensity (wild type)). The percent distribution of proteins in the functional groups is illustrated in a pie chart relative to the number of total proteins found to be regulated. The composed group of energy metabolism is accentuated. B, close-up views of selected protein-spots regulated in ⌬asc1 cells. Protein names are given according to mass spectrometric identifications. The corresponding proteins belong to the joined group of energy metabolism making up 46% of all regulated proteins. For each of the proteins, its average regulation is displayed relative to that of the wild type (Ø (⌬asc1/wild type)).
In the context of energy metabolism, it is especially striking that all five proteins assigned to the process of fermentation were uniformly increased at least 2-fold in the absence of Asc1p. This group includes, among others, the alcohol dehydrogenase Adh1p and the pyruvate decarboxylases Pdc1p and Pdc5p, which are key enzymes of alcoholic fermentation in yeast.
Taken together, the majority of processes found to be affected in the ⌬asc1 strain are involved in energy metabolism. Additionally, a strong down-regulation of proteins taking part in cell wall biogenesis and maintenance could be observed. According to scintillation counting, the de novo biosynthesis of insoluble proteins was reduced by 25% in ⌬asc1 cells.
Furthermore, significant changes in the expression of several proteins involved in the maintenance of cell morphology were revealed in the ⌬asc1 strain. Proteins involved in energy metabolism, including glycolysis, mitochondria biogenesis and respiration, oxidative stress, and fermentation, amount to almost half of all regulated proteins. In particular, proteins involved in fermentation showed a distinct increase in response to the deletion of ASC1, suggesting elevated fermentation in ⌬asc1 cells.
Transcriptome Analysis: Asc1p Is Involved in Several Cellular Signaling Pathways-The microarray analyses revealed 80 genes that were considered differentially expressed in the ⌬asc1 strain relative to the wild type (Fig. 3), according to FIG. 3. Functional classification of mRNAs regulated in ⌬asc1 cells. The regulation for each mRNA is given as average relative to the wild type (wt) (spot intensity (⌬asc1)/spot intensity (wt)). The groups resulting from the respective function of the mRNA products are expressed as the percentage of the total number of regulated mRNAs. The complete data set is supplied in supplemental Table S3. stringent thresholds and experimental settings (see Experimental Procedures). Significant up-regulation was found for mRNAs identified as transposable elements. This group consists solely of retrotransposons of the Ty1 and Ty2 type. Corresponding to their common promoter, they uniformly showed up-regulation in the ⌬asc1 strain.
Additionally, the expression of many mRNAs involved in energy metabolism was affected in the ⌬asc1 strain. They are presented in the groups of glucose transporters, mitochondria/respiration, and oxidative stress (Fig. 3). The affected mRNAs of the first group code for the low-affinity transporter Hxt3p and the two almost identical high-affinity transporters Hxt6p and Hxt7p. All three corresponding proteins reside in the plasma membrane and showed significant down-regulation in their mRNA level in the mutant strain. Among the mRNAs assigned to the "oxidative stress" group, DDR2 and PEP4 could be found as down-regulated in ⌬asc1 cells. They code for proteins expressed in response to multiple stressors and oxidative-damage-induced protein turnover, respectively. Additionally, the expression of HBN1, coding for a putative nitroreductase, was 3-fold up-regulated in the mutant strain. The group of mitochondria and respiration consists of seven regulated mRNAs. MDH1, RIB4, ATP1, ATP2, and INH1 were down-regulated in the ⌬asc1 strain. MDH1 codes for the mitochondrial malate dehydrogenase of the tricarboxylic acid cycle. Rib4p is required for riboflavin synthesis, which is a precursor of the flavin coenzymes flavin adenine dinucleotide (FAD) and flavin mononucleotide. As a non-covalently bound part of the NADH:ubiquinone oxidoreductase Ndi1p, FAD is involved in electron transfer from the tricarboxylic acid cycle to ubiquinone in the electron transport chain at the inner mitochondrial membrane (50,51). ATP1 and ATP2 code for the ␣ and ␤ subunits, respectively, of the F 1 sector of the mitochondrial F 1 F 0 ATPase synthase. ATPase synthase activity is required for the synthesis of ATP, which is inhibited by Inh1p. Additionally, IRC18 was found to be up-regulated, a known indicator for respiratory deficiency (52). The transcription of YHB1, encoding for NO oxidoreductase, was increased in the mutant strain.
Another group of special interest in the context of energy metabolism consists of genes encoding proteins involved in iron homeostasis, as iron is directly involved in respiration via heme-biosynthesis and iron-sulfur cluster formation (53). All proteins corresponding to the mRNAs in this group have their function in the uptake of iron, except for Tis11p, which is an mRNA-binding protein promoting mRNA degradation under iron starvation. FRE1 codes for the major cell-surface iron reductase, reducing iron prior to its uptake. Ccc2p is required for the translocation of the Fet3p/Ftr1p complex to the plasma membrane, which mediates high-affinity iron uptake. Enb1p, Sit1p, and Arn1p are directly linked to iron uptake as members of the Arn family of siderophore iron transporters. Except for Ccc2p, all corresponding proteins involved in iron uptake showed down-regulation in their mRNA level in the ⌬asc1 mutant strain and reside in the plasma membrane. The negative influence on the transcription of genes coding for membrane proteins in ⌬asc1 cells suggested by the mRNAs associated with iron uptake and glucose transport further becomes apparent in the group of cell wall/morphology. CWP2, coding for a major constituent of the cell wall, as well as HOR7, HPF1, and SCW4, coding for additional cell wall components, were clearly down-regulated in ⌬asc1 cells. Additionally, PSA1, coding for an essential protein required for a normal cell wall structure, was found to be down-regulated in the mutant strain. Furthermore, the two mRNAs up-regulated in this group-namely, UTR2 and WSC3-are described as being induced under cell wall stress. The integral membrane protein Wsc3p acts as sensor-transducer of the Pkc1p-MAPK cell wall integrity pathway (54,55).
According to the presented transcriptome data, transposition seems to be a process up-regulated in the ⌬asc1 mutant strain. Additionally, Asc1p takes part in the regulation of iron homeostasis and herein seems to be specifically required for the uptake of iron. This is supported by an elevated sensitivity of the ⌬asc1 strain to the iron (II) chelator BPS relative to the wild type, suggesting that the absence of Asc1p leads to lower iron concentrations within the cell (Fig. 4).
The transcriptional changes further indicate a distinctive role for Asc1p in cell wall integrity. Furthermore, processes involved in energy metabolism, represented by the regulated groups of glucose transporters, mitochondria/respiration, and oxidative stress, are affected by the deletion of ASC1. Especially the particular regulation of mRNAs involved in respiration suggests the significance of Asc1p in the processes of oxygen and energy metabolism. The transcriptome analysis corroborates the previous indication obtained via proteome analysis that Asc1p has a distinct impact on energy metabolism ( Fig. 2A). The cumulative data obtained via proteome and microarray analyses reveal that several distinct biological processes are affected by Asc1p and suggest its involvement in various cellular signaling pathways. The ⌬asc1 Strain Shows Enhanced Sensitivity to Nitrosative Stress-Two especially interesting candidates found to be regulated in both the cellular proteome and the transcriptome are HBN1/Hbn1p, assigned to the group of mitochondria and respiration, and YHB1/Yhb1p, assigned to the group of oxidative stress. Both rank among the most significantly upregulated candidates in both data sets. Their protein abundances, determined by means of Western hybridization, emphasize the significant impact of Asc1p on their expression (Fig. 5A). The total protein amounts of Hbn1p and Yhb1p showed up-regulation in the ⌬asc1 strain by 8-and 4-fold, respectively. Because of these strong effects and the involve-ment of Hbn1p and Yhb1p in protection against reactive nitrogen oxide species (RNOS), the sensitivity of the ⌬asc1 strain, as well as that of the ⌬hbn1, ⌬yhb1, respective double deletion strains and the triple mutant ⌬asc1⌬hbn1⌬yhb1, to the NO-donor (Z)-1-[N-(2-aminoethyl)-N-(2-ammonioethyl)amino]diazen-1-ium-1,2-diolate (DETA-NONOate) was tested via agar diffusion assay (Fig. 5B). The ⌬asc1 strain and the ⌬hbn1 and ⌬yhb1 strains showed larger areas of no growth (halos) than the wild type strain, indicating increased intracellular nitrosative stress in the mutant strains and/or impaired RNOS detoxification. The ⌬asc1 cells were slightly more sensitive to NO when they were additionally lacking either Yhb1p FIG. 5. Evaluation regarding nitrosative stress. A, immuno-detection for candidates significantly regulated on the mRNA and protein levels in ⌬asc1 cells. myc 3 -tagged variants of Hbn1p and Yhb1p were detected by anti-myc antibody. eIF2␣ signals display an internal loading control and served for normalization of Hbn1p and Yhb1p signals. The results are illustrated relative to wild type in the adjacent graphs. B, agar diffusion/halo assay to determine sensitivity toward the nitric oxide donor DETA-NONOate. Cells were plated on MV plates (pH ϭ 7.2) embedded in liquid top agar, and this was followed by the application of filter papers containing 2 l or 4 l DETA-NONOate (1 M). Respective halo sizes were measured after 2 days of growth at 30°C. The graph shows the resistance of the deletion strains relative to the wild type, determined by the reciprocal values of the halo diameters. C, spot assay to score total growth. Growth of 10-fold diluted cell suspension identical to cultures used in halo assays was monitored on MV plates (pH ϭ 7.2). or Hbn1p, as was also observed for the ⌬hbn1⌬yhb1 double mutant. The triple deletion strain ⌬asc1⌬hbn1⌬yhb1 seemed to be more sensitive to DETA-NONOate, which might be attributed to its strong general growth defect (Fig. 5C). This observation implies a genetic interaction of ASC1 with both YHB1 and HBN1. It further suggests that the up-regulation of Yhb1p and Hbn1p in the ⌬asc1 strain is a protective response to an increased nitrosative burden, possibly intrinsically generated in the mutant strain. Additionally, the asc1SNR24 strain, deficient in Asc1p but still expressing the snoRNA U24, encoded by the intron of the ASC1 locus, was analyzed. The displayed sensitivity to DETA-NONOate and its general growth phenotype correspond to the results obtained for the total ASC1 deletion strain ⌬asc1, showing that NO sensitivity is an Asc1p-, and not a U24-, dependent phenotype.
Cell Wall Integrity is Compromised in ⌬asc1 Cells-Because of the indications of compromised cell wall integrity in ⌬asc1 cells, we performed growth assays on medium supplemented with the chitin-interacting reagent congo red (56). The ⌬asc1 strain showed significantly enhanced sensitivity to congo red relative to the wild type (Fig. 6A). This is further evidence of a controlling function of Asc1p in the maintenance of cell wall integrity, as has been indicated by the global expression data. The role of Asc1p in cell wall integrity is further displayed by the differences in colony morphology of the ⌬asc1 mutant and the wild type strain under reduced glucose concentrations (0.4%) (Fig. 6B). Wild type colonies formed a distinct loopy structure when grown with a reduced glucose supply, as has been described previously for yeast mat formation (57). This effect is linked to the expression of the cell surface glycoprotein Flo11p and is therefore absent in a ⌬flo11 strain. Similar to the ⌬flo11 strain, ⌬asc1 colonies also lacked the structure formation displayed by the wild type strain and merely displayed a slightly smaller colony size (Fig. 6B). This effect might be attributed to the significantly impaired ability of the ⌬asc1 strain to express the cell wall gene FLO11, indicated by the lack of adhesion under glucose or amino acid limitations (10) (Fig. 6C). The results for the ASC1 mutant strain asc1SNR24 confirm that the observed phenotypes are a consequence of the absence of Asc1p in the ⌬asc1 strain, and not of the snoRNA U24.
Respiration is Restrained in ⌬asc1 Cells-The presented proteome and transcriptome analyses suggest that Asc1p promotes respiration, whereas its absence induces fermentation. To confirm this influence of Asc1p on energy metabolism, the glucose consumption and ethanol production of wild type and ⌬asc1 cultures were monitored over a period of 8 h (Fig. 7A). The data show that ⌬asc1 cells consume increasingly more glucose than wild type cells. This suggests ele-FIG. 6. Investigation of cell wall integrity. A, growth assays. Cells were serially 10-fold diluted and spotted on YEPD plates with and without 125 g/ml congo red. Growth was documented after 3 days of incubation at 30°C. B, colony morphology alterations under glucose limitation. Wild type (wt) cells, as well as ⌬asc1, asc1SNR24, and ⌬flo11 mutant cells, were plated on YEPD medium containing 2% (no limitation) or 0.4% glucose (glucose limitation). Colonies were investigated via binocular after 3 days of incubation at 30°C. C, asc1p-dependent adhesive growth under amino acid starvation. Strains RH2817 (wt), RH3263 (⌬asc1), RH3510 (asc1SNR24), RH2681 (⌬flo11), and RH3220 (⌬sfl1; Sfl1p is a transcriptional repressor of FLO11) were patched on solid YNB medium supplemented with 10 mM 3AT. After 3 days of incubation at 30°C, plates were photographed before and after washing. Non-adhesive strains were washed off the agar surface. vated fermentation rates in the ⌬asc1 mutant strain, as more glucose is required for fermentation to obtain the same energy yield as respiration. Elevated ethanol production by the ⌬asc1 strain further supports a shift of energy metabolism from respiration toward fermentation in the mutant cells (Fig. 7A). The specific dependence of the respiratory chain on Asc1p was scored via TTC assay (Fig. 7B). TTC functions as an electron acceptor and is reduced in the course of electron transfer within the respiratory chain, leading to the production of red formazan when colonies are actively respiring (42). Under the given growth conditions, nearly all of the wild type colonies were actively respiring, as indicated by their red pigmentation, whereas colonies without Asc1p remained white. This drastic reduction in respiratory activity illustrates a significant impairment in the respiratory chain as consequence of the deletion of ASC1. To further assess the respiratory capability of the ⌬asc1 strain, a growth assay of wild type and ⌬asc1 cells on agar plates, containing either a fermentable (2% glucose) or a non-fermentable (2% EtOH) carbon source, was conducted (Fig. 7C). The results show that colony growth was rather similar in the wild type and ⌬asc1 strains when glucose was readily available as a carbon source. However, when the cells were forced to respire during growth on EtOH as a non-fermentable carbon source, ⌬asc1 cells showed significantly reduced colony growth. The colonies of both strains further demonstrated altered morphology and reduced colony size, and ⌬asc1 colonies exhibited a significantly stronger phenotype than wild type colonies. ⌬asc1 colony size was halved with the growth on EtOH as a carbon source and was only 40% of the wild type colony size on EtOH. This "petite" phenotype is an additional indicator of restrained respiration in the ⌬asc1 strain (58). The ASC1 mutant strains ⌬asc1 and asc1SNR24 displayed corresponding results for the TTC and growth assays. 7. Evaluation of respiratory and fermentative activity in ⌬asc1 cells. A, comparative analysis of glucose consumption and ethanol production in wild type and ⌬asc1 cells. Wild type and ⌬asc1 cells were grown to OD 600 ϭ 0.5 before the first sample drawing (t ϭ 0). Over a period of 8 h, samples were drawn at the given time points to determine glucose or ethanol concentrations in the medium. Additionally, the cell density was determined. The data were obtained in three biologically independent analyses. The glucose concentration in the medium was measured and the glucose consumption (⌬c) was determined for wild type and ⌬asc1 relative to the cell density (OD 600 ) at the given time points (⌬c (nM glucose)/OD 600 ). ⌬c is defined as the difference in glucose concentration from the first sample drawing (t ϭ 0) to the concentration at the given time points. The ethanol concentration was determined in wild type and ⌬asc1 cell cultures at the given time points and related to the respective cell density (⌬c (nM ethanol)/OD 600 ). The illustrated graph displays the average ethanol production relative to the cell density at t ϭ x (min). B, assay to determine respiratory activity in wild type and ⌬asc1 cells. Colonies were grown on YNB under limited glucose conditions (0.4% glucose) for 3 days and subsequently overlaid with 1.2% agar containing 0.5% triphenyltetrazolium chloride (TTC). Colonies were photographed after 20 min of incubation with TTC at 30°C. C, growth assay to evaluate respiratory colony growth. Cells were spotted in serial 10-fold dilutions on plates containing a fermentable (2% glucose) or non-fermentable (2% EtOH) carbon source. After incubation at 30°C for 3 days, colony diameters were determined via binocular. The results are displayed relative to the wild type in the adjacent graphs.
Taken together, the data demonstrate the requirement of Asc1p for full respiratory chain activity, as respiration is significantly reduced in ⌬asc1 cells. In contrast, fermentation seems to be up-regulated in ⌬asc1 cells, possibly as a secondary effect to compensate for the reduced respiratory activity when Asc1p is absent.
Asc1p Post-transcriptionally Regulates Transcription Factor Abundances-The data presented so far illustrate that Asc1p influences different cellular processes that are controlled by various signaling pathways and, ultimately, by specific transcription factors. Changes in their availability depict an efficient and rapid way of globally adjusting gene expression. In order to identify Asc1p-regulated transcription factors, the downstream effectors causing the changes apparent on the protein-and mRNA-levels, as well as the phenotypical observations noted for the ⌬asc1 strain, were analyzed.
The transcriptome data generated for the ⌬asc1 strain served as a starting point for determining candidate-transcription factors possibly regulated by Asc1p. The entity of genes/mRNAs that were found to be regulated in the ⌬asc1 strain was submitted to the online program YEASTRACT (May 2010) to identify the transcription factors that are known to bind to their promoters (59,60). Respective transcription fac- The list of regulated genes was submitted to YEASTRACT (59,60) to determine the respective transcription factors. The search criteria were set to consider only direct evidence for transcription factor binding. The ten top transcription factors are listed according to the number of genes regulated by them. This number is given as a percentage relative to the total number of genes in the list. The target genes for each transcription factor are separated according to the up-or down-regulation of their respective mRNAs in the ⌬asc1 strain. tors are displayed in Table III. The transcription factors binding most of the genes are Ste12p, Phd1p, Tec1p, and Rap1p, each interacting with over 20% of the genes regulated in the ⌬asc1 strain. Furthermore, these transcription factors all bind to the promoter of FLO11, whose expression cannot be effectively induced in the ⌬asc1 strain (10).
We analyzed whether the protein abundances for these transcription factors are affected by the absence of Asc1p by conducting comparative Western hybridizations in the ASC1 mutant strains ⌬asc1 and asc1SNR24 and in the wild type (Fig. 8A). Additionally, the protein abundance of the transcription factor Flo8p, found at position seven on the ranking list, was analyzed because this factor is known to be crucial for FLO11 expression (61). All five transcription factors showed significant changes in abundance in response to the absence of Asc1p (Fig. 8A). Rap1p was down-regulated by 2-to 3-fold in the asc1SNR24 and ⌬asc1 strains, respectively. Protein levels of Tec1p, Phd1p, and Flo8p were also significantly reduced to 10% to 30% of wild type levels in the absence of Asc1p. Ste12p showed an opposite behavior and was upregulated at rates 6-and 9-fold greater in the ⌬asc1 and asc1SNR24 strains, respectively. The significance of the difference in protein levels between wild type and Asc1pdeficient strains suggests that the analyzed transcription factors are indeed involved in the induction of the majority of proteome and transcriptome changes in the ⌬asc1 strain, FIG. 8. Regulation of transcription factor abundance in wild type and ⌬asc1 strains. A, Western hybridization for transcription factors in the wild type and Asc1p-deficient backgrounds ⌬asc1 and asc1SNR24. Tec1p and Rap1p were detected by protein-specific antibodies. The myc 3 -tagged variants of Ste12p, Phd1p, and Flo8p were detected with anti-myc antibody. Signals were quantified and normalized against eIF2␣ levels detected by anti-eIF2␣ antibody. The adjacent graphs illustrate the fold changes relative to wild type signal intensities. B, 5ЈUTR-mediated translational regulation in wild type and ⌬asc1 cells. Wild type and ⌬asc1 strains were transformed with the testing vector (TV) with or without specified transcription factor 5ЈUTRs. The changes in ␤-galactosidase activity upon the introduction of the respective 5ЈUTRs are illustrated relative to the basal ␤-galactosidase activity measured for the empty TV without 5ЈUTR in the wild type and ⌬asc1 mutant. as well as in its described phenotypical characteristics. Thus, these global changes can be, at least partially, considered as a consequence of the function of Asc1p in regulating the abundances of the mentioned transcription factors.
Interestingly, our transcriptome data (supplemental Table  S3) and quantitative real-time PCR (supplemental Fig. S4 and data not shown) did not indicate an alteration in the mRNA levels of Ste12p, Tec1p, Rap1p, Phd1p, or Flo8p for the ⌬asc1 mutant. This implies that Asc1p does not alter the abundance of these transcription factors through transcription, but rather influences these proteins posttranscriptionally.
Asc1p Controls the Translation of the Transcription Factor mRNAs of Ste12p, Phd1p, and Tec1p through Their Respective mRNA-5ЈUTRs-According to our data, the transcription of STE12, PHD1, TEC1, RAP1, or FLO8 genes themselves is not affected in ⌬asc1 cells, whereas the determined protein abundances of the corresponding transcription factors changed significantly from wild type levels in the absence of Asc1p (Fig. 8A), indicating their post-transcriptional regulation. Given the involvement of RACK1/Asc1p in mRNA translation, and specifically in the initiation process (9, 62), we assessed the influence of the respective mRNA-5ЈUTRs of these candidates on translation initiation in wild type and ⌬asc1 background via lacZ-reporter assay (35). The basal lacZ-expression for the wild type and ⌬asc1 strain is expressed by the ␤-galactosidase activity measured for the empty testing vector without any 5ЈUTR-insertion (Fig. 8B). After the introduction of the STE12-5ЈUTR, the ␤-galactosidase activity in the ⌬asc1 mutant strain was twice as high as that in the wild type background. A significantly reduced expression in the ⌬asc1 strain could be observed for the 5ЈUTRs of the PHD1-and TEC1-mRNA. For the PHD1-5ЈUTR, the ␤-galactosidase activity stayed at basal levels in the ⌬asc1 mutant strain, whereas it nearly doubled in the wild type. The insertion of the TEC1-5ЈUTR led to the highest induction of activities in both strains, whereas ␤-galactosidase activities in the ⌬asc1 strain amounted to only 50% of wild type levels.
For the mRNAs of STE12, PHD1, and TEC1, this strongly suggests Asc1p-dependent regulation during the initiation of translation that is mediated by the 5ЈUTR sequences. Further, the changes in ␤-galactosidase activity from the wild type to the ⌬asc1 strain correlate with the observed changes in protein abundance for these candidates. These data support Asc1p's role as a ribosomal regulator in the translation of STE12, PHD1, and TEC1.

DISCUSSION
Eukaryotic RACK1/Asc1p has strong effects on cellular signaling and thereby defines the developmental fate of a cell. Uncontrolled cell growth and cancer formation in metazoa have been connected to distorted RACK1 expression, desig-nating it as a biomarker for breast carcinoma (26,27,63). As its deletion in higher eukaryotes, but not in yeast, is lethal at early developmental stages, the unicellular model organism S. cerevisiae was utilized here to analyze the consequences of a total ASC1 deletion on a genome-wide scale. Asc1p was found to affect diverse cellular processes, including translation, iron homeostasis, pheromone response, cell wall integrity, and energy metabolism (glycolysis, respiration, fermentation, oxidative stress). In addition to the ASC1 gene, the ASC1 locus encodes for the snoRNA U24. An ASC1 mutant strain, asc1SNR24, expressing the snoRNA U24 but not Asc1p, was generated to determine Asc1p specific effects (supplemental Fig. S2). Our data show that the phenotypes and the changes in transcription factor expression observed in the ⌬asc1 strain can be attributed to the absence of Asc1p.
Our analysis suggests a distinct function of Asc1p as a central ribosomal node, taking in a variety of different signals (e.g. from the Kss1p-and Pkc1p-MAPK pathways) and thereupon enabling the respective cellular reprogramming. Our detailed analyses link the Asc1p-mediated control to the initiation of translation of the transcription factors responsible for the majority of transcriptional alterations observed in the ⌬asc1 strain. Asc1p regulates translation initiation of the mRNAs coding for the transcription factors Ste12p, Tec1p, and Phd1p through their respective 5ЈUTR sequences. This finding might stem from a direct interaction of Asc1p with factors involved in translation initiation as demonstrated recently for the translation initiation factor eIF3 (62). The deletion of ASC1 results in a reduced eIF3-binding affinity toward the 40S subunit, thereby inhibiting translation initiation.
One central consequence of the deletion of ASC1 is the down-regulation of mRNAs coding for siderophore transporters and the high-affinity iron-uptake complex Fet3p/Ftr1p (Fig. 9), leading to an imbalance in iron homeostasis in ⌬asc1 cells. The resulting iron limitation would severely compromise biosynthetic processes of, for example, iron-sulfur-clustercontaining proteins and heme in ⌬asc1 cells (53). In support of this, a negative genetic interaction of ASC1 with HEM1 and ISU2, encoding proteins involved in heme biosynthesis and the synthesis of mitochondrial and cytosolic Fe-S proteins, respectively, has previously been established (7). Fe-S proteins are directly involved in the electron transport chain at the inner mitochondrial membrane, whereas heme is critical in supplying oxygen to mitochondria, where it serves as a final electron acceptor in the respiratory chain (Fig. 9). Heme deficiency results in hypoxic (low-oxygen) cellular conditions. Human RACK1 is implicated in the degradation of the hypoxiainduced transcription factor HIF-1 (64), the master regulator of hypoxic genes. HIF-1 further controls the gene expression of key players in angiogenesis, erythropoiesis, and glucose/energy metabolism (65). Yeast cells grown under oxygen limitation down-regulate Asc1p (66). This indicates that Asc1p is specifically required during aerobic respiration, as demonstrated in this study via TTC assay and the impaired growth of ⌬asc1 cells on EtOH as a non-fermentable carbon source. The requirement of Asc1p for respiration is further reflected by the severely reduced expression of MDH1, encoding the tricarboxylic acid cycle malate dehydrogenase, and RIB4, required for riboflavin synthesis in ⌬asc1 cells (Fig. 9). Electron transfer across the inner mitochondrial membrane will most likely be impaired by the prevailing compromised respiration in ⌬asc1 cells. In agreement, ATP1 and ATP2, coding for components of the F 1 F 2 -ATPase, are down-regulated in the deletion strain, suggesting reduced respiratory ATP production (67). Additionally, ASC1 shows a negative genetic interaction with MDJ2, involved in stimulating ATPase activity (7).
In the course of respiration, electrons can leak along the way through the inner mitochondrial membrane, leading to the production of ROS (48). Hypoxia and inhibition of the respiratory chain further increase the production of ROS such as superoxide anions and lead to the formation of RNOS (68). In ⌬asc1 cells, these conditions are indicated by the strong up-regulation of Hbn1p and Yhb1p in the mutant strain on the mRNA level and the protein level. Furthermore, agar diffusion assays with the NO-donor DETA-NONOate indicate that nitrosative stress is inflicted by the deletion of ASC1 and that the up-regulation of Hbn1p and Yhb1p in the ⌬asc1 mutant strain is required for protection against nitrosative stress. Hbn1p has previously been reported to be up-regulated upon iron deficiency (69) in order to activate superoxide dismutase (SOD) (49). Yhb1 protein levels are generally low in respiring yeast cells but strongly increase under hypoxic conditions, when the electron transfer chain is compromised and oxidative and nitrosative stress is enhanced (70 -72). Accordingly, Yhb1p functions indirectly in oxidative stress response as NO reductase (46) (Fig. 9). Yhb1p consumes NO, thereby diminishing NO's inhibitory effect on mitochondrial respiration through the modification of cytochrome c oxidase (73,74) and the consequently enhanced ROS production (47) (Fig. 9). Despite the lack of NO synthases, yeast cells produce NO by passing electrons from the cytochrome c oxidase to NO 2 Ϫ as an alternative final electron acceptor when cellular oxygen levels are reduced (72). NO and the superoxide anion (O 2 Ϫ ) combine and form peroxinitrite (ONOO 2 ), which in turn can nitrate specific protein tyrosine residues (72, 75) (Fig. 9). The experimental data in this study suggest the presence of an excessive amount of ROS, as well as of NO, in ⌬asc1 cells, consequently suggesting elevated ONOO 2 levels and tyrosine nitration. In mammalian cells, this modification specifically activates PKC through its increased interaction with RACK FIG. 9. Cellular context of processes affected by Asc1p. Expression data and phenotypical observations are incorporated according to the cellular localization of the corresponding protein or process, respectively. Regulations supported by direct evidence are illustrated in red for a down-regulation in the ⌬asc1 strain and green for an up-regulation. (76). Increased protein tyrosine nitration has been reported as a consequence of hypoxic growth conditions (72) and might represent a very effective mechanism by which to influence protein activities upon ROS-and NO-inducing conditions, especially in the mitochondria. A proteome study for tyrosine nitrated proteins in rat revealed nitration targets especially within the functional groups of oxidative stress, energy production, and fatty acid metabolism, as well as the Asc1p homologue RACK1 (77). Probably closely associated with cellular nitration is the function of Hbn1p through its NAD(P)Hdependent activity as putative nitroreductase and its involvement in SOD activation (49,78).
Although the nitration of yeast Pkc1p or Asc1p has not yet been shown, the Pkc1-dependent maintenance of cell wall integrity is significantly compromised in the ⌬asc1 strain. This results in increased sensitivity to the cell-wall-disturbing reagents calcofluor white, zymolyase (10), and congo red. Additionally, previous studies have identified negative genetic interactions of ASC1 with several genes, both essential and non-essential for cell wall synthesis (5,7,79,80). The misregulation of multiple mRNAs and proteins involved in cell wall biogenesis and morphology, as well as the reduced de novo biosynthesis of radioactively labeled proteins of the insoluble fraction in the ⌬asc1 strain, corroborates this result (Fig. 9). Despite these unambiguous phenotypes, several findings indicate that the corresponding cell wall integrity signaling pathway is in fact activated (supplemental Fig. S3). The mRNA for WSC3, coding for the sensor-transducer of the pathway, is up-regulated in the mutant strain. Additionally, the MAP kinase of the cell wall integrity pathway, Slt2p, physically interacts with Asc1p (6) and, more important, shows elevated phosphorylation when ASC1 is deleted (81). This discrepancy between the activation of the cell wall integrity pathway and the obvious deficiencies in cell wall integrity suggests that the MAPK pathway of cell wall integrity might be influenced downstream of its MAP kinase Slt2p in the ⌬asc1 mutant. Consistent with this, in our analysis, the bulk of target genes of the cell wall integrity pathway were not up-regulated and even showed down-regulation despite pathway activation when ASC1 was deleted ( Fig. 3 and Table III). These data indicate that Asc1p might influence the Slt2p targeted transcription factors Rlm1p, Skn7p, and Swi4p and thereby cause the observed transcriptional changes, which is a question that needs to be addressed in the future.
A similar discrepancy, supporting a specific requirement of Asc1p for the appropriate synthesis of transcription factors, is observed in the ⌬asc1 strain for the signaling cascade of adhesive and pseudohyphal growth, divided between the cAMP-dependent PKA pathway and the Kss1p-MAPK pathway (supplemental Fig. S3). Both pathways are constitutively activated when ASC1 is absent, as monitored via elevated cAMP-levels and the enhanced phosphorylation of the MAP kinase Kss1p (17). The activation of the PKA pathway in wild type cells leads to the phosphorylation of the transcription factor Flo8p, inducing FLO11 and other target genes (61). Activated Kss1p mediates the transcriptional activity of the transcription factors Ste12p and Tec1p on the FLO11 promotor. In the course of activation, it phosphorylates Ste12p and its inhibitors Dig1p/Dig2p, leading to the release and activation of Ste12p (82,83). Despite the activated FLO11-inducing pathways, FLO11 expression is shut down in ⌬asc1 cells, and Flo11p-dependent adhesive/pseudohyphal growth and mat formation cannot be induced (10,57) (Figs. 6B and 6C). This again suggests a direct effect of Asc1p on the transcription factors. Consistently, the abundances of Ste12p, Tec1p, and Flo8p and the PKA and MAPK pathway-independent FLO11promoter-binding transcription factors Phd1p and Rap1p are in part drastically altered when Asc1p is absent. Whereas the protein levels of Tec1p, Flo8p, Phd1p, and Rap1p are significantly reduced, Ste12p is more abundant in ASC1 mutant cells. In compliance, the transcriptome data reveal a downregulation for the majority of target genes regulated by Tec1p, Flo8p, Phd1p, and Rap1p in the ⌬asc1 strain, whereas Ste12p target genes are mostly up-regulated (Table III).
A function of Asc1p in translational regulation is strongly supported by our analysis of specific transcription factor-5ЈUTR sequences via reporter system in this study. Our data show that the changes in Tec1p, Phd1p, and Ste12p abundances depend on Asc1p. Their protein levels, as determined by Western experiments, and 5ЈUTR activities in the testing system significantly correlate. This strongly supports a specific function of Asc1p in the control of the initiation of translation of specific transcription-factor-encoding mRNAs, thereby regulating the cellular response via affiliated signaling pathways. Additionally, Asc1p might influence the stability of the transcription factors. In order to evaluate the ASC1-dependent transcription factor degradation rates, shut-off Western experiments were performed for those transcription factors that can be deleted and overexpressed (Tec1p, Phd1p, and Flo8p). The differences in transcription-factor-degradation rates observed in Western shut-off experiments are not sufficient to explain the differing steady-state levels of the transcription factors in the wild type and ⌬asc1 backgrounds under their native conditions and with the GAL1 promoter (supplemental Fig. S4). However, we consider the possibility of an Asc1p-mediated coupling of transcription factor translation with native polypeptide-chain folding and premature protein degradation.
In conclusion, our study provides evidence that Asc1p is enmeshed in the regulation of diverse cellular processes, ranging from iron homeostasis and energy metabolism to intracellular signaling through MAPK cascades and the cAMP-dependent PKA pathway. Asc1p is involved in balancing different energy-producing strategies according to oxygen-and carbon-source availability. This supports the requirement of Asc1p/RACK1 to regulate the normoxic-hypoxic transition, a process of significance in cellular aging, neurodegeneration, and cancer in mammalian cells (84 -86).
Asc1p physically interacts with several components of signal transduction pathways that, when activated, might lead to changes in Asc1p modification and/or conformation at the ribosome, thereby transferring their message into the translational machinery (6,17). Some of these possibilities are currently being explored in our laboratory, but it will take further extensive research to identify the exact mechanisms by which Asc1p confers signal-pathway-sensitive translational regulation. On the basis of the discrepancies between the activation of signaling pathways and the transcriptionfactor-mediated phenotypical read-out, we hypothesize that the interactions of the ribosomal protein Asc1p with upstream signal transduction proteins specifically alters ribosomal settings and thereby couples signal transduction events with the translational process (supplemental Fig. S5). We imagine that as a result of the deletion of ASC1, signal transduction and translation become at least partially uncoupled, and, as we were able to show for the process of adhesive/pseudohyphal growth, the pathway is turned on but the expected cellular response is not achieved, a discrepancy previously observed (17). Another important aspect supporting this idea is that Asc1p indeed influences the modification and abundance of translation initiation and elongation factors (10; unpublished data). This, in combination with regulatory element-harboring 5ЈUTR sequences, could determine the translational rates of specific transcription factor mRNAs or confer ribosome specificity for distinct groups of mRNAs.