Correspondence: Germán Larriba, Departamento de Microbiología, Facultad de Ciencias, Universidad de Extremadura, 06071 Badajoz, Spain. Tel./fax: +34 924289428; e-mail: firstname.lastname@example.org
ybl006cΔ was selected using an Alcian-blue-based screen aimed to identify nonessential genes involved in the regulation of mannosylphosphorylation. When cells of this deletant were mixed with the cationic dye Alcian blue in a typical assay, they remained white, indicating a low number of mannosylphosphate groups on the cell surface. ybl006cΔ cells did not show any defect in growth rate nor in the glycosylation or secretion rate of the major exoglucanase Exg1. Transcriptome analysis of ybl006cΔ using macroarrays showed at least two-fold changes in the expression of 52 genes (<0.9% of the genome). Three of these have previously been reported to be directly (MNN6 and MNN4) or indirectly (MNN1) implicated in the transfer of mannosylphosphate to the N- and O-oligosaccharides. Alterations in the expression of these genes were confirmed by Northern analysis. YBL006C product was recently identified as a subunit (Rsc14) of the RSC chromatin-remodelling complex of Saccharomyces cerevisiae. It follows that remodelling of chromatin can be an important regulatory mechanism for the maturation of cell wall mannoproteins.
Mannoproteins are major components of the cell wall of a number of yeasts, including Saccharomyces cerevisiae, Candida albicans, Schizosaccharomyces pombe, and Kluyveromyces lactis. In S. cerevisiae, for which they have been extensively studied, mannoproteins account for about half of the dry weight of the cell wall. Among other roles, mannoproteins contribute to the regulation of the cell wall porosity, and therefore control both the exit of secretory proteins and the entrance of macromolecules from the environment. In fact, most mannoproteins are first secreted into the periplasmic space, and from there some of them are incorporated into the cell wall whereas others can diffuse into the culture medium, where they accumulate (reviewed in Klis et al., 2006; Lesage & Bussey, 2006). A few species of mannoproteins, which have been characterized as proteases, are secreted into the vacuole (Vida et al., 1991; Van Den Hazel et al., 1996).
The cell wall mannoproteins of S. cerevisiae are highly glycosylated polypeptides; carbohydrate often contributes to 50–95% of their molecular mass (Orlean, 1997). During intracellular transport, these proteins acquire N- and O-linked oligosaccharides, and some of them can also receive a GPI anchor at their carboxyl terminus that will link the protein to β-1,6-glucan (Kapteyn et al., 1996). N-oligosaccharides are composed of a conserved inner core (GlcNAc2Man8) synthesized in the endoplasmic reticulum (ER) and a variable branched outer chain of up to 200 mannose units added in the Golgi apparatus (Ballou, 1990), whereas O-oligosaccharides consist of one to five mannoses (Strahl-Bolsinger et al., 1999). In wild-type cells, N- and O-oligosaccharides contain mannosylphosphate residues that confer a net negative charge to the cell wall. Within the N-oligosaccharides, mannosylphosphorylation can occur at four positions, namely two sites in the inner core and two sites in the outer chain (nonreducing end and branches of the mannose outer chain), whereas O-oligosaccharides carry mannosylphosphate only in a single position (Ballou, 1990; Nakayama et al., 1998; Jigami & Odani, 1999). This pattern is not universal, however, but varies from yeast to yeast. It is speculated that the several variants serve to adapt each species to its environmental niche, but for each species of yeast the precise significance of these acidic residues is still under discussion.
To date, only two genes, MNN4 and MNN6, have been shown to be directly involved in the addition of mannosylphosphate residues. mnn4 and mnn6 mutants were initially isolated on the basis of their deficiency in binding the cationic dye Alcian blue. Further analysis of the carbohydrate of their mannoproteins showed an almost total absence of phosphate groups (Ballou, 1990). Mnn6 is a mannosylphosphate transferase that seems to be implicated in the mannosylphosphorylation of the O-linked glycans and the outer chain, but not of the inner core, of the N-oligosaccharides (Wang et al., 1997; Nakayama et al., 1998). The function of Mnn4 remains unclear. This protein shares a conserved signature sequence with the fukutin family of proteins, an observation that would support the notion that Mnn4 is a mannosylphosphate transferase (Aravind & Koonin, 1999). However, the dominant nature of the mnn4 mutation and the failure to demonstrate mannosyltransferase activity in Mnn4 in several in vitro assays point to a regulatory role for this protein (Odani et al., 1996, 1997)
In a screen based on alterations in the binding of the cationic dye Alcian blue carried out over 622 deletion strains during the EUROFAN project we isolated 48 strains with abnormal amounts of phosphate in the surface of the cell (Conde et al., 2003). The mutant ybl006cΔ exhibited one of the most dramatic decreases in the binding of Alcian blue as well as in the binding of QAE-beads, suggesting a very low amount of phosphate. YBL006C was later reported to complement the ldb7 mutant (Mañas et al., 1997; Corbacho et al., 2004).
For the present report we characterized strain ybl006cΔ further, and undertook an analysis of its transcriptome. We show that the decrease in the amount of phosphate is not the result of defects in the size of the outer chain of N-oligosaccharides but may be explained by an alteration in the transcription rates of three genes involved in mannosylphosphorylation: MNN1, MNN6 and MNN4. While this work was in progress, YBL006C was reported as RSC14, which encodes a subunit of chromatin-remodelling complex (RSC) (Wilson et al., 2006). The connection of the altered expression of these ORFs to chromatin remodelling is therefore discussed.
Materials and methods
Yeast strain and growth conditions
Wild-type Saccharomyces cerevisiae BY4741 (MAT a; his3Δ1, leu2Δ0, met15Δ0, ura3Δ0) and its derivatives ybl006cΔ (YBL006C::kanMX4), yer001wΔ (YER001W::kanMX4), ykl201cΔ (YKL201C::kanMX4) and ypl053cΔ (YPL053C::kanMX4) were obtained from the EUROSCARF collection (http://web.uni-frankfurt.de/fb15/mikro/euroscarf/).
Cells were routinely grown at 28°C in YEPD medium (2% glucose, 1% yeast extract, 2% bactopeptone) or synthetic dextrose (SD) medium (0.7% yeast nitrogen base without amino acids, 2% glucose) supplemented with the appropriate nutrients in amounts specified in (Sherman, 1991). YEPD and SD media were solidified with 2% agar. For the production of external exoglucanase, cells were grown in liquid minimal medium supplemented with the appropriate amino acids plus uracil (Cueva et al., 1996) until the middle of the exponential phase of growth.
Determination of exoglucanase activity, and purification and characterization of Exg1p
Secreted and periplasmic space-associated exoglucanase activities were determined as previously described (Hernández et al., 1986). For purification of the various glycoforms of exoglucanse, culture supernatants obtained by centrifugation of cells were concentrated and dialyzed using Amicon PM10 membranes and/or Centricon filters. Samples were fractionated by ion-exchange chromatography (TSK gel DEAE-5PW, 7.5 mm × 7.5 cm, TosoHaas). Exoglucanase activity was determined using p-nitrophenol-β-d-glucoside as a substrate (Cueva et al., 1998). Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and Western blots using rabbit polyclonal antibodies anti-Exg were performed as reported in (Conde et al., 2004). Antiserum binding was visualized using SuperSignal West Pico Chemiluminescent Substrate (PIERCE) following the manufacturer's instructions.
Alcian blue is a cationic stain that binds to cells carrying a negative charge on the cell surface (Friis & Ottolenghi, 1970). We used the method described by Ballou to measure binding (Ballou, 1990). A 0.1% stain solution was prepared in 0.02 N HCl, and the suspension was centrifuged to eliminate insoluble precipitates. A small number of cells grown for 48 h on solid YEPD was placed in a disposable glass test tube. Cells were washed with 0.5 mL of 0.9% NaCl and then resuspended in 0.2 mL of an Alcian-blue solution (0.1% Alcian blue in 0.02 N HCl). The mixture was vortexed until the cells and stain were mixed, and the suspension was then maintained in a static condition at room temperature for 10–15 min and centrifuged at 4000 g. The pellet was washed twice with 2 mL of 0.02 N HCl and the last supernatant was retained above the undisturbed pellet. This allowed us to compare aligned samples in a test-tube rack by looking at the bottom of the tubes.
Macroarray hybridization and data analysis
Total RNA from logarithmically growing ybl006cΔ mutant and isogenic wild-type cultures was isolated as described (Sherman et al., 1986) but using a Fast-Prep (BIO101 Inc.) device to break the cells. RNA was incubated with DNase to remove any traces of DNA and then analysed electrophoretically. Thirty micrograms of RNA was retrotranscribed into cDNA using 33P-dCTP as described (Alberola et al., 2004). Nylon-membrane macroarrays containing PCR products representing full-length ORFs for 6049 genes of S. cerevisiae (Belli et al., 2004) were prepared at the Servicio de Chips de DNA of the University of Valencia (Spain). Prehybridization was carried out at 65°C for 1 h in 5 mL of 5 × SSC, 5 × Denhart's solution, 0.5% SDS. The prehybridization solution was then replaced with 5 mL of the same solution, containing 33P-labelled cDNA (3 × 106 dpm mL−1). Hybridization was performed at the same temperature for 18 h. Washing steps were as follows: once at 65°C for 20 min in 2 × SSC, 0.1% SDS; and twice at 65°C for 30 min in 0.2 × SSC, 0.1% SDS. Filters were exposed to a BAS-MP screen (Fujifilm) for 24 h. Images were photographed using a FujiFilm FLA3000 Phosphorimager. Spot intensities were measured using array vision 7.0 software (Imaging Research Inc.), considering as valid those signals 1.45 times over background. The experiment was repeated three times with each of the strains. Each mutant replicate was normalized against the wild type hybridized on the same membrane. Reproducibility of the replicates was tested using arraystat software (Imaging Research, Inc.), considering the data as independent and allowing the program to take a minimum number of valid replicates of two in order to calculate mean values for every gene. Application of a Z-test for independent data provided the differences in individual gene expression between the two strains. A P-value of 0.05 and the false discovery rate method were used to monitor the overall false-positive error rate. Only those genes showing a differential expression of at least twofold as compared with the wild type were considered.
Total RNA was prepared as described above, separated in formaldehyde-agarose gels (12 μg lane−1), and transferred to nylon membranes (BiodyneR A 0.45 μ Pall Corporation). DNA probes were labelled by random primed incorporation of 32P dCTP using a Random Primed DNA Labeling Kit (Roche) according to the manufacturer's instructions. Hybridization was carried out in 0.2% SDS, 0.4 M NaCl, 0.08 M sodium phosphate pH 6.5, 4 mM EDTA pH 8.1, 10% dextran sulphate at 65°C for 24 h. The filters were washed at room temperature twice for 10 min in 2 × SSC, twice for 45 min in 2 × SSC, 1% SDS, and finally once for 20 min in 0.1 × SSC. Signal quantification was performed in a Phosphorimager Molecular Imager Fx (BioRad). The following probes were used: a 450-bp XbaI/ClaI fragment internal to the MNN6 ORF, a 520-bp EcoRI fragment internal to MNN4, a 500-bp fragment of the coding region of MNN1 obtained by PCR (forward oligonucleotide: 5′-AGAAATCCGAGAGATTGTATCGC-3′; reverse oligonucleotide: 5′-GACAGCTAATGGAAATTGGTTGG-3′), and a 1700-bp HindIII/BamHI fragment internal to ACT1.
ybl006cΔ is defective in phosphomannosyl content
ybl006cΔ was selected in our laboratory in a screen to identify nonessential genes that regulate the mannosylphosphate content of the cell wall mannoproteins using the Alcian-blue test (Conde et al., 2003). As shown in Fig. 1, ybl006cΔ cells did not bind the dye, appearing as white as their counterparts from mnn6Δ, whereas wild-type cells bound the dye and turned deep blue in the same assay. As mentioned above, MNN6 encodes a phosphomannosyl transferase, and therefore its deletion results in cells with very low levels of mannosylphosphate that do not bind Alcian blue. The defect of ybl006cΔ was the result of the deletion rather than of secondary mutations, as it was complemented with its cognate gene in a centromeric vector (Fig. 1).
ybl006cΔ cells carry out normal glycosylation and secretion of the major exoglucanase (Exg1) of Saccharomyces cerevisiae
The decrease in the net negative charge of the cell wall of ybl006cΔ cells could derive from general defects in glycosylation that result in a loss of acceptor sites for the addition of Man-P. To address this issue, we analysed the nature of the Exg1 glycoforms secreted by the mutant. Concentrated and dialysed culture supernatants were fractionated by ion-exchange chromatography (HPLC). As shown in Fig. 2a, the elution profile of the exoglucanase activity secreted by the mutant is identical to that of the wild type, with a heterogeneous minor peak (Exg1a 10%) and a sharp major one (Exg1b 90%) (Larriba et al., 1995). It should be noted that the elution profile of the Exg1 glycoforms in this system depends on the number and length of N-oligosaccharides (Basco et al., 1994). By contrast, the profile is independent of the presence of an active Mnn6p, because a mnn6 mutant yielded the same profile as the wild type (not shown) in spite of the fact that 77% of the Exg1b molecules were phosphorylated on one (34%) or both (43%) positions of the inner core (Hernández et al., 1992). Likewise, Western blot analysis (Fig. 2b and c) revealed that the sizes of both hyperglycosylated and glycosylated glycoforms were the same in the wild-type and ybl006cΔ strains. These results demonstrate that neither inner core glycosylation nor the elongation of the outer chain was impaired in ybl006cΔ, and accordingly the low amount of mannosylphosphate is unrelated to the number of acceptor sites, as reported for the mnn9 mutant. On the other hand, we found no differences in the total amount of secreted exoglucanase activity in the culture medium and periplasmic space between exponentially growing ybl006cΔ and wild-type cells, suggesting that the global secretion rate is unaffected by the mutation.
Analysis of the transcriptome of ybl006cΔ
In order to detect any transcriptional alterations in ybl006cΔ that could explain the deficiency in mannosylphosphate, we assessed the overall pattern of gene expression using cDNA macroarray technology. The transcription of 52 genes was affected at least two-fold; of those, 33 were expressed at higher levels (Table 1) and 19 were expressed at lower levels (Table 2). Genes were grouped into functional categories according to MIPS (http://mips.gsf.de/) and the Saccharomyces Genome Database (http://www.yeastgenome.org/).
Table 1. Up-regulated genes
Biogenesis of cellular components
rRNA gene silencing and nucleolar integrity
Mannosidase activity. Cell wall biogenesis
Cell cycle and DNA processing
Meiosis-specific serine/threonine protein kinase
Protein phosphatase S phase
Conjugation. Pheromone receptor
Metabolism and energy
Cytochrome c oxidase. Mitochondrial electron transport
Alcohol dehydrogenase II
Ribosome biogenesis and assembly
Translation elongation factor
Intranuclear transport of ribosomal precursors
Ribosome biogenesis and assembly
HSP70 family. Nuclear transport
HSP70 family. SRP-dependent cotranslational protein-membrane targetting and translocation;
Guanosine difosfatase. Glycosylation
HSP70 family. SRP-dependent cotranslational protein-membrane targetting and translocation
Retrograde transport, endosome to Golgi
RNA helicase. Prespliceosome formation
Nucleolar protein. rRNA processing
Nucleolar protein. rRNA processing
Transcription factor required for oxidative stress tolerance, drug and metal resistance
Table 2. Down-regulated genes
Biogenesis of cellular components
Spore wall maturation
Structural constituent of cell wall
a-agglutinin adhesion subunit
Aspartyl protease cleaves and inactivates alpha factor
Metabolism and energy
Protein kinase. rRNA processing
ER to Golgi vesicle-mediated transport
Putative regulator of Mnn6
Mitochondrial translational activator
Nuclear mRNA splicing
Nuclear mRNA splicing
Analysis of the data indicates that the genes altered are functionally heterogeneous, but that three ORFs are specifically related to the glycosylation and/or mannosylphosphorylation processes, namely YER001W (MNN1), YKL201C (MNN4) and YPL053C (MNN6). MNN1 is up-regulated (5.18-fold change), and MNN4 and MNN6 are down-regulated (0.13- and 0.44-fold change, respectively) (Fig. 3b). To confirm the macroarray data further, we performed Northern blot analyses. As shown in Fig. 3a, a visual inspection of the blots indicated that MNN1 was up-regulated, whereas MNN4 and MNN6 were down-regulated. Furthermore, quantification of the intensity of the bands indicated 3.91-, 0.26- and 0.38-fold changes for MNN1, MNN4, and MNN6, respectively, which are close to those obtained in the macroarrays (Fig. 3b). With the exception of YEL042W, which is involved in the transport of GDP-mannose into the Golgi lumen, none of the other genes selected in the transcriptome seems to be related to the mannosylphosphate content of the cell wall, as they were not identified in previous screens (Conde et al., 2003; Corbacho et al., 2005).
In a previous work we identified YBL006C as an ORF involved in the regulation of mannosylphosphorylation in S. cerevisiae. In the present study we have investigated the several alterations that could account for the drastic drop in the amount of mannosylphosphate in the cell wall of the mutant. There is no doubt that the white phenotype shown by the deletant in the Alcian-blue test resulted from the deletion, because the defect could be complemented by the corresponding cognate gene in a centromeric plasmid.
An attractive possibility is that this phenotype is the result of a decrease in the length of the N- and O-oligosaccharides, and therefore in the number of acceptor sites for the addition of mannosyl-P residues. As noted, the outer chain of each N-oligosaccharide carries several potential acceptor sites for mannosyl-P, whereas the inner core has only two (Ballou, 1990). For instance, the pale blue colour shown by the mnn9 mutant in the Alcian-blue test is thought to be caused by the absence of the outer chain. However, no changes in the number of N-residues attached to the exoglucanase secreted into the culture medium or in the average size of the outer chain of these N-oligosaccharides were observed. In addition, no alteration in the secretion rate of exoglucanase was detected, suggesting that mutant cells have a normal complement of cell wall mannoproteins.
Genome-wide analysis of ybl006cΔ using macroarrays indicated that only 52 genes (<0.9% of the genome) were altered by a factor of two or more. Of these, 19 were down-regulated and 33 were up-regulated. Interestingly, we found altered expression of three genes whose translation products determine to a great extent the levels of mannosylphosphate. MNN6, which encodes the only well-characterized phosphomannosyl transferase, was down-regulated about three-fold. MNN4, which appears to have a direct positive regulatory role in mannosylphosphorylation, was one of the genes whose expression was more drastically reduced (almost ten-fold in transcriptome and fourfold in Northern analysis). Finally, MNN1, which encodes an α-1,3 mannosyltransferase, was one of the genes more significantly up-regulated (fivefold). It has been suggested that the enzymes Mnn1 and Mnn6 both compete for the addition of their respective residues (Jigami & Odani, 1999). In this scenario, an increase in Mnn1 and a concomitant decrease in Mnn6 would lead to an increase of the α-1,3-mannose residues that decorate the N-oligosaccharides to the detriment of the number of mannosylphosphate residues, especially when a stimulator of the addition of the latter, such as Mnn4, is also significantly diluted.
During the course of this study, it was reported that the product of YBL006C is a component (Rsc14) of the RSC chromatin-remodelling complex, where it interacts with Rsc7, Htl1, Rsc3 and Rsc30 to form a fungal-specific module (Wilson et al., 2006). RSC is an essential member of the SWI/SNF family of ATP-dependent chromatin-remodelling complexes (Cairns et al., 1996; Martens & Winston, 2003; Mohrmann & Verrijzer, 2005). Genome-wide location analysis has suggested that the RSC complex is generally recruited to RNA polymerase III promoters and specifically recruited to RNA polymerase II promoters by transcriptional activators and repressors (Ng et al., 2002). Interestingly, data indicate a genetic link between Rsc3, an essential RSC subunit, and the cell wall integrity pathway (Angus-Hill et al., 2001). Similarly, rsc14Δ exhibits phenotypes typical of some mutants in cell wall components, such as sensitivity to 1.2 M NaCl and 15 mM caffeine at 30°C, and to 2% formamide and 800 ng mL−1 congo red at elevated temperatures (38°C)(Wilson et al., 2006). It is unlikely that these phenotypes are a consequence of the defects in mannosylphosphorylation, as they have not been described for the specific mutants mnn4 and mnn6.
Despite the small number of up- and down-regulated genes provided by the transcriptome, we could not find a rational explanation for their alteration. A literature survey failed to show other mutants that share a similar expression pattern. In addition, the genes altered either lack known regulation or are regulated by different transcription factors. Furthermore, in spite of the strong interaction between Rsc14 and Rsc3, our transcriptome results do not show any overlap with those obtained by Angus-Hill (Angus-Hill et al., 2001) for rsc3 and rsc30Δ strains, which were also different between them. It therefore appears that the different components of the RSC complex influence the expression of different genes. Interestingly, the algorithm blast identifies Rsc14 orthologues in closely related hemiascomycete yeasts, including Candida glabrata, Eremothecium gossypii, and Kluyveromyces lactis (blastP-values of 2.4 × e−25, 1.9 × e−20 and 9.3 × e−12 respectively), but not in other species. A likely possibility is that Rsc14 is a recent acquisition of the RSC complex in S. cerevisiae and other related lineages, where it has evolved to serve a role in the maintenance of a net negative charge, contributing in that way to the fitness or survival of the species.
This study was supported by grants BIO2003-09180-C02-02 from CICYT (Spain) and BIO4CT 97-2294 from the European Union to G.L. R.C. received a fellowship from Junta de Extremadura.