Systematic identification in silico of covalently bound cell wall proteins and analysis of protein–polysaccharide linkages of the human pathogen Candida glabrata

Michael Weig1, Lothar Jänsch2, Uwe Groß1, Chris G. De Koster4, Frans M. Klis3 and Piet W. J. De Groot3,4

1 Institute of Medical Microbiology and National Reference Center for Systemic Mycoses, University of Göttingen, Germany
2 GBF, Braunschweig, Germany
3 SILS-Microbiology, University of Amsterdam, The Netherlands
4 SILS-Biomacromolecular Mass Spectrometry, University of Amsterdam, The Netherlands

Correspondence
Piet W. J. de Groot
pgroot{at}science.uva.nl


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Candida glabrata is an important cause of systemic candidiasis in humans. This paper reports a systematic analysis of the putative glycosylphosphatidylinositol-modified (GPI) proteins of C. glabrata, a large part of which are covalently bound to the cell wall glucan network and the remainder of which are retained in the plasma membrane, and of cell wall proteins (CWPs) which are covalently bound in a mild-alkali-sensitive manner. In silico genomic analysis revealed 106 putative GPI proteins. Fifty-one of these GPI proteins could be categorized as adhesive proteins, potentially implicated in fungus–host interactions or biofilm formation during the development of fungal infections. Eleven proteins belonged to well-known GPI protein families of glycoside hydrolases, probably involved in cell wall expansion and remodelling during growth. Other identified GPI proteins included phospholipases, aspartic proteases, homologues of ScEcm33p and ScKre1p, and structural CWPs. Interestingly, the GPI algorithm predicted three orthologues of an abundant CWP in S. cerevisiae, Cwp1p, which is absent in Candida albicans. To evaluate the in silico predictions, isolated cell walls were extracted using HF-pyridine, which specifically cleaves phosphodiester bonds, to release GPI-CWPs. Immunological analysis of the extract using one-dimensional SDS-PAGE and anti-ScCwp1p antiserum indicated the presence of a Cwp1p homologue in C. glabrata cell walls. Further analysis by two-dimensional gel electrophoresis and electrospray ionization tandem mass spectrometry (ESI-MS/MS) confirmed the presence of two of the predicted Cwp1p proteins, Cwp1.1p and Cwp1.2p. Crh1p, a putative 1,3-{beta}-glucan remodelling enzyme, was also identified. In silico genomic analysis further revealed five putative Pir proteins (Pir1–5p) and five members of the Bgl2 glycoside hydrolase family 17, belonging to a class of putative CWPs that can be extracted with NaOH. Immunological analysis of mild-alkali-extracted CWPs showed the presence of a ScPir2p homologue. Together, these experimental data and in silico predictions represent the first systematic analysis of the C. glabrata cell wall proteome.


Abbreviations: 1D, one-dimensional; 2D, two-dimensional; CAI, codon adaptation index; ConA, concanavalin A; CWP, cell wall protein; ER, endoplasmic reticulum; ESI-MS/MS, electrospray ionization tandem mass spectrometry; GPI, glycosylphosphatidylinositol (-modified)


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The opportunistic yeast Candida glabrata is now the second most common Candida species isolated from the bloodstream of patients in the USA and accounts for approximately 18 % of all yeast blood culture isolates (Pfaller et al., 2003b). C. glabrata infections are difficult to treat, as the organism is naturally less susceptible to fluconazole and amphotericin B compared to Candida albicans (Rex et al., 2000; Pfaller et al., 2002, 2003a). In the pleomorphic yeast C. albicans, virulence is related to its ability to switch between yeast and hyphal morphology (Lo et al., 1997), and both yeast and hyphal specific cell wall proteins (CWPs) have been identified (Klis et al., 2001). Although C. glabrata is capable of forming germ-like, nucleus-free tubes (Lachke et al., 2002) and ‘pseudohypha-like projections' at the periphery of colonies grown on nitrogen-starvation medium (Csank & Haynes, 2000; Calcagno et al., 2003), true hyphae have not been observed during infection. This suggests that, in contrast to C. albicans, pathogenicity of C. glabrata is solely mediated by the yeast growth form (Fidel et al., 1999).

The first contact between pathogenic yeast cells and human tissue occurs at the cell surface and is attributed to fungal cell wall components. For the pathogen C. albicans and for the baker's yeast Saccharomyces cerevisiae, models that describe their cell wall structure and organization have been developed. The major components of their walls are the polysaccharides 1,3-{beta}-glucan, 1,6-{beta}-glucan and chitin, and two types of covalently incorporated CWPs, glycosylphosphatidylinositol-modified (GPI) proteins, and proteins with internal repeats (Pir proteins). These proteins undergo post-translational modifications during secretion and they are often heavily mannosylated. The inner part of the wall is a skeletal layer formed by covalently connected 1,3-{beta}-glucan and chitin molecules to which 1,6-{beta}-glucan molecules are attached that in turn tether GPI-CWPs. These GPI-CWPs have been associated with various functions, such as maintaining cell shape, limiting permeability, hydrophobicity, and cell wall biosynthesis and remodelling (Scherrer et al., 1974; Zlotnik et al., 1984; Van der Vaart et al., 1995; Mouyna et al., 2000; Rodriguez-Pena et al., 2000). Pir proteins are directly linked to 1,3-{beta}-glucan via a currently unknown linkage, which is sensitive to NaOH treatment (Mrsa & Tanner, 1999; De Groot et al., 2004). In C. albicans several CWPs have been identified that mediate adhesion to host tissue and virulence (i.e. adhesins of the Als-family and Hwp1p) (Hoyer et al., 1999; Hoyer, 2001; Sundstrom, 2002; Cheng et al., 2003). Also, other functions related to Candida virulence, such as adhesion to plastics, biofilm formation and antigenicity, can be attributed to CWPs (Gomez et al., 1996; Chaffin et al., 1998; Nisini et al., 2001; Li & Palecek, 2003; Garcia-Sanchez et al., 2004). Recently, a genome-wide in silico survey identified 66 and 104 putative GPI proteins in S. cerevisiae and C. albicans (De Groot et al., 2003), respectively. The final destiny of these fungal GPI proteins will be either the cell wall or the plasma membrane and is influenced by the protein sequence in the C-terminal region after GPI modification (Vossen et al., 1997; Hamada et al., 1998, 1999; Terashima et al., 2003; Frieman & Cormack, 2003). However, the precise mechanism of cell wall incorporation of GPI proteins is still unknown. Besides GPI-CWPs, in both organisms a small family of Pir-CWPs was also identified. In S. cerevisiae, this family consists of four members which differ mainly in the number of internal tandem repeats. In C. albicans, two proteins have Pir protein characteristics, but only Pir1 was shown to be covalently bound to the cell wall (De Groot et al., 2004; Martínez et al., 2004). Recently, we have demonstrated that C. albicans Scw1p/MP65, a presumed 1,3-{beta}-glucanase/1,3-{beta}-glucanosyl transferase, is also (partly) covalently bound to the cell wall in a Pir-like manner (De Groot et al., 2004).

For C. glabrata, much less is known with respect to the glycan and protein composition of the cell wall. Only a small number of GPI proteins have been identified in this organism so far: the Epa1p adhesin (Cormack et al., 1999; Frieman et al., 2002), Epa2–5p (De Las Penas et al., 2003), Epa6–7p (NCBI database), and a family of three Gas/Phr orthologues (Weig et al., 2001). Using the second release of the annotated C. glabrata proteome, which was made publicly available in June 2004 by the Génolevures consortium (http://cbi.labri.fr/Genolevures/elt/CAGL), we have now performed a systematic, genome-wide prediction of GPI proteins of C. glabrata. GPI proteins can be identified by screening the genome for C-terminal consensus sequences for GPI modification. Essentially, we have applied the same algorithm that was used to predict GPI proteins of S. cerevisiae, C. albicans, Schizosaccharomyces pombe and Neurospora crassa (De Groot et al., 2003), but made a minor modification to make it more comprehensive. Additionally, putative Pir proteins were identified by screening for proteins with a characteristic pattern of four cysteine residues in the C-terminal region, whereas Bgl2 family members were identified by performing BLAST searches. In order to validate our in silico predictions, we also aimed to identify covalently linked CWPs from isolated cell walls using tandem mass spectrometry, and we studied the covalent linkages between CWPs and the cell wall polysacchararides.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In silico analysis.
The second release of the C. glabrata proteome, which was made publicly available in June 2004, was retrieved from the Génolevures website (http://cbi.labri.fr/Genolevures/elt/CAGL). In this release, comprising 5272 translated ORF sequences, two systematic numbers are provided for each ORF. One is a systematic individual protein file (IPF) number, used during the annotation process; the other indicates the position of the ORF on artificially assigned chromosomes (numbered A–M), and can be used to link the ORFs to NCBI protein files. Pattern searches for proteins having a GPI protein consensus sequence or the Pir-specific four-cysteine domain at their C-terminus (Fig. 1) were performed using the program FUZZPRO from the EMBOSS software package at http://www.hgmp.mrc.ac.uk/Software/EMBOSS/. The selection criteria to identify GPI proteins are described in De Groot et al. (2003). In short, ORFs were screened for (1) the presence of a C-terminal, fungal-specific, consensus sequence for GPI modification, (2) the presence of an N-terminal signal peptide for translocation across the membrane of the endoplasmic reticulum, and (3) the absence of internal transmembrane domains (Fig. 1a). The C-terminal GPI-modification algorithm is described as follows: [GNSDAC]-[GASVIETKDLF]-[GASV]-X(4,19)-[FILMVAGPSTCYWN](10)>, in Prosite format, where > indicates the C-terminal end of the protein. To optimize this GPI algorithm, we allowed the presence of glutamine in the hydrophobic tail. The presence of an N-terminal ER-import signal was analysed using the program SignalP V3.0 at http://www.cbs.dtu.dk/services/SignalP-3.0/. The standardized threshold value for signal peptides in the two algorithms [SignalP-NN (Smean) and SignalP-HMM (Sprob)] used in SignalP V3.0 was 0·5. The presence of internal membrane domains and protein localization was analysed using TMHMM (http://www.cbs.dtu.dk/services/TMHMM/) and PSORTII (http://psort.nibb.ac.jp). BLASTP searches were performed at NCBI (http://www.ncbi.nlm.nih.gov/blast). The results of our in silico GPI protein analysis are accessible at http://www.pasteur.fr/recherche/unites/Galar_Fungail/. The big-PI GPI protein prediction webserver is available at http://mendel.imp.univie.ac.at/gpi/fungi/gpi_fungi.html (Eisenhaber et al., 2004). Global protein sequence identity levels were calculated with the Needleman–Wunsch method of DNA Strider 1.3 using default settings. Theoretical pI values, molecular masses and the serine and threonine content of proteins were also calculated using DNA Strider 1.3. Multiple protein alignments and an unrooted phylogenetic tree were created using the webserver CLUSTALW at http://www.ebi.ac.uk/clustalw/index.html. Codon adaptation index (CAI) values were determined with the codon usage measurement tool CAI Calculator 2 (UMBC, University of Maryland; http://www.evolvingcode.net/codon/CalculateCAIs.php) using S. cerevisiae as a template. The maximum value of the CAI is 1·0 and the minimum value is 0. Highly expressed proteins have CAI values greater than 0·500.



View larger version (13K):
[in this window]
[in a new window]
 
Fig. 1. Schematic structure of GPI proteins (a) and Pir proteins (b). C-terminal motifs are in Prosite format. The GPI algorithm was adapted from De Groot et al. (2003). A minor modification in that algorithm, concerning the possible presence of glutamine (Q) in the hydrophobic tail, is underlined. GPI modification ({omega}) and Kex2p cleavage sites are indicated by arrows. SP, signal peptide for translocation across the membrane of the endoplasmic reticulum; TMD, transmembrane domain; S/T, serine/threonine; GPI, C-terminal GPI-modification signal peptide.

 
Yeast strains and growth conditions.
C. glabrata wild-type stain ATCC 90876 was used in this study. The isolate was routinely cultured on YPD plates at 30 °C. For cell wall analysis yeast cells were grown at 30 °C in synthetic complete (SC) medium, containing 2 % (w/v) glucose, 1·1 % (w/v) Casamino acids (Difco), 0·8 % (w/v) yeast nitrogen base with ammonium sulphate without amino acids (Becton Dickinson), 110 µg ml–1 L-leucine (Sigma), 55 µg ml–1 L-tyrosine (Sigma), 55 µg ml–1 L-tryptophan (Sigma) and 55 µg ml–1 adenine sulphate (Fluka).

Isolation of cell walls.
Cell walls of C. glabrata were isolated using a method that was previously described for C. albicans (Kapteyn et al., 1995; De Groot et al., 2004). Briefly, yeast cells were grown to exponential phase, washed and homogenized using glass beads in a FastPrep Instrument FP120 (Bio 101). The cell walls were isolated and washed extensively with 1 M NaCl. Subsequently, the cell walls were boiled twice for 5 min in the presence of 2 % SDS, 100 mM sodium EDTA, 40 mM {beta}-mercaptoethanol and 50 mM Tris/HCl, pH 7·8, to remove any contaminants derived from the cytosol and/or plasma membrane from the walls (Mrsa et al., 1997; Klis et al., 1998). Purified walls were freeze-dried, weighed and stored at –20 °C until use.

Chemical and enzymic release of cell wall proteins.
GPI proteins were extracted from isolated cell walls using hydrogen fluoride (HF)-pyridine as described previously (De Groot et al., 2004). Briefly, freeze-dried walls were incubated with HF-pyridine at 0 °C for 3 h. The reaction was quenched by diluting the reaction mixture with an equal amount of ice-cold H2O, followed by dialysis against H2O. After centrifugation, the supernatant containing the GPI-CWPs was freeze-dried. Alternatively, GPI-CWPs were released from cell walls using purified recombinant Trichoderma harzianum endo-1,6-{beta}-glucanase (Bom et al., 1998) (0·8 U per g wet weight of cell walls) as described elsewhere (Kapteyn et al., 1996, 1999). In order to extract alkali-sensitive CWPs, isolated cell walls were incubated with 30 mM NaOH for 17 h at 4 °C. The reaction was stopped by adding neutralizing amounts of acetic acid (Kapteyn et al., 1999).

Analysis of cell wall proteins.
Solubilized CWP fractions were separated by SDS-PAGE using a linear gradient of 2·2 % to 20 %. Proteins were stained either with silver (Bio-Rad) or colloidal Coomassie according to Neuhoff et al. (1988). For Western blot analysis, the proteins were transferred onto an Immobilon PVDF membrane (Millipore). To verify that the isolated cell wall fractions were not contaminated with cytosolic proteins after SDS/{beta}-mercaptoethanol extraction, the PVDF membranes were probed with polyclonal antibodies raised against the glycolytic enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH) of S. cerevisiae (Gozalbo et al., 1998) (results not shown). To visualize mannosylated proteins, PVDF membranes were probed with peroxidase-labelled concanavalin A (ConA; 1 µg ml–1, Sigma) in phosphate-buffered saline (PBS), containing 3 % (w/v) BSA, 2·5 mM CaCl2 and 2·5 mM MnCl2, as described elsewhere (Klis et al., 1998). We also used polyclonal antisera raised against the S. cerevisiae CWPs Pir2p (Russo et al., 1992) and Cwp1p (Shimoi et al., 1995), following the protocol described by Kapteyn et al. (1996, 1999). To increase the specificity of the antisera against Cwp1p and Pir2p, the membranes were pre-treated with 50 mM periodic acid and 100 mM sodium acetate (pH 4·5) for 30 min prior to the blocking step. To visualize binding of polyclonal antibodies to cell wall components, goat anti-rabbit IgG peroxidase secondary antibodies (Pierce) at a dilution of 1 : 10 000 in PBS/5 % (w/v) milk powder were used. Blots were developed using ECL Western blotting detection reagents (Amersham).

Two-dimensional (2D) gel electrophoresis.
For first-dimension gel electrophoresis, precast immobilized pH-gradient (IPG) strips (pH 3–10) 18 cm long were used for isoelectric focusing (IEF). IEF was carried out in an IPGphor device (Amersham). Approximately 200 µg protein was loaded for colloidal Coomassie stain and 50 µg was loaded for silver staining. Protein concentrations were determined using the Bradford protein assay (Merck). Rehydration was done for 10 h (30 V) using 8 M urea, 2 % w/v CHAPS, 1 % IPG buffer (pH 3–10, Amersham) and 70 mM DTT in a volume of 360 µl at 20 °C. The following gradient was applied: 300 V for 3 h, 600 V for 3 h and 1000 V for 3 h. The final focusing step at 8000 V was extended to 100 000 Volt-hours (Vh). For second-dimension gel electrophoresis, gradient gels (8 % to 16 %) were applied. After treatment of the IPG strips in SDS equilibration buffer, electrophoresis was performed in a Hoefer DALT SE 600-system (Amersham). To determine the relative molecular masses of the proteins, a pre-stained broad-range Marker (Bio-Rad) was added in the second dimension. Gels for mass spectrometry purposes were stained with Coomassie colloidal blue R250 as described by Neuhoff et al. (1988). The reproducibility of the protein profiles on 2D gels was evaluated by comparing HF-pyridine extracts from three independent cultures and cell wall preparations (data not shown).

Protein identification using electrospray ionization tandem mass spectrometry (ESI-MS/MS).
Protein spots were excised from the gel and incubated once in 100 µl 25 mM NH4HCO3 and twice in a solution containing 50 % acetonitrile and 25 mM NH4HCO3 for 15 min. Subsequently, the excised gel pieces were dried at room temperature under reduced pressure. In-gel digestion was performed on the dried gel pieces by treatment with 20 to 40 µl trypsin (2 µg ml–1) in 50 mM NH4HCO3 (sequencing grade; Promega) overnight at 37 °C. The supernatant was removed and stored, and the gel pieces were incubated in 30 to 60 µl 5 % formic acid for 30 min. The same volume of acetonitrile was added, and incubation was continued for 15 min. The supernatant was removed and pooled with the first supernatant, and the volume was reduced with a Speed Vac to approximately 10 µl. The peptides generated were desalted by using C18-ZipTip (Millipore). The purified peptides were eluted with 3 µl 65 % methanol/0·5 % formic acid.

Sequence data of internal peptides were acquired with a quadrupole time-of-flight mass spectrometer (Q-TOF-MS) by performing ESI MS/MS. For electrospray analysis and subsequent peptide sequencing, Au/Pd-coated nanospray glass capillaries (Protana) were filled with 3 µl ZipTip purified sample. The tip of the capillary was placed orthogonally in front of the entrance hole of a quadrupole time-of-flight mass spectrometry instrument (Q-TOF I; Micromass) equipped with a nanospray ion source. A capillary voltage between 750 V and 1000 V and a cone voltage of 35 V were applied. Doubly and triply charged peptides were chosen for collision-induced MS/MS fragmentation experiments and the corresponding parent ions were selectively transmitted from the quadrupole mass analyser into the collision cell. Argon was used as collision gas and the kinetic energy was set between 20 eV and 35 eV. The resulting daughter ions were separated by an orthogonal time-of-flight mass analyser. De novo peptide sequencing was carried out with the program Peptide-Sequencing within the BioLynx software (version 3.4; Micromass). Protein identifications were obtained by comparing the sequences of trypsin fragments with the proteins predicted from the C. glabrata genome sequence. For in silico digestion of proteins we used PeptideMass (http://ca.expasy.org/tools/peptide-mass.html).


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In silico identification of GPI proteins
To make an inventory of the proteins that are putatively covalently linked to the cell wall of C. glabrata, we first searched the proteome of C. glabrata for putative GPI proteins, comprising the major class of CWPs and plasma-membrane proteins. For this, we utilized a selection method that was developed to identify GPI proteins from S. cerevisiae, C. albicans, Sch. pombe, and N. crassa (see Methods). Applying the GPI-modification algorithm described by De Groot et al. (2003) as the first selection step (Fig. 1a), we identified 163 candidates in the C. glabrata proteome. Of these preselected ORFs, 81 met the additional selection criteria (see Methods) and were therefore designated putative GPI proteins (Table 1). For two ORFs, the transmembrane (TM) domain predictions were not fully unambiguous because PSORTII indicated a possible internal TM domain whereas TMHMM did not. These ORFs, CAGL0I10362g/CAGL-IPF6446 and CAGL0J02552g/CAGL-IPF6528, have a serine/threonine content of 28 % and 27 %, respectively, have low pI values, and show similarity to the Flo adhesin GPI protein family of S. cerevisiae (Halme et al., 2004) and to other C. glabrata adhesive proteins. Therefore, we have included them in our predicted GPI protein list. Additionally, 15 other ORFs, most of which lack a start codon, do not give a positive N-terminal signal peptide prediction but otherwise show all features that are often encountered in GPI proteins and are homologous to other GPI proteins (adhesive proteins and aspartic proteases). These ORFs probably represent C-terminal parts of larger GPI proteins and are therefore also included in Table 1.


View this table:
[in this window]
[in a new window]
 
Table 1. In silico identification and functional classification of putative GPI proteins of C. glabrata

 
In C. glabrata strain BG2, Epa1p, which is a member of a large GPI protein family, mediates 95 % of the adherence to human epithelial cells (Cormack et al., 1999). Five members (Epa1–5p) of this family have been published so far (De Las Penas et al., 2003), and recently, two new members (Epa6p and Epa7p) of this family have been submitted to the NCBI database. A BLAST search with Epa6p identified CAGL0C00110g/CAGL-IPF9053 as its closest homologue in the proteome of strain CBS138. This protein was not identified with our algorithm because it contains a glutamine residue in the hydrophobic tail (the C-terminal 10 amino acids), which is not commonly found in fungal GPI proteins (De Groot et al., 2003). This prompted us to test a modified GPI algorithm that would allow Q within the hydrophobic tail. This analysis resulted in 43 additional preselected ORFs, eight of which appear to be (part of) genuine GPI proteins. Thus, glutamines, although not frequently, may occur in the hydrophobic tail of GPI signal peptides. Allowing Q in the hydrophobic tail therefore has made our GPI algorithm less selective but more comprehensive. With this approach, we have in total identified 106 putative GPI proteins of C. glabrata. Eighty-six of these predicted GPI proteins are confirmed by applying the recently published big-PI Fungal Predictor (Eisenhaber et al., 2004). Of the 20 others, at least 17 have homology with known fungal GPI proteins, as predicted both by our method and by Eisenhaber's method, confirming that our method is comprehensive.

Functional classification of GPI proteins of C. glabrata
The 106 identified GPI proteins can be classified into different functional categories (Table 1).

Adhesive proteins.
No fewer than 51 of the predicted C. glabrata GPI proteins have features that suggest adhesive functions at the cell surface and thus are destined for cell wall incorporation. For 27 of these proteins, significant similarity with ‘adhesive’ proteins was directly indicated by homology searches using NCBI-BLAST. Among them are 13 proteins that can be grouped in the CgEpa-family. Other proteins in this group show homology with flocculins of S. cerevisiae and to a lesser extent with the Epa family, or to sexual agglutinins of S. cerevisiae or Hyr1p of C. albicans. All these proteins are predicted to mediate interaction of the fungal cell with its surroundings, be they other fungal cells, host cells, or inert materials like the plastics of a catheter. Five other proteins have significant similarity with the GPI proteins Yol155c of S. cerevisiae and Awa1p of sake yeast. Awa1p and Yol155c are not yet fully characterized; however, the former was shown to render yeast cells more hydrophobic (Shimoi et al., 2002). Both Yol155c and Awa1p have adhesin-like features: they are large proteins, have a high serine/threonine content and contain internal tandem repeats. For 19 translated ORFs, BLAST analysis indicated that they have homology with other members in the ‘adhesive proteins' category. Only seven of these proteins appear to be full-length. All these seven proteins are large (>=870 aa), have a serine/threonine content of at least 26 %, pI values of 5·1 or below, and they all contain internal repeats, indicating that they also belong to the ‘adhesive proteins' category. Interestingly, among the ‘adhesive’ proteins are 29 proteins, 14 of which seem to comprise partial ORFs, that represent subtelomeric sequences. All these subtelomeric proteins are homologous to ‘adhesive proteins' of different subcategories, including the Epa family. Consistent with this, Epa1p–Epa5p have all been localized to subtelomeric loci, and at least six other subtelomeric Epa-related proteins have been identified (De Las Penas et al., 2003).

Glycoside hydrolases.
Twelve of the predicted GPI proteins have homology with glycoside hydrolases (Coutinho & Henrissat, 1999). Five proteins belong to the Gas/Phr family of 1,3-{beta}-glucanosyl transferases (Mouyna et al., 2000). In S. cerevisiae and C. albicans, also five members of this family have been identifed. Two C. glabrata GPI proteins belong to the Crh family, which are classified in the glycoside hydrolase (GH) family 16, mainly consisting of 1,3-{beta}-glucanases and 1,3-1,4-{beta}-glucanosyl transferases. By performing a BLAST search, we identified a third member of the Crh family, CAGL0J08910g/CAGL-IPF6994, which like its closest homologues in S. cerevisiae (Crr1p) and C. albicans (Crh12p) does not conform to the GPI algorithm. With the nine Crh proteins from these organisms we constructed an unrooted phylogenetic tree to reveal possible functional subgroups within this family (Fig. 2a). Interestingly, Crh1p-like proteins from each organism grouped together, as was also the case for Utr2p/Crh2p-like proteins. The Crr1p-like proteins of C. glabrata and S. cerevisiae are also closely related to each other, but not to CaCrr1p. All nine proteins have a putative catalytic domain similar to that of bacterial endo-1,3-1,4-{beta}-glucanases (Fig. 2b) (Tabernero et al., 1994).



View larger version (30K):
[in this window]
[in a new window]
 
Fig. 2. The Crh families of C. glabrata, C. albicans and S. cerevisiae. (a) Phylogram of the Crh isoenzyme families based on amino acid sequences. (b) Alignment of the amino acid sequences of the putative catalytic domains of CAGL0G09449g/CAGL-IPF1559 (CgCrh1p), CAGL0C02211g/CAGL-IPF1070 (CgUtr2p), CAGL0J08910g/CAGL-IPF6994 (CgCrr1p), IPF14360.2 (CaCrh11p), IPF8080.2 (CaUtr2p), IPF5772.2 (CaCrh12p), Ygr189c (ScCrh1p) Yel040w (ScUtr2p) and Ylr213c (ScCrr1p). The numbers show the positions of the given amino acids in the protein sequence. The position of the proposed catalytic nucleophile is indicated by a single asterisk; the proposed catalytic proton donor is indicated by two asterisks (Hoj et al., 1992; Viladot et al., 1998). Conserved amino acid residues are given in boldface and amino acids that are identical in all sequences are boxed.

 
Two other proteins are homologous to the Dfg5/Dcw1 family belonging to glycoside hydrolase family 76, which comprises {alpha}-1,6-mannanases, whereas CAGL0D02530g/CAGL-IPF4864 is homologous to glucan exo-1,3-{beta}-glucosidases of glycoside hydrolase family 5 and CAGL0M08756g/CAGL-IPF8154 to the endoglucanase Egt2p of S. cerevisiae. All these proteins are orthologues of GPI proteins that are also present in other in silico-searched ascomycetous proteomes (De Groot et al., 2003; Eisenhaber et al., 2004), suggesting that they are ubiquitous in Ascomycetes. Finally, CAGL0A01452g/CAGL-IPF3033 is an orthologue of ScCwh41p, a processing glucosidase I enzyme (glycoside hydrolase 63) involved in modifying nascent N-glycosylation chains in the ER. Since ScCwh41p is not GPI-modified, and the big-PI algorithm did not detect GPI modification sites in CAGL0A01452g/CAGL-IPF3033, we suspect that this protein may be a false positive in our list. On the other hand, deletion of CWH41 in S. cerevisiae causes a 50 % reduction of 1,6-{beta}-glucan in the cell wall (Jiang et al., 1996).

Aspartic proteases.
Nine proteins have significant similarity with aspartic proteases of S. cerevisiae. Of the six aspartic proteases that have been identified in S. cerevisiae, only three are predicted GPI proteins according to our selection criteria. In the human pathogen C. albicans, ten aspartic proteases (Saps) have been identified; however, only Sap9p has GPI protein characteristics that conform to our GPI algorithm (De Groot et al., 2003).

Phospholipases.
The GPI algorithm identified three phopholipase B proteins, which belong to a class of well-established plasma-membrane-localized fungal GPI proteins.

Cell wall biogenesis.
Five proteins are proposed to play an important, as yet uncharacterized, role in cell wall biogenesis. Four have homology to the Ecm33 family, also an ubiquitous fungal GPI protein family. Deletion of ECM33 in S. cerevisiae results in swollen cells that secrete increased levels of 1,6-{beta}-glucosylated CWPs into the growth medium and are hypersensitive to drugs that interfere with cell wall biosynthesis (De Groot et al., 2001). Recently, we have shown that Ecm33.3p of C. albicans is covalently attached to the cell wall glycan network (De Groot et al., 2004). The last protein in this category, CAGL0M04169g/CAGL-IPF8285, has similarity with Kre1p of S. cerevisiae and C. albicans. Neither CAGL0M04169g/CAGL-IPF8285 nor ScKre1p and CaKre1p is predicted by big-PI to be GPI-modified. However, immunofluorescence and cell fractionation studies using epitope-tagged fusion proteins confirmed that ScKre1p is localized at the cell surface (Roemer & Bussey, 1995). Furthermore, the C-terminal region of ScKre1p, comprising the putative GPI-modification signal peptide, was shown to be vital for its proper localization and functioning (Breinig et al., 2002). Absence of KRE1 in S. cerevisiae resulted in a 40 % reduced level of cell wall 1,6-{beta}-glucan and a shorter chain length of this cell wall polymer (Boone et al., 1990).

Structural wall proteins.
Fourteen of the predicted GPI proteins have homology to CWPs of S. cerevisiae that are suggested to have a structural function. Deletion of their corresponding genes in S. cerevisiae did not result in strong alterations of the cell wall composition. However, some cell-wall-related phenotypes, such as increased sensitivities to the cell-wall-perturbing agents Congo red and Calcofluor white, have been observed (Van der Vaart et al., 1995; Garcia et al., 2004). Eight of the C. glabrata proteins in this group are homologous to the Srp1p/Tip1p family and three others to the abundant GPI-CWP Cwp1p of S. cerevisiae (Fig. 3). Homologues of the latter protein have not been identified in C. albicans.



View larger version (56K):
[in this window]
[in a new window]
 
Fig. 3. Alignment of the CgCwp1p family of C. glabrata and ScCwp1p of S. cerevisiae. The degree of conservation in each position is indicated by ‘*’ for identical residues, by ‘:’ for conserved substitutions, and by ‘.’ for semi-conserved substitutions in all sequences in the alignment. Peptide sequences of CgCwp1.1p and CgCwp1.2p derived from the Q-TOF MS/MS analysis are depicted in bold. Predicted N-terminal ER-import signals are underlined. The signal peptidase cleavage site of CgCwp1.1p was experimentally validated by ESI-MS/MS. Putative GPI attachment sites according to our GPI algorithm in the C-terminal region are boxed. Pir protein repeat-like sequences (in the C-terminal half of the proteins) are also underlined.

 
Others.
Five proteins have homology with S. cerevisiae proteins that cannot be directly associated with the cell wall. For the C. glabrata proteins with similarity to ScSpo19p and ScBud28p, it is not certain whether they are true GPI proteins, because their closest homologues in S. cerevisiae are not predicted to be GPI-modified.

Proteins of unknown function.
The remaining seven proteins have unknown functions.

In silico identification of putative mild-alkali-sensitive proteins
Established members of the class of covalently bound CWPs that can be extracted with mild alkali are Pir proteins (Kapteyn et al., 2001). Pir proteins share common features, namely, an N-terminal signal peptide for secretion, a Kex2p protease cleavage site, a central domain with a variable number of characteristic tandem repeats, and a C-terminal region composed of four cysteine residues with conserved spacing (Fig. 1b). Using the conserved cysteine pattern as the primary selection criterion, we identified five putative Pir proteins in C. glabrata. These five proteins also have sequence homology with known Pir proteins; therefore, we tentatively named these proteins CgPir1–5p (Table 2). CgPir1–4p have all of the above-described Pir protein features, but CgPir5p has no Kex2p cleavage site, no internal repeats, and a very low pI value compared to CgPir1–4p (Table 2). Since studies with mutated versions of ScPir4p/Cis3p indicated that the presence of the glutamine-rich repetitive sequence is required for its covalent attachment to wall carbohydrates (Castillo et al., 2003), it is uncertain whether CgPir5p can be covalently bound to the cell wall.


View this table:
[in this window]
[in a new window]
 
Table 2. The Pir protein family of C. glabrata

 
Recently, by mass spectrometric analysis of cell wall fractions of C. albicans, we discovered that besides Pir1p, also (part of) Scw1p/MP65 is covalently bound to the cell wall and can be extracted with mild alkali (De Groot et al., 2004). Scw1p is orthologous to the Bgl2 family in S. cerevisiae, generally known as proteins that can be released from cell walls with reducing agents (Cappellaro et al., 1998). Using the four S. cerevisiae orthologues and C. albicans Scw1p, BLAST analysis identified five members of this protein family in C. glabrata (Table 3). Interestingly, only two of these proteins, CAGL0G00308g/CAGL-IPF3933 and CAGL0M13805g/CAGL-IPF6053, appear to be closely related to C. albicans Scw1p, as well as to S. cerevisiae Scw4p and Scw10p (Table 3). CAGL0A01474g/CAGL-IPF3032 and CAGL0E02915g/CAGL-IPF9200, on the other hand, seem to be close relatives of ScScw11p, whereas CAGL0G00220g/CAGL-IPF8919 is much more closely related to ScBgl2p. Therefore, subgroups may exist within the Bgl2 protein family, belonging to glycoside hydrolase family 17, with respect to protein localization and function.


View this table:
[in this window]
[in a new window]
 
Table 3. Homology (% sequence identity) between C. glabrata Bgl2 family members and orthologues of S. cerevisiae and C. albicans

The size (number of amino acids) of the proteins is indicated in parentheses. Close relatives (sequence identity>50 %) are indicated in bold.

 
SDS-PAGE and Western blot analysis
As indicated by studies on the adhesin Epa1p (Frieman et al., 2002; Frieman & Cormack, 2003) and by our in silico analysis, C. glabrata contains covalently bound GPI-CWPs. In order to study CWPs on a proteomic level, cell walls of C. glabrata strain ATCC 90876 were isolated, treated with SDS, and incubated with HF-pyridine to extract GPI proteins (De Groot et al., 2004). The released material was separated by one-dimensional (1D) gradient SDS-PAGE and stained with silver, resulting in a major band of approximately 37 kDa and a high-molecular-mass smear (Fig. 4, lane 2). When the HF-pyridine-released material was blotted and probed with ConA–peroxidase, a 37 kDa protein, three additional bands, of approximately 80 kDa, 150 kDa and 250 kDa, and a high-molecular-mass smear were detected (Fig. 4, lane 4), suggesting that several GPI-CWPs were released. Immunoblot analysis further showed that the major 37 kDa band reacted with antiserum raised against S. cerevisiae Cwp1p, indicating that this abundant CWP of C. glabrata is homologous to ScCwp1p (Fig. 4, lane 3).



View larger version (115K):
[in this window]
[in a new window]
 
Fig. 4. C. glabrata contains HF-pyridine- and NaOH-extractable covalently bound CWPs. SDS-PAGE of protein fractions extracted with HF-pyridine (lanes 2–4) or mild-alkali (lanes 5–7) from isolated cell walls of C. glabrata. Visualization by silver staining (lanes 2 and 5), or by Western blot analysis using ConA–peroxidase (lanes 4 and 7), antiserum raised against ScCwp1 (lane 3) and antiserum raised against ScPir2p (lane 6). Lane 1, Size markers.

 
As in silico analysis indicated that C. glabrata also contains five Pir-related proteins, we analysed its cell walls for the presence of Pir CWPs. NaOH-released material, separated by gradient SDS-PAGE and stained with silver, appeared as a smear of high molecular mass (Fig. 4, lane 5). Western blot analysis using antiserum raised against S. cerevisiae Pir2p detected a distinct protein band of approximately 75 kDa (Fig. 4, lane 6). This indicates that at least one Pir protein is incorporated in the cell wall of C. glabrata via an alkali-sensitive linkage. Probing the same sample with ConA–peroxidase also visualized the 75 kDa protein, as well as additional bands and smears of approximately 30 kDa, 37 kDa, 150 kDa and >250 kDa, showing that multiple mannoproteins can be released from the cell wall of C. glabrata by mild alkali treatment (Fig. 4, lane 7). Together with the in silico identification of five Pir proteins and members of the ScBgl2p/CaScw1p protein family, these results indicate that multiple Pir proteins or Bgl2 family members or both are present in cell walls of C. glabrata.

Mass spectrometric identification of HF-pyridine-released proteins
To identify GPI-CWPs of C. glabrata by de novo sequencing, HF-pyridine-released material was separated by 2D gel electrophoresis. Three major groups of protein spots were observed after staining the gels with colloidal Coomassie. Each group consisted of protein spots of similar molecular mass but different pI (Fig. 5, arrows). The first group had a molecular mass of approximately 150 kDa and acidic pIs of 4–5. These spots probably correspond to the 150 kDa band that was detected when 1D gels with HF-pyridine extracts were blotted and probed with ConA–peroxidase (Fig. 4, lane 4). For the second group we estimated a molecular mass of approximately 37 kDa and pIs of 6–8, and the third group of protein spots showed a molecular mass of 33–34 kDa. The latter two groups seem to correspond to the abundant 37 kDa band that was observed on the 1D PAGE gel. We therefore speculated that these two groups of spots represented modifications of a single CWP. Six reproducible protein spots (1 to 6) were selected and excised for mass spectrometry purposes (Fig. 5).



View larger version (33K):
[in this window]
[in a new window]
 
Fig. 5. HF-pyridine-extracted CWPs separated by 2D gradient gel electrophoresis. Proteins were extracted from isolated cell walls of C. glabrata, separated by 2D gradient gel electrophoresis (8 % to 16 %) and stained with colloidal Coomassie. Isoelectric focusing was performed using precast IPG-strips (pH 3–10). The arrows point to the three groups of protein spots at 150, 37 and 34 kDa, respectively. Spots that were excised for ESI-MS/MS purposes are indicated by numbers. Size markers are shown on the right.

 
Determined peptide masses and sequences by tandem mass spectrometry of trypsin fragments of the excised spots are given in Table 4. All these masses and sequences perfectly matched with in silico tryptic digests of proteins from the C. glabrata proteome, leading to unambiguous identifications of all six protein spots. Spot 1 corresponded to CAGL0G09449g/CAGL-IPF1559, BLAST analysis of which indicated that its closest homologue in S. cerevisiae is Crh1p. Hence we named this protein CgCrh1p. Spots 2, 4, 5 and 6 were identified as CAGL0F07601g/CAGL-IPF545, and spot 3 as CAGL0F07579g/CAGL-IPF543 (Table 4). These two C. glabrata proteins are 95 % identical to each other and both have 50 % identity with Cwp1p of S. cerevisiae. These proteins are therefore termed CgCwp1.1p (CAGL0F07601g/CAGL-IPF545) and CgCwp1.2p (CAGL0F07579g/CAGL-IPF543). An alignment of ScCwp1p with the CgCwp proteins, in which the sequenced peptides of Cwp1.1p and Cwp1.2p are indicated in bold, is given in Fig. 3. These MS/MS data confirmed the results obtained from SDS-PAGE and Western blot analyses, namely, the presence of (an) abundant homologue(s) of ScCwp1p in the cell wall of C. glabrata. Furthermore, the results of the 2D gels and ESI-MS/MS analyses unequivocally show that Cwp1.1p and Cwp1.2p are both expressed and covalently bound to the cell wall, and also revealed the signal peptidase cleavage site of Cwp1.1p (Table 4).


View this table:
[in this window]
[in a new window]
 
Table 4. Identification of HF-pyridine-extracted CWPs from C. glabrata cell walls by ESI-MS/MS

Protein spots 1 to 6 were excised from 2D gels, digested with trypsin and analysed using ESI-MS/MS.

 
CgCwp1.1p is a 218 aa protein with a predicted molecular mass of 21·3 kDa (unprocessed precursor). CgCwp1.2p contains 212 aa and its calculated mass is 20·7 kDa. The higher molecular mass of 33–37 kDa observed on the gel as compared to the theoretical molecular masses of the proteins are probably due to post-translational modifications such as O-glycosylation. CgCwp1.1p and CgCwp1.2p possess serine/threonine-rich regions that may act as multiple acceptor sites for O-glycosylation but they do not contain potential acceptor sites for N-glycosylation (N-X-S/T). In S. cerevisiae, O-linked carbohydrate side-chains may contain mannosylphosphate groups, which confer a negative charge upon the proteins (Nakayama et al., 1998). Additions of phosphorylated O-linked carbohydrate side-chains on Cwp1.1p and Cwp1.2p molecules may therefore explain why protein spots comprising these proteins are found at different pIs. In the C-terminal region immediately upstream of the predicted GPI-attachment sites, adjacent basic residues are not present in these proteins, and also not in Crh1p. This supports the idea that adjacent basic residues may negatively influence cell wall incorporation of GPI proteins (Vossen et al., 1997; Frieman & Cormack, 2003), and is in agreement with the notion that Cwp1.1p and Cwp1.2p are among the major protein constituents of the cell wall of C. glabrata. High expression levels of Cwp1.1p and Cwp1.2p are also in agreement with codon adaptation index (CAI) values that were determined with the codon usage measurement tool CAI Calculator 2 (Sharp & Li, 1987). When S. cerevisiae was chosen as the template, CAI values that were obtained for CgCwp1.1p and CgCwp1.2p were 0·919 and 0·917, respectively. As a result of the in silico GPI protein search, a third homologue of ScCwp1p was discovered in the proteome of C. glabrata, named CgCwp1.3p (Table 1, Fig. 3). This protein is composed of 251 aa and has 40 % and 42 % identity with CgCwp1.1 and CgCwp1.2, respectively. This protein also does not contain a dibasic motif, suggesting that it is predominantly cell wall localized. However, we did not detect it with ESI-MS/MS. Consistent with its lower CAI value of 0·366, this protein is possibly expressed at lower levels compared to CgCwp1.1p and CgCwp1.2p. Interestingly, in the C-terminal region of all three GPI-modified CgCwp1 proteins a short sequence similar to the internal repeat sequences of Pir proteins is present. These internal repeats have been shown to be essential for covalent coupling of ScPir4 to the cell wall glycan network (Castillo et al., 2003). Such a sequence is also present in ScCwp1, which has been shown to become linked to the cell wall both through GPI modification and in a mild-alkali-sensitive manner (Kapteyn et al., 2001).


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In this paper, we present a first systematic analysis of the cell wall proteome of the human pathogen C. glabrata. We have also studied the covalent linkages between CWPs and the glycan network, and performed confirmative mass spectrometric identifications of CWPs. In S. cerevisiae and C. albicans, the major class of covalently bound CWPs are GPI-CWPs. Extraction of CWPs with HF-pyridine, which specifically cleaves phosphodiester bridges under the conditions used, indicated that C. glabrata also contains mannosylated GPI-CWPs that are linked to 1,6-{beta}-glucan through phosphodiester bridges. This architectural model was strengthened by data of 1,6-{beta}-glucanase digestions, which also liberated mannosylated CWPs (results not shown). Finally, mass spectrometric analysis of HF-pyridine-extracted CWPs resulted in the identification of three predicted GPI proteins (Crh1p, Cwp1.1p and Cwp1.2p). A similar strategy has been applied by Pitarch and coworkers to characterize the interactions of CWPs with other wall components in C. albicans (Pitarch et al., 2002). Our preliminary mass spectrometric analysis of 2D-separated proteins was not expected to yield a complete image of the cell wall proteome. However, using a combination of sequential protein fractionation, 2D PAGE and mass spectrometry we confirmed the presence of GPI-CWPs in C. glabrata and validated our in silico method to predict GPI proteins in this organism. In addition, particularly for Cwp1.1p, our analyses demonstrated that CWPs undergo post-translational modifications.

As shown by SDS-PAGE and Western blotting, the identified Cwp1.1p and Cwp1.2p correspond with the most abundant protein band on 1D gels of the HF-pyridine extracts. The homologous Cwp1p, the first identified GPI-CWP of S. cerevisiae, is also one of the most abundant CWPs (Van der Vaart et al., 1995), indicating a structural similarity between cell walls of these evolutionarily closely related yeasts. In contrast, a homologue of ScCwp1 is not present in C. albicans.

To make a comprehensive inventory of the putative GPI proteins in C. glabrata, we performed an in silico analysis by applying a slightly modified version of a fungal-specific algorithm which was previously developed to identify GPI proteins in S. cerevisiae and C. albicans (De Groot et al., 2003). Of the 106 predicted GPI proteins, 51 could be classified as putative adhesive proteins, possibly involved in interactions of C. glabrata with other yeast cells or adhesion to host cells or inert materials, and thus are likely to be incorporated in the cell wall. In agreement with this, only one of these proteins has adjacent basic residues immediately upstream of putative GPI-modification sites (Table 1). These proteins therefore may play an important role in establishing fungal infections. Consistent with this idea, it has been shown that the GPI-modified protein CgEpa1p mediates adhesion to cultured human epithelial cells (Cormack et al., 1999). Four other members of this family have been described so far, but more Epa-like proteins have been discovered (De Las Penas et al., 2003; NCBI database). Of the 13 Epa family members that we have identified with our GPI algorithm, CAGL0D06732g/CAGL-IPF8202 and CAGL0C05643g/CAGL-IPF4055 comprise 428 and 379 amino acids, respectively, and have significant similarity with only the N-terminal part of Epa proteins. This might suggest that these translated ORFs constitute C-terminally truncated versions of proteins that may not be present in full length in the current release of the C. glabrata proteome. On the other hand, in C. albicans, members of the Rbt5p/Csa1p GPI protein family, which also contain features that are indicative of proteins with adhesive properties, vary in length and are homologous only in their N-terminal regions (De Groot et al., 2003). Also, CAGL0D06732g/CAGL-IPF8202 and CAGL0C05643g/CAGL-IPF4055 were recognized by the GPI algorithm, suggesting that they are smaller full-length representatives of the Epa family. In contrast, CAGL0A00121g/CAGL-IPF9521, the closest homologue of Epa4p in the proteome of strain CBS138, was not selected by our algorithm and seems to comprise only the N-terminal part of Epa4p, as indicated by BLAST analysis. Possibly, large allelic variations may occur in the Epa adhesin family of C. glabrata, as is also the case for the Als family of C. albicans (Hoyer, 2001; Zhao et al., 2003).

The nonpathogenic S. cerevisiae contains only 11 known adhesive proteins, including agglutinins, Awa1p-like proteins, which mediate the foam-forming capacity of sake yeast (Shimoi et al., 2002), and flocculins. The human pathogens C. glabrata and C. albicans seem to contain a much larger number of putative adhesive proteins. In C. albicans, 24–26 of the predicted GPI proteins have adhesin-like properties, including the Als, Hwp1p/Rbt1p, Csa1p/Rbt5p and Hyr1p families. Conceivably, the high number of adhesive proteins may be related to the ability of the fungus to cause infections. In the polymorphic yeast C. albicans, differential regulation has been shown for Als3p, Hwp1p and Hyr1p, which are all specifically expressed or upregulated during hyphal development (Bailey et al., 1996; Hoyer et al., 1998; Staab et al., 1999), a growth form that does not seem to occur in C. glabrata. Hyphal-specific Hwp1p has been shown to mediate virulence in a mouse model of systemic infection (Staab et al., 1999; Sundstrom, 2002). Nevertheless, the fact that the nonfilamentous C. glabrata is a frequent cause of fungal infections indicates that, rather than the ability to undergo morphological development, a large repertoire of isofunctional adhesins may be required to develop infections. Possibly, differentially regulated adhesins enable human pathogenic fungi to make quick adaptations of their adhesion capacity to the altering conditions that are present in different human tissues, and to escape the hostile environment that is created by the host immune response.

At least 12 of the identified putative GPI proteins specify glycoside hydrolases, and most of these belong to GPI protein families that seem ubiquitous among ascomycetous fungi (De Groot et al., 2003; Eisenhaber et al., 2004). In C. albicans, we have shown recently that GPI-modified enzymes belonging to glycoside hydrolase families are present in HF-pyridine extracts of cell walls and are covalently bound to 1,6-{beta}-glucan (De Groot et al., 2004). We hypothesize that such a mechanism allows fungi to retain a desired enzyme activity in the wall, which otherwise would be lost to the surrounding growth environment. This would enable the organism to perform an effective and targeted remodelling and expansion of its cell wall structure during growth. The Crh1p protein that we have identified in HF-pyridine extracts is more closely related to ScCrh1p and CaCrh11p than to other members of this protein family. Interestingly, CaCrh11p is also covalently bound to the cell wall lattice (De Groot et al., 2004), and, using green fluorescent protein as a fusion marker, ScCrh1p was shown to be localized at the cell surface in a cell-cycle-dependent manner. ScCrh1p expression seems to correlate with chitin deposition in the cell wall (Rodriguez-Pena et al., 2000), suggesting a role, possibly indirect by remodelling the 1,3-{beta}-glucan network, for this enzyme in chitin incorporation in the wall. Notably, in contrast to the filamentous C. albicans and N. crassa, GPI proteins specifying chitinase activity were not identified in the nonfilamentous yeasts C. glabrata, S. cerevisiae and Sch. pombe. GPI modification of C. albicans chitinase 2 was confirmed by its identification in HF-pyridine extracts of cell walls (De Groot et al., 2004). We speculate that, in terms of the glycan composition, in particular chitin, differences in cell wall structure between fungi may be related to the ability to undergo filamentation.

Nine aspartic proteases of C. glabrata are predicted to be GPI-modified. These proteins are probably predominantly localized in the plasma membrane since they all have adjacent basic residues in their C-terminal regions. Interestingly, in C. albicans, secreted aspartic proteases (Saps) have been demonstrated to be virulence attributes, crucial for mucosal and systemic infections and involved in evasion of host immune responses (Hube & Naglik, 2001). However, in contrast to C. glabrata, only CaSap9p is predicted to be GPI-modified, suggesting differences in their defence strategy when approaching hostile environments.

In C. albicans and S. cerevisiae, a minor class of covalently incorporated CWPs can be released with mild alkali. The best-characterized members of this class are Pir proteins, five of which were identified in C. glabrata. In C. albicans cell walls, only a single Pir protein was found (De Groot et al., 2004), whereas S. cerevisiae walls contain at least three (Mrsa et al., 1997). In C. albicans, a 1,3-{beta}-glucanase with homology to the Bgl2 family of S. cerevisiae, CaScw1p, is also incorporated in a mild-alkali-sensitive manner. Of the five Bgl homologues that we identified in C. glabrata, two are closely related to CaScw1p, making them likely candidates to be incorporated in a similar way. SDS-PAGE and Western blot analysis using ConA confirmed the presence of mannosylated mild-alkali-sensitive proteins in C. glabrata. Among those are Pir proteins, as was shown by reactivity of the extracts with anti-ScPir2p antiserum. Thus, although its CWP composition may differ significantly, in terms of protein–glycan linkages, C. glabrata conforms to the cell wall architectural models that have been described for S. cerevisiae (Klis et al., 2002) and C. albicans (Klis et al., 2001). These models can therefore successfully be used to explore the cell wall structure of the human pathogen C. glabrata.


   ACKNOWLEDGEMENTS
 
Candida glabrata genome sequencing was performed by the Génolevures consortium. We thank B. Dujon and C. Hennequin for communication of unpublished results, A. Goretzki for excellent technical assistance, and D. Gozalbo, H. Shimoi and P. Russo for providing antibodies. P. W. J. de G. was supported by the European Commission (QLRT-1999-30795). M. W. was supported by the Forschungsförderprogramm of the University of Göttingen, by the MSD-grant Mycology 2001 and by the DFG (WE 3537/1-1).


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bailey, D. A., Feldmann, P. J., Bovey, M., Gow, N. A. & Brown, A. J. (1996). The Candida albicans HYR1 gene, which is activated in response to hyphal development, belongs to a gene family encoding yeast cell wall proteins. J Bacteriol 178, 5353–5360.[Abstract]

Bom, I. J., Dielbandhoesing, S. K., Harvey, K. N., Oomes, S. J., Klis, F. M. & Brul, S. (1998). A new tool for studying the molecular architecture of the fungal cell wall: one-step purification of recombinant trichoderma beta-(1-6)-glucanase expressed in Pichia pastoris. Biochim Biophys Acta 1425, 419–424.[Medline]

Boone, C., Sommer, S. S., Hensel, A. & Bussey, H. (1990). Yeast KRE genes provide evidence for a pathway of cell wall beta-glucan assembly. J Cell Biol 110, 1833–1843.[Abstract]

Breinig, F., Tipper, D. J. & Schmitt, M. J. (2002). Kre1p, the plasma membrane receptor for the yeast K1 viral toxin. Cell 108, 395–405.[CrossRef][Medline]

Calcagno, A. M., Bignell, E., Warn, P., Jones, M. D., Denning, D. W., Muhlschlegel, F. A., Rogers, T. R. & Haynes, K. (2003). Candida glabrata STE12 is required for wild-type levels of virulence and nitrogen starvation induced filamentation. Mol Microbiol 50, 1309–1318.[CrossRef][Medline]

Cappellaro, C., Mrsa, V. & Tanner, W. (1998). New potential cell wall glucanases of Saccharomyces cerevisiae and their involvement in mating. J Bacteriol 180, 5030–5037.[Abstract/Free Full Text]

Castillo, L., Martínez, A. I., Garcera, A., Elorza, M. V., Valentin, E. & Sentandreu, R. (2003). Functional analysis of the cysteine residues and the repetitive sequence of Saccharomyces cerevisiae Pir4/Cis3: the repetitive sequence is needed for binding to the cell wall beta-1,3-glucan. Yeast 20, 973–983.[CrossRef][Medline]

Chaffin, W. L., Lopez-Ribot, J. L., Casanova, M., Gozalbo, D. & Martinez, J. P. (1998). Cell wall and secreted proteins of Candida albicans: identification, function, and expression. Microbiol Mol Biol Rev 62, 130–180.[Abstract/Free Full Text]

Cheng, S., Clancy, C. J., Checkley, M. A., Handfield, M., Hillman, J. D., Progulske-Fox, A., Lewin, A. S., Fidel, P. L. & Nguyen, M. H. (2003). Identification of Candida albicans genes induced during thrush offers insight into pathogenesis. Mol Microbiol 48, 1275–1288.[CrossRef][Medline]

Cormack, B. P., Ghori, N. & Falkow, S. (1999). An adhesin of the yeast pathogen Candida glabrata mediating adherence to human epithelial cells. Science 285, 578–582.[Abstract/Free Full Text]

Coutinho, P. M. & Henrissat, B. (1999). Carbohydrate-active enzymes: an integrated database approach. In Recent Advances in Carbohydrate Bioengineering, pp. 3–12. Edited by H. J. Gilbert, G. Davies, B. Henrissat & B. Svensson. Cambridge: Royal Society of Chemistry.

Csank, C. & Haynes, K. (2000). Candida glabrata displays pseudohyphal growth. FEMS Microbiol Lett 189, 115–120.[CrossRef][Medline]

De Groot, P. W., Ruiz, C., Vázquez de Aldana, C. R. & 14 other authors (2001). A genomic approach for the identification and classification of genes involved in cell wall formation and its regulation in Saccharomyces cerevisiae. Comp Funct Genomics 2, 124–142.[CrossRef]

De Groot, P. W., Hellingwerf, K. J. & Klis, F. M. (2003). Genome-wide identification of fungal GPI proteins. Yeast 20, 781–796.[CrossRef][Medline]

De Groot, P. W., De Boer, A. D., Cunningham, J., Dekker, H. L., De Jong, L., Hellingwerf, K. J., De Koster, C. & Klis, F. M. (2004). Proteomic analysis of Candida albicans cell walls reveals covalently bound carbohydrate-active enzymes and adhesins. Eukaryot Cell 3, 955–965.[Abstract/Free Full Text]

De Las Penas, A., Pan, S. J., Castano, I., Alder, J., Cregg, R. & Cormack, B. P. (2003). Virulence-related surface glycoproteins in the yeast pathogen Candida glabrata are encoded in subtelomeric clusters and subject to RAP1- and SIR-dependent transcriptional silencing Genes Dev 17, 2245–2258.[Abstract/Free Full Text]

Eisenhaber, B., Schneider, G., Wildpaner, M. & Eisenhaber, F. (2004). A sensitive predictor for potential GPI lipid modification sites in fungal protein sequences and its application to genome-wide studies for Aspergillus nidulans, Candida albicans, Neurospora crassa, Saccharomyces cerevisiae and Schizosaccharomyces pombe. J Mol Biol 337, 243–253.[CrossRef][Medline]

Fidel, P. L., Jr, Vazquez, J. A. & Sobel, J. D. (1999). Candida glabrata: review of epidemiology, pathogenesis, and clinical disease with comparison to C. albicans. Clin Microbiol Rev 12, 80–96.[Abstract/Free Full Text]

Frieman, M. B. & Cormack, B. P. (2003). The omega-site sequence of glycosylphosphatidylinositol-anchored proteins in Saccharomyces cerevisiae can determine distribution between the membrane and the cell wall. Mol Microbiol 50, 883–896.[CrossRef][Medline]

Frieman, M. B., McCaffery, J. M. & Cormack, B. P. (2002). Modular domain structure in the Candida glabrata adhesin Epa1p, a beta1,6 glucan-cross-linked cell wall protein. Mol Microbiol 46, 479–492.[CrossRef][Medline]

Garcia, R., Bermejo, C., Grau, C., Perez, R., Rodriguez-Pena, J. M., Francois, J., Nombela, C. & Arroyo, J. (2004). The global transcriptional response to transient cell wall damage in Saccharomyces cerevisiae and its regulation by the cell integrity signaling pathway. J Biol Chem 279, 15183–15195.[Abstract/Free Full Text]

Garcia-Sanchez, S., Aubert, S., Iraqui, I., Janbon, G., Ghigo, J. M. & D'Enfert, C. (2004). Biofilms of Candida albicans: a developmental state associated with specific and stable gene expression patterns. Eukaryot Cell 3, 536–545.[Abstract/Free Full Text]

Gomez, M. J., Torosantucci, A., Arancia, S., Maras, B., Parisi, L. & Cassone, A. (1996). Purification and biochemical characterization of a 65-kilodalton mannoprotein (MP65), a main target of anti-Candida cell-mediated immune responses in humans. Infect Immun 64, 2577–2584.[Abstract]

Gozalbo, D., Gil-Navarro, I., Azorin, I., Renau-Piqueras, J., Martinez, J. P. & Gil, M. L. (1998). The cell wall-associated glyceraldehyde-3-phosphate dehydrogenase of Candida albicans is also a fibronectin and laminin binding protein. Infect Immun 66, 2052–2059.[Abstract/Free Full Text]

Halme, A., Bumgarner, S., Styles, C. & Fink, G. R. (2004). Genetic and epigenetic regulation of the FLO gene family generates cell-surface variation in yeast. Cell 116, 405–415.[CrossRef][Medline]

Hamada, K., Terashima, H., Arisawa, M. & Kitada, K. (1998). Amino acid sequence requirement for efficient incorporation of glycosylphosphatidylinositol-associated proteins into the cell wall of Saccharomyces cerevisiae. J Biol Chem 273, 26946–26953.[Abstract/Free Full Text]

Hamada, K., Terashima, H., Arisawa, M., Yabuki, N. & Kitada, K. (1999). Amino acid residues in the omega-minus region participate in cellular localization of yeast glycosylphosphatidylinositol-attached proteins. J Bacteriol 181, 3886–3889.[Abstract/Free Full Text]

Hoj, P. B., Condron, R., Traeger, J. C., McAuliffe, J. C. & Stone, B. A. (1992). Identification of glutamic acid 105 at the active site of Bacillus amyloliquefaciens 1,3-1,4-beta-D-glucan 4-glucanohydrolase using epoxide-based inhibitors. J Biol Chem 267, 25059–25066.[Abstract/Free Full Text]

Hoyer, L. L. (2001). The ALS gene family of Candida albicans. Trends Microbiol 9, 176–180.[CrossRef][Medline]

Hoyer, L. L., Payne, T. L., Bell, M., Myers, A. M. & Scherer, S. (1998). Candida albicans ALS3 and insights into the nature of the ALS gene family. Curr Genet 33, 451–459.[CrossRef][Medline]

Hoyer, L. L., Clevenger, J., Hecht, J. E., Ehrhart, E. J. & Poulet, F. M. (1999). Detection of Als proteins on the cell wall of Candida albicans in murine tissues. Infect Immun 67, 4251–4255.[Abstract/Free Full Text]

Hube, B. & Naglik, J. (2001). Candida albicans proteinases: resolving the mystery of a gene family. Microbiology 147, 1997–2005.[Medline]

Jiang, B., Sheraton, J., Ram, A. F., Dijkgraaf, G. J., Klis, F. M. & Bussey, H. (1996). CWH41 encodes a novel endoplasmic reticulum membrane N-glycoprotein involved in beta 1,6-glucan assembly. J Bacteriol 178, 1162–1171.[Abstract]

Kapteyn, J. C., Montijn, R. C., Dijkgraaf, G. J., Van den, E. H. & Klis, F. M. (1995). Covalent association of beta-1,3-glucan with beta-1,6-glucosylated mannoproteins in cell walls of Candida albicans. J Bacteriol 177, 3788–3792.[Abstract]

Kapteyn, J. C., Montijn, R. C., Vink, E., de la, Cruz. J., Llobell, A., Douwes, J. E., Shimoi, H., Lipke, P. N. & Klis, F. M. (1996). Retention of Saccharomyces cerevisiae cell wall proteins through a phosphodiester-linked beta-1,3-/beta-1,6-glucan heteropolymer. Glycobiology 6, 337–345.[Abstract]

Kapteyn, J. C., Van Egmond, P., Sievi, E., Van den Ende, H., Makarow, M. & Klis, F. M. (1999). The contribution of the O-glycosylated protein Pir2p/Hsp150 to the construction of the yeast cell wall in wild-type cells and beta 1,6-glucan-deficient mutants. Mol Microbiol 31, 1835–1844.[CrossRef][Medline]

Kapteyn, J. C., ter Riet, B., Vink, E., Blad, S., De Nobel, H., Van den Ende, H. & Klis, F. M. (2001). Low external pH induces HOG1-dependent changes in the organization of the Saccharomyces cerevisiae cell wall. Mol Microbiol 39, 469–479.[CrossRef][Medline]

Klis, F. M., Ram, A. F. J., Montijn, R. C., Kapteyn, J. C., Caro, L. H. & Vossen, J. H. (1998). Posttranslational modifications of secretory proteins. Methods Microbiol 26, 223–238.

Klis, F. M., de Groot, P. & Hellingwerf, K. (2001). Molecular organization of the cell wall of Candida albicans. Med Mycol 39 Suppl 1, 1–8.[Medline]

Klis, F. M., Mol, P., Hellingwerf, K. & Brul, S. (2002). Dynamics of cell wall structure in Saccharomyces cerevisiae. FEMS Microbiol Rev 26, 239–256.[CrossRef][Medline]

Lachke, S. A., Joly, S., Daniels, K. & Soll, D. R. (2002). Phenotypic switching and filamentation in Candida glabrata. Microbiology 148, 2661–2674.[Medline]

Li, F. & Palecek, S. P. (2003). EAP1, a Candida albicans gene involved in binding human epithelial cells. Eukaryot Cell 2, 1266–1273.[Abstract/Free Full Text]

Lo, H. J., Kohler, J. R., DiDomenico, B., Loebenberg, D., Cacciapuoti, A. & Fink, G. R. (1997). Nonfilamentous Candida albicans mutants are avirulent. Cell 90, 939–949.[Medline]

Martínez, A. I., Castillo, L., Garcerá, A., Elorza, M. V., Valentín, E. & Sentandreu, R. (2004). Role of Pir1 in the construction of the Candida albicans cell wall. Microbiology 150, 3151–3161.

Mouyna, I., Monod, M., Fontaine, T., Henrissat, B. & Lechenne, B., Latgé, J. P. (2000). Identification of the catalytic residues of the first family of beta(1-3)glucanosyltransferases identified in fungi. Biochem J 347, 741–747.[CrossRef][Medline]

Mrsa, V. & Tanner, W. (1999). Role of NaOH-extractable cell wall proteins Ccw5p, Ccw6p, Ccw7p and Ccw8p (members of the Pir protein family) in stability of the Saccharomyces cerevisiae cell wall. Yeast 15, 813–820.[CrossRef][Medline]

Mrsa, V., Seidl, T., Gentzsch, M. & Tanner, W. (1997). Specific labelling of cell wall proteins by biotinylation. Identification of four covalently linked O-mannosylated proteins of Saccharomyces cerevisiae. Yeast 13, 1145–1154.[CrossRef][Medline]

Nakayama, K., Feng, Y., Tanaka, A. & Jigami, Y. (1998). The involvement of mnn4 and mnn6 mutations in mannosylphosphorylation of O-linked oligosaccharide in yeast Saccharomyces cerevisiae. Biochim Biophys Acta 1425, 255–262.[Medline]

Neuhoff, V., Arold, N., Taube, D. & Ehrhardt, W. (1988). Improved staining of proteins in polyacrylamide gels including isoelectric focusing gels with clear background at nanogram sensitivity using Coomassie Brilliant Blue G-250 and R-250. Electrophoresis 9, 255–262.[Medline]

Nisini, R., Romagnoli, G., Gomez, M. J., La Valle, R., Torosantucci, A., Mariotti, S., Teloni, R. & Cassone, A. (2001). Antigenic properties and processing requirements of 65-kilodalton mannoprotein, a major antigen target of anti-Candida human T-cell response, as disclosed by specific human T-cell clones. Infect Immun 69, 3728–3736.[Abstract/Free Full Text]

Pfaller, M. A., Messer, S. A., Hollis, R. J., Jones, R. N. & Diekema, D. J. (2002). In vitro activities of ravuconazole and voriconazole compared with those of four approved systemic antifungal agents against 6,970 clinical isolates of Candida spp. Antimicrob Agents Chemother 46, 1723–1727.[Abstract/Free Full Text]

Pfaller, M. A., Diekema, D. J., Messer, S. A., Boyken, L. & Hollis, R. J. (2003a). Activities of fluconazole and voriconazole against 1,586 recent clinical isolates of Candida species determined by Broth microdilution, disk diffusion, and Etest methods: report from the ARTEMIS Global Antifungal Susceptibility Program, 2001. J Clin Microbiol 41, 1440–1446.[Abstract/Free Full Text]

Pfaller, M. A., Messer, S. A., Boyken, L., Tendolkar, S., Hollis, R. J. & Diekema, D. J. (2003b). Variation in susceptibility of bloodstream isolates of Candida glabrata to fluconazole according to patient age and geographic location. J Clin Microbiol 41, 2176–2179.[Abstract/Free Full Text]

Pitarch, A., Sanchez, M., Nombela, C. & Gil, C. (2002). Sequential fractionation and two-dimensional gel analysis unravels the complexity of the dimorphic fungus Candida albicans cell wall proteome. Mol Cell Proteomics 12, 967–982.

Rex, J. H., Walsh, T. J., Sobel, J. D., Filler, S. G., Pappas, P. G., Dismukes, W. E. & Edwards, J. E. (2000). Practice guidelines for the treatment of candidiasis. Infectious Diseases Society of America. Clin Infect Dis 30, 662–678.[CrossRef][Medline]

Rodriguez-Pena, J. M., Cid, V. J., Arroyo, J. & Nombela, C. (2000). A novel family of cell wall-related proteins regulated differently during the yeast life cycle. Mol Cell Biol 20, 3245–3255.[Abstract/Free Full Text]

Roemer, T. & Bussey, H. (1995). Yeast Kre1p is a cell surface O-glycoprotein. Mol Gen Genet 249, 209–216.[CrossRef][Medline]

Russo, P., Kalkkinen, N., Sareneva, H., Paakkola, J. & Makarow, M. (1992). A heat shock gene from Saccharomyces cerevisiae encoding a secretory glycoprotein. Proc Natl Acad Sci U S A 89, 3671–3675.[Abstract]

Scherrer, R., Louden, L. & Gerhardt, P. (1974). Porosity of the yeast cell wall and membrane. J Bacteriol 118, 534–540.[Medline]

Sharp, P. M. & Li, W. H. (1987). The Codon Adaptation Index – a measure of directional synonymous codon usage bias, and its potential applications. Nucleic Acids Res 15, 1281–1295.[Abstract]

Shimoi, H., Iimura, Y. & Obata, T. (1995). Molecular cloning of CWP1: a gene encoding a Saccharomyces cerevisiae cell wall protein solubilized with Rarobacter faecitabidus protease I. J Biochem 118, 302–311.[Abstract]

Shimoi, H., Sakamoto, K., Okuda, M., Atthi, R., Iwashita, K. & Ito, K. (2002). The Awa1 gene is required for the foam-forming phenotype and cell surface hydrophobicity of sake yeast. Appl Environ Microbiol 68, 2018–2025.[Abstract/Free Full Text]

Staab, J. F., Bradway, S. D., Fidel, P. L. & Sundstrom, P. (1999). Adhesive and mammalian transglutaminase substrate properties of Candida albicans Hwp1. Science 283, 1535–1538.[Abstract/Free Full Text]

Sundstrom, P. (2002). Adhesion in Candida spp. Cell Microbiol 4, 461–469.[CrossRef][Medline]

Tabernero, C., Coll, P. M., Fernandez-Abalos, J. M., Perez, P. & Santamaria, R. I. (1994). Cloning and DNA sequencing of bgaA, a gene encoding an endo-beta-1,3-1,4-glucanase, from an alkalophilic Bacillus strain (N137). Appl Environ Microbiol 60, 1213–1220.[Abstract]

Terashima, H., Hamada, K. & Kitada, K. (2003). The localization change of Ybr078w/Ecm33, a yeast GPI-associated protein, from the plasma membrane to the cell wall, affecting the cellular function. FEMS Microbiol Lett 218, 175–180.[CrossRef][Medline]

Van der Vaart, J. M., Caro, L. H., Chapman, J. W., Klis, F. M. & Verrips, C. T. (1995). Identification of three mannoproteins in the cell wall of Saccharomyces cerevisiae. J Bacteriol 177, 3104–3110.[Abstract]

Viladot, J. L., de Ramon, E., Durany, O. & Planas, A. (1998). Probing the mechanism of Bacillus 1,3-1,4-beta-D-glucan 4-glucanohydrolases by chemical rescue of inactive mutants at catalytically essential residues. Biochemistry 37, 11332–11342.[CrossRef][Medline]

Vossen, J. H., Muller, W. H., Lipke, P. N. & Klis, F. M. (1997). Restrictive glycosylphosphatidylinositol anchor synthesis in cwh6/gpi3 yeast cells causes aberrant biogenesis of cell wall proteins. J Bacteriol 179, 2202–2209.[Abstract]

Weig, M., Haynes, K., Rogers, T. R., Kurzai, O., Frosch, M. & Muhlschlegel, F. A. (2001). A GAS-like gene family in the pathogenic fungus Candida glabrata. Microbiology 147, 2007–2019.[Medline]

Zhao, X., Pujol, C., Soll, D. R. & Hoyer, L. L. (2003). Allelic variation in the contiguous loci encoding Candida albicans ALS5, ALS1 and ALS9. Microbiology 149, 2947–2960.[CrossRef][Medline]

Zlotnik, H., Fernandez, M. P., Bowers, B. & Cabib, E. (1984). Saccharomyces cerevisiae mannoproteins form an external cell wall layer that determines wall porosity. J Bacteriol 159, 1018–1026.[Medline]

Received 19 April 2004; revised 6 August 2004; accepted 12 August 2004.