Previous Article | Next Article ![]()
Eukaryotic Cell, June 2004, p. 675-684, Vol. 3, No. 3
1535-9778/04/$08.00+0 DOI: 10.1128/EC.3.3.675-684.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.
Program in Biochemistry and Cell Biology,1 Department of Molecular Genetics and Microbiology, State University of New York, Stony Brook, New York 11794-52222
Received 17 September 2003/ Accepted 25 March 2004
|
|
|---|
|
|
|---|
The filamentous growth of hyphae occurs by localized morphogenesis at the hyphal tip (20). As in other eukaryotic organisms, this polarized growth is mediated by the actin cytoskeleton. Actin and other morphogenesis proteins are highly enriched at the leading edge of hyphal growth. Interestingly, recent studies of other organisms have indicated that a polarized distribution of membrane lipids can also contribute to morphogenesis. In particular, these studies have focused on membrane microdomains known as lipid rafts that have been observed in organisms from yeast to humans (2, 8). Lipid rafts have a distinct membrane composition in that they are highly enriched in sterols and sphingolipids. An asymmetric distribution of lipid rafts in the plasma membrane has been implicated in promoting the cell polarity that mediates cell migration in a wide range of cells, including leukocytes, carcinomas, and epithelial cells (33). A remarkable example of membrane lipid asymmetry identified in a fungus is the dramatic polarization of sterol-rich membrane domains at the tips of pheromone-induced mating projections in Saccharomyces cerevisiae (4). Lipid polarization was not detected during budding growth and is thought to contribute to the highly polarized morphogenesis that occurs to create the mating projection. In view of the observation of lipid raft polarization in a wide variety of cell types, it seems likely that polar distribution of raft domains may be a common underlying feature of cell polarity and morphogenesis.
Lipid rafts are thought to contribute to morphogenesis in part because their distinct lipid composition allows for the partitioning of specific proteins into these domains. This results in the formation of specialized zones that are enriched in proteins involved in signal transduction pathways, cell adhesion, and other cell polarity processes. Interestingly, one subset of proteins that are enriched in raft domains are glycosylphosphatidylinositol (GPI)-anchored proteins (8, 9). This is significant in regard to hyphal morphogenesis in that several proteins involved in cell wall biogenesis are thought to contain GPI anchors (28). In addition, the adhesin proteins (e.g., Hwp1 and Als1p) are GPI anchored (46). Therefore, in the present study we examined whether polarized membrane domains are found in C. albicans. The results indicate that a highly polarized ergosterol-rich domain is present specifically during hyphal growth. This suggests that lipid rafts contribute to hyphal morphogenesis in C. albicans and are likely to also contribute to the presentation of virulence factors by this pathogen.
|
|
|---|
Fluorescence microscopy. Cells were stained with 10 µg of filipin/ml for 10 min and then analyzed by fluorescence microscopy. For colocalization studies, C. albicans strain yAW2, expressing CDC10-GFP under the control of the ADH promoter, was stained with filipin and analyzed by fluorescence microscopy. Rhodamine-phalloidin staining of actin was performed as described previously (1). Samples were photographed under an Olympus BH2 microscope by using a Zeiss Axio Cam camera run by Openlab 3.0.8 software from Improvision.
DRM isolation and Western blot analysis.
The isolation of detergent-resistant membranes (DRMs) was performed essentially as previously described (3). Briefly, cell pellets were lysed by agitation with glass beads in 100 µl of lysis buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 5 mM EDTA) supplemented with a protease inhibitor mixture (1 µg of pepstatin A/ml, 1 mM benzamidine, and 0.5 mM phenylmethylsulfonyl fluoride). The lysates were centrifuged at a low speed to remove glass beads and unbroken cells. Cleared lysates were incubated in the presence of cold 1% Triton X-100 for 30 min at 4°C. After the detergent treatment, lysates were subjected to density gradient centrifugation using a step gradient of different concentrations of Optiprep (Sigma). Fractions of equal volumes were removed from the tops of gradients and analyzed by Western blotting using a polyclonal antibody raised against Pma1p from Neurospora crassa (kindly provided by Amy Chang), a polyclonal anti-Cdc11p antibody (Santa Cruz Biotechnology, Santa Cruz, Calif.), or concanavalin A conjugated to horseradish peroxidase (Sigma). The antibodies were diluted in a solution of 1% powdered milk in TBS (10 mM Tris [pH 7.5], 150 mM NaCl), and concanavalin A was diluted in a solution of 0.3% bovine serum albumin in TBS. Control concanavalin A blots were preincubated with 2 mM methyl-
-D-mannopyranoside. Blots probed with antibodies were incubated with horseradish peroxidase-conjugated secondary antibodies, and then the cross-reacting proteins were detected by chemiluminescence by use of a SuperSignal West Dura kit as specified by the manufacturer (Pierce Inc., Rockford, Ill.). Concanavalin A blots were washed and incubated with the SuperSignal West Dura substrate to visualize glycosylated proteins. The protein concentrations of the cell lysates were assayed by using a BCA protein assay kit (Pierce Inc.).
|
|
|---|
![]() View larger version (41K): [in a new window] |
FIG. 1. Polarization of membrane ergosterol during hyphal growth of C. albicans. The images show filipin staining of strain BWP17 grown at 30°C in YPD to promote bud formation (A), at 37°C to promote pseudohyphal growth (B), at 37°C in medium containing 10% bovine calf serum (BCS) to promote hyphal growth (C and E), or at 37°C in Lee's medium to promote hyphal growth (D). The cells were stained with 10 µg of filipin/ml for 10 min and then analyzed by fluorescence microscopy. Bar, 10 µm.
|
![]() View larger version (37K): [in a new window] |
FIG. 2. Filipin staining of farnesol-treated cells. (A) Hyphal growth was induced in the presence of 10% BCS at 37°C. (B) Farnesol was added to serum-containing cultures at a final concentration of 1 mM to prevent the formation of hyphae. The cells were stained with 10 µg of filipin/ml for 10 min and then analyzed by fluorescence microscopy. Bar, 10 µm.
|
Analysis of DRMs in C. albicans. Since lipid rafts are enriched in ergosterol, we examined whether C. albicans contains lipid rafts by taking advantage of the detergent-resistant properties of the raft domains (8, 9). For this analysis, cell lysates were extracted with cold 1% Triton X-100 and then the DRMs were isolated by density gradient centrifugation. Fractions from the gradient were then analyzed in a Western blot probed with antibodies against Pma1p, a plasma membrane ATPase that has been shown in S. cerevisiae to be present in DRMs (2). Pma1p was detected in the DRM fraction of both budding and hyphal cells, indicating that lipid rafts also exist in C. albicans (Fig. 3). As a control for DRM isolation, fractions were analyzed by Western blotting with a concanavalin A probe to detect glycosylated membrane proteins. As shown in Fig. 3B, many glycosylated proteins were detected in fractions at the bottom of the gradient, indicating that they were efficiently solubilized by 1% Triton X-100. This is consistent with the previously reported ability of 1% Triton X-100 to discriminate between detergent-resistant and nonresistant fractions (43).
![]() View larger version (70K): [in a new window] |
FIG. 3. Pma1p is present in the DRM fraction of budding and hyphal cells. DRMs were isolated from budding and hyphal cells by density gradient centrifugation after membrane extraction with cold 1% Triton X-100, as described in Materials and Methods. Fractions of equal volumes were removed from the top of the gradient and analyzed by Western blotting using a polyclonal antibody raised against Pma1p (A) or concanavalin A (B). In panel A, fraction 2 from the first density gradient (spin 1), which consistently contained most of the Pma1p, was separated through a second density gradient (spin 2). The relative positions of protein standards, with masses indicated in kilodaltons, are shown to the left.
|
Time course analysis of membrane polarization during hyphal growth. The pattern of membrane polarization during hyphal growth was examined by staining cells with filipin at various time points after induction with serum (Fig. 4). At early time points, cells lacking a hyphal outgrowth showed intense staining at the presumptive site of germ tube formation. During the subsequent initiation of hyphal growth, staining was detected at the leading edge of growth.
![]() View larger version (75K): [in a new window] |
FIG. 4. Polarization of membrane ergosterol at different stages of hyphal growth. (A) Strain BWP17 was induced to form hyphae in medium containing 10% BCS, and then samples were collected for filipin staining at various time points up to 3 h. Representative cells from different times of induction are shown. The arrow indicates a cell that has completed septum formation. (B) Patterns of filipin staining quantified by different lengths of hyphae observed during the time course. Legend: none, uniform membrane staining; tip, staining only at hyphal tips; dual, staining at hyphal tips and presumptive septation sites; dual (+), dual staining after septum formation; tip (+), tip staining only after septum formation. At least 300 cells were counted for each hyphal length from 0 to 10 µm, at least 100 cells were counted for each hyphal length from 11 to 40 µm, and 67 cells were counted for hyphae of >40 µm. Bar, 10 µm.
|
Septins colocalize with ergosterol-rich membranes during hyphal growth. The pattern of filipin staining described above was reminiscent of that of septin proteins, which also localize to the hyphal tip and to sites of septation (45, 49). For a comparison of the localization of septins and ergosterol-rich membranes, CDC10-GFP cells, which produce the Cdc10 septin fused to green fluorescent protein, were grown in serum for various time points and analyzed by fluorescence microscopy. Septins colocalized with ergosterol-rich membranes at the septation site in two distinct phases (Fig. 5). At early stages, septins formed an immature single ring at the incipient septation site. Immature ring formation occurred independent of ergosterol-rich membranes, as no colocalization with filipin staining was observed at this stage (data not shown). At somewhat later stages, septins were localized adjacent to ergosterol-rich membranes at the boundary between the growing hyphal tip and the mother cell. Filipin staining extended back from the hyphal tip until it reached the septin ring and no further (Fig. 5A to D). Finally, when septum formation was under way, the septins separated into characteristic double rings and overlapped directly with filipin staining. The filipin staining extended away from the septin double ring in both directions for a short distance before becoming uniform with the rest of the membrane (Fig. 5E to H). These data suggest a role for septins in forming a boundary for lipid domains during hyphal growth, consistent with reports that septins act as a boundary domain for membrane proteins and actin during the budding growth of S. cerevisiae (6, 47). Septins also localized to the leading edge of growth, similar to ergosterol, but this is not obvious in Fig. 5 because septin localization at the tip is more diffuse and requires longer exposure times.
![]() View larger version (48K): [in a new window] |
FIG. 5. Septin colocalization with ergosterol-rich domains during hyphal growth. Strain yAW2, expressing the septin CDC10-GFP fusion construct under the control of the ADH promoter (49), was grown in 10% BCS at 37°C to promote hyphal formation and then was stained with filipin. Filipin staining is indicated in red, while CDC10-GFP staining is depicted in green. Images were merged with Openlab 3.0.8 software from Improvision. (A to D) Hyphal cells prior to division of the septin ring; (E to H) hyphal cells after division of the septin ring. (A and E) Filipin stain; (B and F) Cdc10-GFP stain; (C and G) merged images; (D and H) close-ups of merged images. Bar, 10 µm.
|
Actin is involved in membrane polarization. The actin cytoskeleton is involved in polar growth in many cell types, including hyphal growth in C. albicans (21, 39). To test whether the actin cytoskeleton is also involved in lipid raft polarization, we induced cells with serum containing either a high dose of latrunculin A (400 µM), which blocks budding and hyphal growth, or a low dose (200 µM) that was previously reported to prevent hyphal growth, but not budding (21). After 1 h of growth in the presence of the drug, ergosterol polarization was detected in only 53% ± 4.2% of cells treated with 200 µM latrunculin A, in contrast to the case for untreated controls (94.5% ± 5%) (Fig. 6A and C). Furthermore, rhodamine-phalloidin staining revealed reduced polarization of actin patches to sites of morphogenesis in cells treated with 200 µM latrunculin A compared to untreated controls (Fig. 6B and D). With 400 µM latrunculin A, polarized morphogenesis was not observed and ergosterol polarization was detected in only a very small minority of cells (Fig. 6E and F). These results suggest a role for actin in the polarization of plasma membrane lipids during C. albicans hyphal growth.
![]() View larger version (93K): [in a new window] |
FIG. 6. Actin is required for polarization of membrane lipids. Strain BWP17 was grown overnight in YPD medium and then diluted into medium containing 10% BCS and grown at 37°C to promote hypha formation. Latrunculin A was added to the cultures at a concentration of 200 or 400 µM. The cells were then grown for 1 h in the presence of the drug and stained with filipin to detect membrane sterols (A, C, and E) or rhodamine-phalloidin to detect actin (B, D, and F). Bar, 10 µm.
|
-demethylation of lanosterol in the ergosterol biosynthesis pathway (16, 37). When cells were treated with a low dose of ketoconazole (100 µM), hyphae grew more slowly than in untreated controls and often appeared swollen at their tips (Fig. 7). In addition, sterol polarization was detected in only 50% of cells, even after 1.5 h of growth in serum (Fig. 7B). When staining was observed, it was frequently localized to a more restricted region of the hyphal tip relative to the case for untreated controls (Fig. 7A, arrows). At a higher dose of ketoconazole (300 µM), hyphal growth was completely blocked and sterol polarization to bud tips was observed in only a small minority of cells. When staining was present, it was sometimes localized to a more restricted region (Fig. 7A, arrows) or uniformly present across the entire surface of the growing bud (Fig. 7A, arrowheads). These results indicate a role for sterol biosynthesis in hyphal growth and membrane lipid polarization in C. albicans.
![]() View larger version (52K): [in a new window] |
FIG. 7. Sterol biosynthesis is required for membrane polarization and hyphal growth. Strain BWP17 was diluted from an overnight culture into medium containing 10% BCS and grown for 1.5 h at 37°C. The cells were harvested at the indicated times. Ketoconazole was added at a final concentration of 100 or 300 µM. (A) Filipin staining of ketoconazole-treated cells and untreated controls. Arrows indicate cells displaying filipin staining at a more restricted region of the hyphal tip, and arrowheads indicate cells with uniform filipin staining across the entire surface of the growing bud. Cells were seeded onto coverslips at a low density to prevent clumping of the hyphae, and representative cells were moved closer together by using Adobe Photoshop. Bar, 10 µm. (B) Percent cells displaying filipin staining at sites of polarized morphogenesis in untreated control cells (white bar) or cells treated with 100 µM (gray bar) or 300 µM (black bar) ketoconazole. Error bars represent standard deviations for three experiments.
|
The addition of myriocin at a concentration (100 µM) that partially inhibited the growth of serum-stimulated cells disrupted hyphal morphogenesis (Fig. 8A). Hyphae appeared to emerge normally from mother cells (Fig. 8A, arrow), but hyphal tips subsequently became swollen and the morphology appeared less regular. The relatively normal morphology during the initial stages of hyphal growth in myriocin-treated samples suggested that cells from the overnight culture had a pool of sphingolipids that was used during the early phases of hyphal growth. Time course experiments were therefore performed to understand how hyphal morphology and sterol polarization change over time in the presence of myriocin (Fig. 8B). After 1 h of growth in the presence of myriocin, the percentage of cells displaying sterol polarization was slightly reduced relative to that for untreated controls (Table 1). By 2 h, the majority of cells exhibited abnormal hyphal morphologies and no longer stained for filipin (Fig. 8; Table 1). An analysis of the DRMs from the myriocin-treated cells also demonstrated that the level of Pma1p was diminished after 3 h (data not shown). This was consistent with the behavior of Pma1p in S. cerevisiae, which is targeted for vacuolar degradation when it is unable to enter DRMs in the secretory pathway (2, 19).
![]() View larger version (90K): [in a new window] |
FIG. 8. Sphingolipid biosynthesis is required for membrane polarization and hyphal morphogenesis. Strain BWP17 was diluted from an overnight culture into medium containing 10% BCS and grown for 2 h at 37°C. Myriocin was added at a final concentration of 100 µM to serum-stimulated cells, and phytosphingosine was added to rescued cultures at a final concentration of 30 µM. Representative morphologies of cells observed by light microscopy after 2 h of growth under the indicated conditions (A) or of cells from each sample stained with filipin after 1 h and 2 h of growth (B) are shown. Cells were seeded onto coverslips at a low density to prevent clumping of the hyphae, and representative cells were moved closer together by using Adobe Photoshop. Bar, 10 µm.
|
|
View this table: [in a new window] |
TABLE 1. Sphingolipid biosynthesis is necessary for sterol polarization to hyphal tips
|
|
|
|---|
Sterol polarization is specific to hyphae, as budding cells and pseudohyphae did not display sterol polarization at sites of active morphogenesis. When hyphal growth was prevented by a treatment with the quorum-sensing molecule farnesol, sterol polarization was still observed at the tips of growing buds. However, although farnesol-treated cells resemble pseudohyphae, they are distinct from pseudohyphae in that they express a hypha-specific reporter gene, HWP1-GFP (data not shown), suggesting that these cells are actively engaged in a hyphal differentiation program at the transcriptional level. This result further indicates that sterol polarization is specific to cells engaged in the hyphal morphogenesis program and is not a result of the unique geometry of the hyphal tip versus that of budding cells. The mechanism by which farnesol disrupts hyphal growth in C. albicans is not understood. However, a possible connection was suggested by the large increase in farnesol production that was observed when ergosterol biosynthesis was prevented downstream of farnesyl pyrophosphate (24). This suggests a link between farnesol and ergosterol production that could influence lipid raft functions and hyphal morphogenesis.
One interesting question that arose from the present study concerns the mechanism that maintains sterol polarization at the hyphal tip. Studies of several mammalian cell types have suggested that lipid raft aggregation at specific locations within the plasma membrane is enhanced by interactions between proteins found within these lipid domains. For example, T-cell-receptor interactions are thought to promote raft clustering and consequently the formation of multimeric signaling complexes at sites of T-cell aggregation (26, 27). It is possible that lipid raft polarization at the hyphal tips in C. albicans is brought about in a similar manner. Also, a role for the septin family of proteins in this process was suggested by the colocalization of septins with sterol-rich domains at hyphal tips and septation sites (Fig. 5) (45, 49). Septins establish domain boundaries between mother cells and buds in S. cerevisiae for a variety of proteins, including a transmembrane protein, Ist2p (6, 47). It is tempting to speculate a more general role for septins in forming domain boundaries for distinct lipid domains. However, our failure to detect significant amounts of septin proteins in DRM fractions suggests that other proteins are also involved.
The significance of lipid polarization during polarized morphogenesis in S. cerevisiae has recently been called into question (48). Valdez-Taubas and Pelham (48) argued that the sterol polarization observed in mating projections of S. cerevisiae results from endocytic recycling and the slow diffusion of membranes at the tip of mating projections. However, they failed to consider a role for septins in establishing a boundary domain at the base of mating projections. Septins localize at the base of mating projections and help to sequester proteins involved in mating to the shmoo tips (12, 14, 17). It is possible that they similarly act to maintain the polarization of lipid domains to this site. In addition, the polarization of sterol-rich domains to the tips of farnesol-treated cells suggests that membrane recycling and diffusion alone are not adequate for membrane polarization. Farnesol-treated cells grow with budding or pseudohyphal morphologies but are nevertheless able to activate some aspects of the hyphal differentiation program (Fig. 2; also data not shown). Thus, factors other than unique geometry and tip dynamics function to promote membrane polarization during hyphal growth.
Relationship between lipid rafts and C. albicans virulence. The detection of polarized lipid domains in C. albicans has many interesting implications for the ability of this organism to act as a pathogen. One aspect of this is the ability of proteins, including acylated proteins, pleckstrin homology (PH) domain-containing proteins, and GPI-anchored proteins, to partition into lipid raft domains. It is particularly interesting that C. albicans produces many GPI-anchored proteins that are specific to the hyphal phase of growth. The proteins thought to contain GPI anchors include members of the adhesin protein family (e.g., Hwp1p and Als1p) that mediate adhesion to host cells and members of the secreted aspartyl protease family (i.e., Sap9p and Sap10p) (28, 42, 46). These virulence factors are upregulated during hyphal growth, and their production is required for the invasive growth and virulence of the organism. Bioinformatic approaches have also predicted other likely GPI-anchored proteins whose functions are not yet known (29). Because many GPI-anchored proteins require sequestration into raft domains for delivery to the plasma membrane in other organisms, including S. cerevisiae, the present studies suggest that the polarization of membrane domains to the hyphal tip may directly contribute to the virulence of C. albicans.
Other GPI-anchored proteins, such as Eap1p, Dfg5p, and Phr1p, are known to be involved in epithelial adhesion, virulence, and proper hyphal growth (30, 41, 44). Phr1p is the closest C. albicans homolog of Gas1p from S. cerevisiae. Gas1p is involved in cell wall biogenesis and is a known lipid raft protein that fails to reach the cell surface when lipid rafts are disrupted (3). Interestingly, the hyphal phenotypes conferred by myriocin were similar to those observed in phr1
/phr1
strains of C. albicans grown at a neutral pH (41). Thus, the hyphal growth defects observed upon treatment with myriocin might be accounted for, in part, by a loss of Phr1p trafficking to the cell surface. Alternatively, Phr1p could have a more direct role in promoting membrane polarization. In some cases, clusters of GPI-anchored proteins are thought to promote the formation of lipid rafts (8). To examine this possibility, we used S. cerevisiae to test whether the overproduction of GPI-anchored proteins could promote ectopic lipid raft polarization in budding cells. However, GAL1 promoter overexpression in budding cells either GAS1 or SAG1, the gene encoding
-agglutinin, did not promote detectable membrane polarization (data not shown), suggesting that other proteins are required for membrane lipid polarization in yeast.
The presence of ergosterol-rich membrane domains at the tips of hyphae is also significant in that amphotericin, a drug that is widely used to treat C. albicans infections, acts by binding ergosterol and forming pores in membranes (15, 38). These observations suggest the interesting possibility that amphotericin analogs may be useful for targeting novel therapeutic agents that neutralize other virulence functions to hyphal tips.
This work was supported by grant RO1 AI47837 from the National Institutes of Health to J.B.K.
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»