Previous Article | Next Article ![]()
Infection and Immunity, October 2008, p. 4509-4517, Vol. 76, No. 10
0019-9567/08/$08.00+0 doi:10.1128/IAI.00368-08
Copyright © 2008, American Society for Microbiology. All Rights Reserved.

UMR Inserm 799, Laboratoire de Mycologie Fondamentale et Appliquée, Universitéde Lille 2, 59045 Lille cedex, France,1 Unité de Glycobiologie Structurale et Fonctionnelle, UMR CNRS 8576, 59655 Villeneuve d'Ascq, France,2 Groupe d'Etude des Interactions Hôte-Pathogène, UPRES EA 3142, UFR des Sciences Pharmaceutiques et d'Ingénierie de la Santé, 16 bd Daviers, 49000 Angers, France,3 Institut für Mikrobiologie, Heinrich-Heine-Universität, Universitätsstrasse, 1/26.12, D-40225 Düsseldorf, Germany4
Received 21 March 2008/ Returned for modification 11 May 2008/ Accepted 12 July 2008
|
|
|---|
|
|
|---|
-1,2 Man chains (see Fig. 3B) (57). β-Mans are epitopes of anti-C. albicans polyclonal antibodies (63, 74) and monoclonal antibodies (MAbs) with finely tuned specificity based on the β-Man chain length or the association with
-1,2-Mans (2, 7, 20, 22, 36, 70). Some of these antibodies have been shown to be protective in experimental models of systemic or vaginal candidiasis (9, 20, 21). These antibodies have also been used to study the distribution of β-Man epitopes on different C. albicans glycoconjugates (35, 49, 66, 70, 71). Initial mapping of β-Man epitopes led to the identification of a 14- to 18-kDa molecule which was characterized later as a cell wall glycolipid termed phospholipomannan (PLM) (69). PLM was the second C. albicans molecule for which structural evidence revealed the presence of β-Mans (73). However, C. albicans β-Man epitope mapping studies strongly suggested that several cell wall mannoproteins (CWMPs) are β-mannosylated (5, 42, 70, 71). C. albicans CWMPs are either noncovalently attached (for example, PPM) or covalently attached via disulfide bridges to other proteins, to β-1,6-glucans via a remnant glycosylphosphatidylinositol (GPI) anchor (GPI-anchored proteins or GpiPs) or, like PIR proteins, directly to β-1,3-glucans (12, 27, 59). Despite extensive studies on the different non-PPM CWMP families, nothing is known about their putative β-mannosylation and the process of attachment of β-Mans to the protein moiety. It is generally postulated that the mode of attachment is similar to that described for PPM, which is far from clear.
![]() View larger version (47K): [in a new window] |
FIG. 3. Mapping of β-1,2 oligomannoside epitopes after release of O-glycans by β-elimination and release of N-glycans by PNGase treatment of cell wall fraction 3 (A) and PPM as a control (B). Deglycosylated mannoproteins were separated by SDS-PAGE and analyzed by lectin staining with ConA or immunoblotting with anti-β-Man MAb 5B2. Mild acid hydrolysis was also performed to remove the acid-labile fraction of heat-extracted PPM (B). A PPM model with deglycosylation targets indicated by arrows is shown in panel B. An additional selective β-elimination procedure was performed with blotted mannoproteins from cell wall fraction 3 before immunodetection with MAb 5B2 (C).
|
|
|
|---|
Lectin and MAbs.
Horseradish peroxidase (HRP)-labeled concanavalin A (ConA) specific for terminal
-D-mannosyl and
-D-glucosyl residues was purchased from Sigma.
MAbs 5B2 and B6.1 (rat-mouse immunoglobulin M [IgM] and mouse IgM, respectively) were used for immunoblot assays. MAb C3.1 was used for enzyme-linked immunosorbent assay (ELISA) inhibition reactions according to its IgG nature. These three antibodies are specific for β-1,2-linked oligomannosides (20, 22, 70).
A rabbit polyclonal antiserum against the N-terminal Hwp1 fragment (34) and MAb 16B1 were used to analyze Hwp1. MAb 16B1, a mouse IgG, was first described as specific for C. albicans hyphal forms (38). Analysis of MAb 16B1 reactivity with different mutant strains lacking hypha-specific genes showed that this antibody is specific for Hwp1 (data not shown). The specificity of MAb 16B1 was confirmed in this study with the Hwp1-Nterm and Hwp1-Cterm recombinant proteins.
Cloning and expression of two recombinant fragments of Hwp1. Two recombinant fragments (176 and 431 amino acids) of Hwp1 (CA2825; http://genolist.pasteur.fr/CandidaDB/), a 634-amino-acid protein, were expressed in Escherichia coli. Two sets of primers (primers 5'CAAGGTGAAACAGAGGAAGCT3' and 5'TCAAGCAGGAATGTTTGGAGTAGT3' and primers 5'CCAACTGATTGGATCCCAGAT3' and 5'TCATTGAAATGTAGAAATAGGAGC3') were designed to clone Hwp1 N-terminal (amino acids 27 to 203) and C-terminal (amino acids 180 to 611) fragments. Amplified PCR fragments were cloned directly in the pEXP5-NT/TOPO vector (Invitrogen). This vector allows expression of recombinant proteins with an N-terminal polyhistidine (six-His) tag. E. coli strain DE3 was transformed with the two vectors and grown overnight in LB medium containing ampicillin (100 µg ml–1; Sigma) at 20°C with continuous shaking at 200 rpm. The cultures were used as inocula for fresh LB media containing ampicillin and were incubated until the optical density at 600 nm was 0.8 before 4 h of induction of the two Hwp1 recombinant fragments with 0.5 mM isopropyl-β-D-thiogalactoside (Sigma). Cells were harvested and suspended in sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) loading buffer.
Cell wall preparation and protein solubilization. Preculture cells were inoculated into YPD broth at a density of 106 cells ml–1 and grown at 37°C with continuous shaking at 150 rpm until the optical density at 600 nm of the culture was 1 (approximately 2 x 107 cells ml–1). Cells were harvested and washed twice with 50 mM Tris buffer (pH 8.0). Cells were suspended in lysis buffer (50 mM Tris [pH 8.0], 10 mM EDTA, protease inhibitors [protease inhibitor cocktail set IV; Calbiochem]), lysed, and homogenized by vortexing with acid-washed glass beads (0.5 mm; Sigma) for 30 min at 4°C. Cell walls were recovered by centrifugation for 5 min at 3,000 rpm at 4°C and washed five times with 1 M NaCl to remove bound cytoplasmic proteins. After two final washes in 10 mM Tris (pH 8.0), cell walls were suspended in 10 mM Tris (pH 8.0), and mannoproteins were solubilized sequentially. The procedure was derived from procedures described previously (10, 48), and a diagram of the procedure is shown in Fig. 1. Briefly, cell walls were heat treated twice for 15 min at 100°C. The pellet was then extracted with 15 mM dithiothreitol (DTT) in 10 mM Tris (pH 8.0) for 1 h at 56°C. The remaining pellet was divided into two fractions; one fraction was hydrolyzed with 15 mM NaOH for 16 h at 4°C and neutralized with acetic acid, and the other was digested with 5 mg ml–1 zymolyase 20T (MP Biomedicals) in 50 mM Tris (pH 7.5) containing 40 mM 2-mercaptoethanol and protease inhibitors for 1 h at 37°C. The solubilized CWMPs included noncovalently bound proteins, proteins bound via disulfide bonds, PIR proteins or proteins directly linked to β-1,3-glucans, and GPI anchor-derived proteins linked to β-1,6-glucans.
![]() View larger version (28K): [in a new window] |
FIG. 1. Diagram of the procedure used for sequential release of C. albicans CWMPs in a soluble form. C. albicans cell walls were first isolated. Noncovalently bound CWMPs were obtained after heating at 100°C (fraction 1). Mannoproteins bound to the cell wall via disulfide bridges were solubilized by DTT treatment at 56°C (fraction 2). Mannoproteins covalently bound to β-glucans were released either by mild alkali treatment (fraction 3) or by zymolyase treatment (fraction 4).
|
Chemical and enzymatic treatments. Twenty micrograms (carbohydrate content) of heat-released (fraction 1) (Fig. 1) and alkali-soluble (fraction 3) (Fig. 1) CWMPs was precipitated in cold 20% trichloroacetic acid on ice for 1 h, washed twice with ice-cold acetone, and suspended in the appropriate buffer for further analysis. N-glycans were released from mannoproteins by peptide-N-glycosidase F (PNGase F) treatment as recommended by the manufacturer (Sigma). O-glycans were β-eliminated from mannoproteins after 16 h of incubation at 4°C in 0.1 M NaOH. Alternatively, a more selective β-elimination method (15) was used to remove O-glycans without affecting N-glycans and protein moieties after SDS-PAGE. Mannoproteins transferred to an Immobilon P membrane were heated for 16 h at 40°C in 0.055 M NaOH. Blots were washed twice with distilled water before immunodetection.
Production of secreted Hwp1p and immunoprecipitation. Preculture cells of the SC5314 and CAHT-21 strains were inoculated into M199 medium (Invitrogen) at a density of 106 cells ml–1 in order to undergo the yeast-to-hypha transition. After 3 h of incubation at 37°C with shaking, medium supernatants were filtered through GF-F membranes (Millipore) and concentrated approximately 500-fold using centrifugal filter devices with a 100-kDa molecular mass cutoff (Centricon Plus-20; Amicon).
Concentrated CAHT-21 medium supernatant was diluted 1:10 in washing buffer (150 mM NaCl, 1% Triton X-100, 50 mM Tris HCl; pH 8.0). To 1 ml of sample, 5 µg of MAb 16B1 and 50 µl of µMACS protein G microbeads (Miltenyl biotec) were added. After 30 min of incubation on ice, the magnetic beads were washed four times with 200 µl of washing buffer and twice with 20 mM Tris HCl (pH 7.5). Immunoprecipitated protein was recovered with SDS-PAGE loading buffer.
SDS-PAGE and blotting. SDS-PAGE was performed by the method of Laemmli (33) using a Hoeffer gel system. Electrophoresis was carried out in 10% (wt/vol) acrylamide gels at 30 mA for 3 h. Prestained SDS-PAGE standards were obtained from Bio-Rad. Subsequently, the proteins on the gels were transferred electrophoretically to nitrocellulose (Whatman) or polyvinylidene difluoride (Millipore) membranes using a semidry blot system (OWL). After transfer, the membranes were washed in distilled water before they were stained with Ponceau red (Sigma) and destained with 5% acetic acid. Membranes were washed twice in TBST (0.05 M Tris HCl [pH 8.0], 0.15 M NaCl, 0.1% Tween 20), blocked for 1 h at room temperature with 5% nonfat milk in Tris-buffered saline, and probed for 2 h at room temperature with MAbs 5B2 (1:2,000 dilution), B6.1 (1:1,000 dilution), or 16B1 (1:1,000 dilution). After several washes in TBST, the membranes were probed for 2 h at room temperature with corresponding alkaline phosphatase-labeled secondary antibodies (Southern Biotech). Bound antibodies were revealed after incubation of the membranes in 0.1 M Tris (pH 9.5), 0.15 M NaCl, 5 mM MgCl2 containing nitroblue tetrazolium/5-bromo-4-chloro-3-indolylphosphate (BCIP) (Promega). Periodate oxidation was used in some cases before immunodetection, as described previously (45).
For lectin staining, membranes were blocked for 2 min at 20°C with PBS containing 2% Tween 20 (Sigma). After two washes with PBS, the membranes were incubated for 16 h at 20°C with 15 µg HRP-labeled ConA in PBS containing 0.05% Tween 20, 1 mM CaCl2, 1 mM MnCl2, and 1 mM MgCl2. Peroxidase activity was detected with diamidobenzidine (SIGMAFAST DAB; Sigma).
For recombinant proteins, six-His tags were detected with a nickel-nitrilotriacetic acid conjugate (nickel-HRP HisDetector; KPL) by following the manufacturer's instructions.
ELISA and inhibition assays. PPM of C. albicans strain SC5314 was extracted by the method of Kocourek and Ballou (32). S. cerevisiae PPM was purchased from Sigma. The acid-labile fraction of SC5314 PPM was obtained by hydrolysis in 10 mM HCl for 30 min at 100°C. After cooling and neutralization with NaOH, acid-stable PPM was removed using centrifugal filter devices with a 10-kDa molecular mass cutoff (Microcon 10K; Amicon). O-glycans from fraction 3 CWMPs were released after 18 h of incubation at 20°C in 0.03 and 0.1 M NaOH. After neutralization with HCl, CWMPs were removed by centrifugal filtration as described above.
Microtiter plates were coated with 50 ng PPM in 0.05 M sodium bicarbonate buffer (pH 9.6). After 1 h of incubation at 37°C and overnight incubation at 4°C, the wells were washed with PBS and blocked with PBS containing 3% bovine serum albumin (Sigma). Optimal concentrations of MAb C3.1 and HRP-labeled anti-mouse IgG (Southern Biotech) were determined by titration. MAb C3.1 was diluted 1:4,000 in PBS supplemented with 1% bovine serum albumin and 0.05% Tween 20 (PBSBT). Inhibition of MAb C3.1 binding to PPM was determined with various concentrations of either SC5314 PPM, S. cerevisiae PPM, oligomannosides released from C. albicans PPM by mild acid hydrolysis, or O-glycans released from fraction 3 CWMPs by β-elimination or with corresponding concentrations of NaCl as a control. After 1 h of incubation at 37°C and washes in PBS supplemented with 0.05% Tween 20, wells were incubated with secondary antibody diluted 1:25,000 in PBSBT. After four washes, HRP conjugates were detected with the trimethylbenzene substrate for 30 min at room temperature in the dark. The reaction was stopped with 1.5 N sulfuric acid, and the absorbance at 450 nm was determined. All assays were performed in triplicate.
|
|
|---|
-D-mannosyl and
-D-glucosyl residues and widely used to detect CWMPs (1, 26, 39), served as a systematic control in this study. As one of the aims of this study was to compare the well-defined PPM to other CWMPs, it was important to release noncovalently bound proteins, like PPM, without detergents, even at a low concentration, to avoid any contamination with covalently bound mannoproteins. As shown in Fig. 1, noncovalently bound CWMPs were released by heat treatment (fraction 1). This fraction contained essentially PPM as a high-molecular-weight polydispersed component reactive with ConA and anti-β-Man MAbs (Fig. 2A, lanes 1). Lower-molecular-weight and more defined CWMPs were also detected with ConA in this fraction, most of which also contained β-Man epitopes. DTT treatment (Fig. 1) solubilized proteins bound by disulfide bonds (fraction 2), which included a mannoprotein or a pool of mannoproteins that had molecular masses of approximately 180 kDa and were both
- and β-mannosylated (Fig. 2A, lanes 2). The remaining CWMPs directly linked either to β-1,3-glucans, like PIR proteins (fraction 3), or to β-1,6-glucans via a GPI remnant (fraction 4) were solubilized by mild alkaline hydrolysis or zymolyase treatment, respectively (Fig. 1). Figure 2A (lanes 3 and 4) shows that CWMPs from fractions 3 and 4 revealed by ConA also displayed β-Mans epitopes. There were some differences between the patterns displayed depending on the anti-β-Man MAb; MAb 5B2 gave a stronger signal for PPM, while MAb B6.1 had better affinity for other CWMPs, especially CWMPs from fraction 3. These differences might correspond to the different paratopes known to adapt to various β-1,2 oligomannose chain lengths and suggest differences in β-1,2-mannosylation between PPM and other CWMPs (55). After cold alkali hydrolysis, strong reactivity with ConA and anti-β-Man antibodies was still observed (lanes 3) due to the very mild conditions, suggesting that there was a partial impact on O-glycans. Similarly, there could have been several possible artifacts in fraction 4 resulting from proteolytic enzymes known to be present in zymolyase (28). However, the presence of high-molecular-weight polydispersed CWMPs in this fraction suggests that this effect could be limited (Fig. 2A, lanes 4). Nevertheless, this would not affect the conclusions about glycosylation processes, which are based on the members of a family of proteins or their subunits generated by proteolysis.
![]() View larger version (76K): [in a new window] |
FIG. 2. Mapping of β-1,2 oligomannoside epitopes on mannoproteins present in different cell wall extracts. Lanes 1 to 4 correspond to fractions 1 to 4 in Fig. 1, respectively. C. albicans (A) and S. cerevisiae (B) CWMPs released by heat treatment (lanes 1) or solubilized by either DTT (lanes 2), NaOH (lanes 3), or zymolyase (lanes 4) treatment were separated by SDS-PAGE before they were blotted and stained with ConA and MAbs 5B2 and B6.1 specific for β-1,2 oligomannoside. The position of PPM released by heat treatment is indicated by an arrow. The gels were loaded using material extracted from 50 µg (protein content) of cell wall. The slight signal observed with MAb 5B2 for the S. cerevisiae NaOH extract (lane 3) was also observed with MAb B6.1 (data not shown) and was interpreted as a nonspecific signal since it was also observed with a nonrelevant IgM MAb.
|
-Man terminal residues, with ConA than of S. cerevisiae CWMPs, suggesting that terminal β-Mans could hinder ConA binding in C. albicans.
β-Man epitopes are distributed in the N- and O-glycan parts of the different families of cell wall molecules, in contrast to PPM, in which these epitopes are present only in N-glycans.
The analysis first focused on the mild alkali extract as it contained a wide range of CWMPs (Fig. 2, lane 3). β-Elimination was carried out to remove O-glycans, while PNGase was used to hydrolyze N-glycans. Western blotting using ConA and MAb 5B2 showed that there was a reduction in
- or β-Man signals after both alkali and PNGase treatments, and there was less
- and β-Man epitope removal with PNGase (Fig. 3A). This suggests that association of the two types of mannose anomers with these proteins results from N-glycosylation processes as well as O-glycosylation processes. Similar results were obtained with the other cell wall extracts (Fig. 1, fractions 1, 2 and 4; data not shown). As this is the first report of an association between β-Mans and O-glycans from CWMPs putatively linked to β-1,3-glucans, we reviewed previous studies of a member of this CWMP family, which was initially designated MP65 by A. Cassone's group and was characterized with a MAb designated MAb AF1 (19, 44, 67). We showed that MAb AF1 was specific for β-Man epitopes with a minimal degree of polymerization of four residues (13, 70). Interestingly, the sequence of SCW1, the new designation for MP65, has only putative O-glycosylation sites (http://www.cbs.dtu.dk/services/NetOGlyc/). The complete absence of N-glycosylation sites on this protein strongly suggests that its characteristics, based on its β-mannosylated moiety, were due to the presence of O-linked β-mannose residues.
However, as these findings could also have resulted from incomplete action of PNGase and/or nonspecific β-elimination, we used PPM as a control; in this molecule the distribution of β-Mans is known to be restricted to the N-glycan part (57).
Use of mild acid hydrolysis to remove β-Mans present in the PPM acid-labile fraction resulted in a shift in the molecular weight and a reduction in the β-Man signal for the remaining acid-stable part. β-Man epitopes were removed completely after PNGase treatment (Fig. 3B), confirming their presence among N-glycans. Interestingly, as expected, β-elimination had no effect on β-Man expression. Based on these results obtained using a high-mannose-content molecule, incomplete hydrolysis of N-glycans by PNGase treatment can be ruled out, and data obtained using other molecules (Fig. 3A) strongly suggest that β-Mans are part of their N- as well as O-glycans. Furthermore, a control consisting of use of a selective β-elimination method to remove O-glycans on blots (15) led to similar conclusions (i.e., reduced binding of MAb 5B2 to mannoproteins) (Fig. 3C).
To confirm the presence of β-Mans in CWMPs, we performed an ELISA inhibition test. In this analysis, O-glycans released by β-elimination from fraction 3 CWMPs (Fig. 1) were used to inhibit binding of MAb C3.1 to PPM. MAb C3.1, which is also specific for β-Mans (22), was used instead of MAb 5B2 or B6.1 IgMs since IgG binding is more susceptible to inhibition by oligomers. As shown in Fig. 4, inhibition of MAb C3.1 binding was observed using the homologous C. albicans antigen; as expected, PPM acid-released oligomannosides, accounting for part of the PPM β-Man epitopes (56, 72), inhibited MAb C3.1 binding to PPM in a dose-dependent manner. Interestingly, similar inhibition was observed when O-glycans released from fraction 3 CWMPs were used. As it was previously, S. cerevisiae was used as a control. The PPM of this organism, widely described as devoid of β-Mans (31, 58, 64), did not prevent any binding. Altogether, these results show that in contrast to PPM, C. albicans CWMP O-glycans carry a β-Man (at least a β-mannotriose) epitope of MAb C3.1 (22).
![]() View larger version (21K): [in a new window] |
FIG. 4. Inhibition of MAb C3.1 binding to SC5314 PPM by different cell wall extracts in ELISA. The results are expressed as percentages of inhibition determined by comparison with the data for MAb C3.1 incubated alone with SC5314 PPM. The extracts involved in the inhibition reactions are indicated at the bottom. Each extract was used after twofold serial dilution to obtain carbohydrate concentrations ranging from 125 to 7.8 ng ml–1 for C. albicans PPM, from 1.9 to 0.12 µg ml–1 for acid-labile oligomannosides released from C. albicans PPM, from 0.9 to 0.125 µg ml–1 for O-linked oligomannoses released from fraction 3 CWMPs (CWMPs O-glycans), and from 250 to 15.62 ng ml–1 for S. cerevisiae PPM used as a control.
|
mutant CAHT-21. This mutant expresses a truncated form of Hwp1 without the predicted GPI anchor addition amino acid motif (60), so that it is secreted in a soluble form when C. albicans is grown under conditions inducing filamentation. The proteins recovered from the culture supernatants of CAHT-21 hyphal cells were analyzed by Western blotting using MAb 5B2, and the results were compared with results for SC5314, which was used as a control (Fig. 5). Two high-molecular-weight proteins reactive with MAb 5B2 were released from CAHT-21 cells (lane 3), while mainly the higher-molecular-weight protein was secreted by SC5314 cells (lane 1). Western blotting using anti-Hwp1 serum (lane 2) revealed that the lower-molecular-weight protein corresponded to the truncated Hwp1 protein. Trace amounts of Hwp1 could also be detected in SC5314 culture supernatants, corresponding to diffusion of Hwp1 into the medium, as reported previously (60). The higher-molecular-weight highly mannosylated protein had the characteristics of PPM, which was released in a soluble form by both strains. β-Man epitopes were lost after PNGase treatment, whereas β-elimination did not alter their expression (Fig. 5, lanes 6 and 4, respectively). Interestingly, truncated Hwp1p displayed the opposite pattern; its β-Man epitopes were conserved after PNGase treatment but were lost after β-elimination.
![]() View larger version (89K): [in a new window] |
FIG. 5. Comparative analysis of β-1,2 oligomannoside epitopes on Hwp1, a member of the GPI cell wall mannoprotein family (fraction 4), and PPM. Culture media of germ tubes from wild-type strain SC5314, which is known to release small amounts of Hwp1 (lane 1), and strain CAHT-21 secreting Hwp1 (lanes 2 to 6) (60) contained two molecular entities behaving like polydispersed material and corresponding to PPM (filled arrow) (50) and Hwp1 (open arrow), respectively. Secreted Hwp1 was stained with a rabbit antiserum raised against the N-terminal sequence of Hwp1 (lane 2), whereas PPM did not react. When stained by MAb 5B2 (lanes 1 and 3 to 6), both mannoconjugates were reactive, but they displayed opposite sensitivities after PNGase and β-elimination treatments. PNGase reduced β-Man signals on PPM and did not affect these signals on Hwp1, while β-elimination had no effect on PPM but dramatically reduced MAb 5B2 staining of Hwp1.
|
![]() View larger version (43K): [in a new window] |
FIG. 6. Confirmation of the presence of O-linked β-Man epitopes on Hwp1. (A) The specificity of MAb 16B.1 for Hwp1 was assessed by using six-His-tagged recombinant N-terminal (lanes 1 and 3) and C-terminal (lanes 2 and 4) fragments of this protein produced in E. coli. Western blots were probed with anti-His (lanes 1 and 2), which recognized both fragments, and MAb 16B.1 (lanes 3 and 4), which recognized the recombinant N-terminal Hwp1 fragment. (B) Lanes 1 and 2 show staining of Hwp1 (arrow) with MAb 16B1 for the culture supernatant and immunoprecipitate, respectively. (C) Lanes 1 and 2 show staining with MAb 5B2 of the culture supernatant and immunopurified Hwp1, respectively. (D) Same as panel C, lane 2, but after chemical release of O-linked oligomannosides on an Immobilon membrane, showing a reduction in the MAb 5B2 β-Man signal. (E) Control with MAb 16B1 staining after β-elimination. Note that the very sharp band that appeared at around 201 kDa in lanes 2 was unrelated to Hwp1 as detected by the secondary antibody alone. The 50-kDa protein corresponds to the immunoglobulin fragment recovered by immunoprecipitation.
|
C. albicans mutants affected in the early steps of O and N glycosylation have reduced β-1,2 oligomannoside expression.
C. albicans mutants lacking enzymes responsible for the first steps of O and N glycosylation are available. The pmts
mutants lack one of the five protein mannosyltransferases responsible for addition of the first mannose to serine or threonine residues of proteins (the first O-mannosylation step), while the och1
mutant lacks outer chain N-glycans (3, 52). It was interesting to check the CWMP β-Man epitope expression profiles of these mutants. Of the five pmts
mutants, we focused mainly on pmt1
as Pmt1p was shown to be partially responsible for Hwp1p glycosylation (60), and Hwp1 was described for the first time in the present study to be β-mannosylated. Western blot analysis of pmt1
fraction 4 (GPI-anchored CWMPs) with MAbs 5B2 and B6.1 revealed that the β-mannosylation was highly reduced compared to that of the wild-type strain extract (Fig. 7C, lanes 1 and 2). Less reduction was observed after mapping with ConA. Similar results were obtained when SC5314 and pmt1
fractions 1 and 3 (soluble and alkali-labile CWMPs) were analyzed (Fig. 7A and B, lanes 1 and 2). However, staining of fractions 1 and 3 from the pmt1
mutant with MAb B6.1 revealed less reduction than staining with MAb 5B2, suggesting that the mutation also affects the length of the β-Man chains. In parallel, cell walls of the five pmts
mutants (52) were isolated and compared to SC5314 isolated cell walls for expression of
- and β-Man epitopes. Even if there is redundant activity in some members of the Pmt family, the pmt1
and pmt2
mutants have defects in their
- and β-Man profiles due to O-mannosylation defects (data not shown). In addition, β-Man expression was investigated in the soluble, mild alkali, and zymolyase extracts (fractions 1, 3, and 4) of the och1
mutant. As expected, the
- and β-Man epitope expression patterns of the och1
mutant were impaired compared to those of the wild-type strain (Fig. 7, lanes 1 and 3), but β-Man epitopes were still detected. Interestingly, MAb B6.1, which is specific for β-mannotriose (20), had greater reactivity with och1
CWMPs than MAb 5B2. This result confirms the observation of Bates et al., who described a β-mannotriose released by acid hydrolysis of whole och1
cells (3). It is not known if this mannotriose is part of the O-glycan moiety of the mutant CWMP or part of the N-linked core structure.
![]() View larger version (89K): [in a new window] |
FIG. 7. Mapping of β-1,2 oligomannoside epitopes on surface soluble, alkali-solubilized, and GPI-anchored cell wall mannoproteins (fractions 1, 3, and 4, respectively) of strains defective in O and N glycosylation. Heat-released (A), alkali-labile (B), and zymolyase (C) extracts of strain SC5314 (lane 1) and pmt1 (lane 2) and och1 (lane 3) mutants were analyzed by Western blotting with ConA and anti-β-Man MAbs 5B2 and B6.1. Upstream prevention of O and N glycosylation resulted in important alterations in both - and β-Man detection for this family of proteins. Gels were loaded with 15 µg (protein content) of each extract.
|
Conclusions. Whatever experimental arguments are presented, this study demonstrated that β-Man epitopes are widely distributed among soluble (32), disulfide-linked (35), β-1,3-glucan-linked, and GPI anchor-derived (10) C. albicans cell wall molecules. The results also show that extrapolation from structural information obtained from PPM is not appropriate since, as it is in nonpathogenic species (68), O glycosylation is part of the C. albicans β-mannosylation process. As glycosylation modulates the function of a molecule within the cell wall and its impact on the host, this unforeseen complexity of virulence factor expression on CWMPs must be considered in order to obtain a greater understanding of C. albicans pathogenic potential (4). In association with the recent identification of a family of nine genes encoding β-mannosyl transfer in C. albicans (41), these findings should increase our understanding of how β-mannosylation is coordinated.
We thank J. E. Cutler (Research Institute for Children, New Orleans, LA) for critical reading of the manuscript and for providing MAbs B6.1 and C3.1. We are grateful to Paula Sundstrom for her helpful suggestions and for sharing strain CAHT-21. We thank J. Pontón (Universidad del País Vasco, Bilbao, Spain) and N. Gow (Department of Molecular and Cell Biology, Aberdeen, United Kingdom) for providing the rabbit polyclonal antibody raised against the N-terminal Hwp1p fragment and the och1
mutant strain, respectively. We thank Valerie Hopwood for editing the English of the manuscript.
Published ahead of print on 21 July 2008. ![]()
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»