Hemoglobin-Induced Binding of Candida albicans to the Cell-Binding ...

Download PDF
5 downloads 2568 Views 198KB Size Report
Comment
Dec 16, 1997 - proteolytic fragments of FN revealed that the cell-binding domain mediated this interaction. ... The adhesion of C. albicans to host tissues,.
INFECTION AND IMMUNITY, May 1998, p. 1904–1909 0019-9567/98/$04.0010 Copyright © 1998, American Society for Microbiology

Vol. 66, No. 5

Hemoglobin-Induced Binding of Candida albicans to the Cell-Binding Domain of Fibronectin Is Independent of the Arg-Gly-Asp Sequence SIZHUANG YAN, RUI G. RODRIGUES,

AND

DAVID D. ROBERTS*

Laboratory of Pathology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892 Received 5 November 1997/Returned for modification 16 December 1997/Accepted 25 February 1998

Hemoglobin specifically induces fibronectin (FN) binding to the pathogenic yeast Candida albicans. When grown in the complex medium Sabouraud broth, C. albicans expresses receptors that bind to several domains of FN. Growth in defined medium supplemented with 0.1% hemoglobin, however, enhanced the binding of FN to a single class of receptors, with a Kd 5 4.6 3 1028 M. Competitive binding assays using recombinant and proteolytic fragments of FN revealed that the cell-binding domain mediated this interaction. A recombinant 40-kDa fragment of FN consisting of type III repeats 9 to 13 had an inhibitory activity similar to that of the entire 120-kDa cell-binding domain, indicating that the C-terminal portion of the cell-binding domain contains the binding site. A recombinant 33-kDa fragment of the cell-binding domain and a 33-kDa fragment with the RGD sequence deleted had the same inhibitory activities, demonstrating that the RGD sequence recognized by some mammalian integrins is not required. The addition of hemoglobin to the culture medium also enhanced Candida cell adhesion to immobilized FN and to 120- and 40-kDa fragments of FN but not to the collagenbinding or fibrin I domains. Using ligand protection, we identified a surface protein from C. albicans with an apparent molecular mass of 55 kDa that was protected by both FN and the 40-kDa fragment derived from the cell-binding domain. Therefore, hemoglobin both induces FN binding and changes the relative affinities of C. albicans for the cell- and collagen-binding domains of FN. Candida albicans is one of the most common human opportunistic pathogens and causes a variety of diseases, from superficial candidiasis to systemic infections in immunocompromised hosts (7). The adhesion of C. albicans to host tissues, mediated through binding to various extracellular matrix proteins such as fibronectin (FN) and laminin, is correlated with pathogenicity (4, 6, 12, 14, 19). FN is a major component of the host extracellular matrix that may play an important role in the initiation and dissemination of C. albicans infections (6, 14, 18). As has been found for mammalian cells (23) and some pathogenic bacteria (5, 20), several domains of FN are recognized by C. albicans. Klotz and Smith (13) suggested that the cell-binding domain of FN mediated its binding to C. albicans. Further evidence indicated that peptides containing the RGD sequence from this domain inhibit the binding of C. albicans to FN both in solution and in adhesion assays (19). Additional domains of FN, however, have been found to interact with C. albicans (18). Four domains of FN can interact with C. albicans grown in Sabouraud dextrose medium (18). Among them, a proteolytic fragment from the gelatin- and collagenbinding domain of FN exhibited the highest affinity, i.e., it was as potent as intact FN. However, all of these interactions of C. albicans with FN were studied by using cultures grown in complex media. We have previously reported that hemoglobin induces a specific enhancement of FN-binding activity in C. albicans grown in defined medium (24). This induction is reversible and is not due to a bridge effect of hemoglobin between a receptor on the organism and FN. In addition, adhesion to immobilized FN was significantly increased for C. albicans grown in hemo-

globin-containing medium. Because cells must be growing in the presence of hemoglobin for several hours to induce binding, expression of a specific FN receptor may be induced by hemoglobin. We have now used these defined growth conditions to determine which domain of FN is recognized by the receptors induced by hemoglobin. We report here that binding and adhesion to the cell-binding domain are specifically induced by hemoglobin. This binding was localized to the 9th through 13th type III repeats of the cell-binding domain of FN but does not require the RGD sequence in the 10th type III repeat. Therefore, hemoglobin in defined medium specifically induces a class of receptors in C. albicans with a specificity different from those previously reported for this organism. MATERIALS AND METHODS Strains and growth conditions. C. albicans ATCC 44807 was used throughout this study. Its binding to FN in Sabouraud medium as well as in the chemically defined media yeast nitrogen base (YNB) and Lee Buckley Campbell broth has been well characterized (18, 24). Cultures were routinely propagated in freshly prepared YNB broth for 48 h at 26°C on a rocking platform except where otherwise indicated. The stationary-phase blastoconidia were aliquoted and frozen at 270°C until used. For each experiment, organisms were thawed, inoculated into 6 ml of YNB medium with or without hemoglobin, and incubated as described above for 20 to 48 h. Under both growth conditions, no germination was found upon microscopic examination. Proteolytic or recombinant fragments of FN and RGD synthetic peptides. FN was purified from frozen human plasma as previously described (18). Proteolytic fragments of FN were obtained from Telios Pharmaceuticals, Inc., San Diego, Calif., or Gibco BRL, Gaithersburg, Md. Recombinant fragments derived from several domains of FN, expressed in Escherichia coli, were purified and refolded as described previously (22). Construction of the cell-binding domain with the RGD sequence deleted was carried out by oligonucleotide site-directed mutagenesis using the 33-kDa expression vector (22). The origins in the FN sequences of the fragments used in this paper are presented schematically in Fig. 1. FN-binding assay. FN was iodinated to a specific activity of 1 to 2 mCi/mg with Iodogen (Pierce, Rockford, Ill.), and unbound iodine was removed by being passed through a PD-10 column (18). In a typical binding assay, 2 3 106 C. albicans cells were exposed to 125I-FN at a final concentration of 0.5 to 1 mg/ml in

* Corresponding author. Mailing address: Bldg. 10, Room 2A33, 10 Center Dr., MSC 1500, NIH, Bethesda, MD 20892-1500. Phone: (301) 496-6264. Fax: (301) 402-0043. E-mail: [email protected]. 1904

HEMOGLOBIN PROMOTES C. ALBICANS BINDING TO FN

VOL. 66, 1998

1905

FIG. 1. FN and proteolytic and recombinant fragments derived from FN. I, II, and III represent, respectively, type I, type II and type III repeating motifs. Proteolytic fragments are indicated by the prefix p, and recombinant fragments are indicated by the prefix r. Numbers following p or r indicate the molecular mass in kilodaltons. The arrows indicate the positions of cell adhesion recognition sequences RGD, LDV, and REDV. The asterisk indicates the position of the RGD sequence deletion from the r33 fragment (r33D).

a total volume of 200 ml of Dulbecco’s phosphate-buffered saline (DPBS) without CaCl2 and MgCl2–0.1% bovine serum albumin (BSA), pH 6.0, in a polypropylene tube and were incubated for 3 h with shaking on a rotary plate at 160 rpm. The cell suspensions were transferred to microtubes, and blastoconidia were separated from unbound 125I-FN by centrifugation through 100 ml of an oil mixture of dibutyl phthalate-dioctyl phthalate (2:1). Radioactive FN bound to the cell pellet was counted in a gamma counter (Packard Instrument Company, Downers Grove, Ill.). For inhibition assays, the binding of radiolabeled protein was determined in the presence of various concentrations of unlabeled FN, proteolytic or recombinant fragments of FN, or synthetic peptides. Adhesion to immobilized FN and its fragments. Either FN or its fragments were used to coat glass slide wells by adding 300 ml of a solution of FN or FN fragments (0.01, 0.1, 1, or 10 mg/ml) in DPBS without Ca21 and Mg21 (pH 7.5) per well to Chamber slides. The slides were incubated at 4°C overnight (24). Candida cells prepared from cultures with or without hemoglobin in the YNB medium were added to each well at a concentration of 2 3 106 CFU/ml and allowed to incubate at room temperature for 2 h; the wells were then washed three times with DPBS. Attached Candida cells were fixed, stained, and counted. Aggregates of Candida cells larger than four cells were not counted. The numbers of cells attached per square millimeter of surface were determined in triplicate and are presented as the means 6 standard deviations (SD). Biotinylation and extraction of surface proteins of C. albicans. To identify surface proteins of hemoglobin-induced C. albicans that bound FN and the recombinant 40-kDa fragment (r40), a modified biotinylation procedure was used (25). Briefly, 20 ml of Candida cultures grown in 43 YNB broth with hemoglobin at 26°C for 48 h were harvested by centrifugation and the cells were washed twice with DPBS. The cultures were divided into three parts. The cells were incubated with 1 ml of FN or r40 (both at the concentration of 100 mg/ml) in DPBS at room temperature for 2 h with shaking, and the remaining fraction was incubated in PBS without a ligand. The cells were centrifuged to remove excess unbound ligand, and the pellets were suspended in 1 ml of sulfosuccinimidyl-3-(4-hydroxyphenyl) proprionate (sulfo-SHPP; 0.5 mg/ml; Pierce) in 50 mM sodium bicarbonate buffer, pH 7.8, and incubated at room temperature for 1 h with shaking. The pellets were then washed with DPBS (four times, 15 min each) to remove bound FN or fragments. Finally, each fraction of cells was subjected to biotinylation to label surface proteins with lysyl residues that were protected by bound proteins. Following biotinylation by sulfohydroxysuccinimidyl-6-(biotin amido) hexanoate (Pierce), cell surface proteins were extracted with Lyticase (Sigma, St. Louis, Mo.) and 8 mM dithiothreitol. Cell pellets were removed by centrifugation, and extracted cell surface proteins in the supernatant fluids were separated by sodium dodecyl sulfate–10% polyacrylamide gel electrophoresis. Following electrophoresis, proteins were transferred onto nitrocellulose membrane, blocked with 3% BSA in 50 mM Tris (pH 7.5)–150 mM NaCl, and incubated with streptavidin-horseradish peroxidase (Amersham, Pittsburgh, Pa.). Biotinylated cell surface proteins were visualized with an ECL chemiluminescent detection kit (Amersham).

FN binding to C. albicans grown in hemoglobin-supplemented medium is similar to that for FN binding to the low-affinity receptors induced by the complex Sabouraud medium, i.e., Kd values of 4.6 6 0.5 3 1028 M (Table 1) and 1.2 3 1027 M (18), respectively. Thus, the low-affinity receptors are up-regulated following growth with hemoglobin. The cell-binding domain of FN mediates hemoglobin-induced binding to C. albicans. Because the binding of FN to C. albicans grown in defined medium in the presence of hemoglobin differed in affinity from that measured in complex medium, we examined whether different domains of FN might mediate these interactions. Several recombinant or proteolytic fragments of FN were used to inhibit the binding of radiolabeled FN to C. albicans (Table 2). Native FN inhibited the binding of radiolabeled ligand with a 50% inhibitory constant (IC50) of 32.5 nM (Table 2). The most potent fragments of FN, the 120-kDa proteolytic fragment [p120] and r40, were derived from the cell-binding domain and had IC50s of 210 and 320 nM, respectively. A proteolytic 30-kDa fragment derived from the collagen-binding domain in the N-terminal domain of FN, which showed the greatest inhibition of FN binding to C. albicans grown in complex medium (18), was less active (Fig. 2), and fragments from the fibrin-binding domain in the N terminus (r12, r18.5, and r31) responsible for FN binding to gram-

RESULTS In contrast to the heterogeneous binding of FN to C. albicans grown in Sabouraud broth (18), an analysis of FN binding to C. albicans grown in the defined medium supplemented with 0.1% hemoglobin with the LIGAND program (17, 21) indicated the presence of only one class of binding sites, with a Ka 5 2.2 3 107 M21 (24). Inhibition of hemoglobin-induced 125 I-FN binding to C. albicans with unlabeled FN was complete and dose dependent (Fig. 2) (24). The dissociation constant for

FIG. 2. Competitive displacement of FN and its derived fragments binding to C. albicans. Unlabeled FN (E), the proteolytic 120-kDa cell-binding domain (F), and proteolytic 30- (ƒ) and 40-kDa fragments (Œ) were used as inhibitors of 125 I-FN binding to C. albicans. C. albicans cells (2 3 106), prepared from a 48-h culture grown in YNB medium supplemented with 1 mg of hemoglobin per ml, were incubated with radiolabeled ligand and the indicated concentrations of unlabeled proteins at 26°C for 3 h. Binding is presented as the percentage of that measured in the absence of inhibitors; values are means 6 SD (n 5 3).

1906

YAN ET AL.

INFECT. IMMUN.

TABLE 1. Binding constants for hemoglobin-induced C. albicans blastoconidia determined by displacement with FN or FN fragmentsa Ligand

Mean Kd (M) 6 SD

FN p120 r40

(4.6 6 0.5) 3 1028 (8.8 6 1.8) 3 1028 (2.6 6 0.8) 3 1027

No. of binding site receptors/cell

4.2 3 105

a Binding constants were determined by using nonlinear regression with the LIGAND program to analyze displacement curves for the indicated labeled and unlabeled ligands. Assessments were performed at each dose in triplicate. Constants were calculated based on a single-site model.

positive bacteria (5) were also less active (Table 2) than the cell-binding domain. To further localize the Candida binding site in the cellbinding domain, four additional fragments of FN were tested in the competition assays. All of these fragments overlap the C-terminal end of the cell-binding domain and flank the RGD sequence located in the 10th type III repeat. A recombinant 40-kDa fragment (amino acids 1380 to 1851) had an inhibitory activity similar to that of p120 containing the cell-binding domain (Table 2). Two other RGD-containing fragments of the cell-binding domain lacking amino acids 1723 to 1851 (r33 and r28) were less active. A 33-kDa recombinant fragment which lacks the RGD sequence had inhibitory activity identical to that of r33 (amino acids 1329 to 1722), which contains the RGD sequence (Table 2). The LIGAND analysis of p120 and r40 binding data indicated that their dissociation constants differed by only threefold (Table 1). Although FN binding to many mammalian cells through its cell-binding domain is mediated by the RGD sequence, the role of the RGD sequence in the recognition of FN by C. albicans is controversial (13, 18, 19). We tested several RGD peptides for inhibition of FN binding to C. albicans induced by hemoglobin. Among the synthetic peptides used, the FN peptide GRGDS, and peptides GRGDNP, GdRGDSPAKK, and EIATRYRGDQDATMS demonstrated no inhibition. Peptides REDV and CS1 (DELPQLVTLPHPNLHGPEILDV PST), recognized by integrin a4b1 (23), also failed to inhibit FN binding to C. albicans (Table 2). Adhesion of C. albicans to immobilized FN and its fragments. Since hemoglobin also induced the adhesion of C. albicans to immobilized FN (24), we compared the adhesion of control and hemoglobin-induced cultures to the FN fragments. C. albicans cells grown with hemoglobin demonstrated a significantly greater adhesion to immobilized FN (P , 0.05 by two-tailed t test) than those grown in YNB medium alone (Fig. 3). Consistent with the results obtained from binding assays in solution, hemoglobin also promoted cell adhesion to the p120 and r40 fragments (P , 0.05 [compared to cultures without hemoglobin]) (Fig. 3). As reported previously (18), the 30-kDa fragment from the collagen-binding domain and the r31 fibrin I domain fragment of FN promoted the adhesion of C. albicans, but hemoglobin did not increase adhesion to these fragments. Adhesion to recombinant 12.5- and 18-kDa subfragments of the fibrin I domain was also not influenced by hemoglobin (results not shown). Adhesion of hemoglobin-induced cultures to the p120 and r40 fragments demonstrated a dose dependence similar to that of intact FN, whereas the collagen-binding p30 and fibrin I r31 fragments were less active (Fig. 4). Identification of surface proteins that recognize both FN and the r40 fragment. Preliminary experiments demonstrated

TABLE 2. Inhibition of radioactive FN binding to C. albicans grown with hemoglobin by FN, fragments, and synthetic peptides Inhibitor

FN p120 r40 r28 r33 (with RGD) r33 kDa (without RGD) p30 r31 r18.5 r12 GRGDS GRGDNP GdRGDSPAKK EIATRYRGDQDATMS CS1 (DELPQLVTLPHPNLH GPEILDVPST) REDV

FN domain

IC50 (nM)a

IIIcsb

32.5 210 320 1,290 3,800 3,900 .5,000 1,100 1,200 1,400 .200,000 .600,000 .349,000 .146,000 .330,000

IIIcs

.484,000

Cell Cell Cell Cell Cell Collagen Fibrin I Fibrin I Fibrin I Cell RGD analog RGD analog

a

Concentrations of peptides and proteolytic or recombinant fragments giving 50% inhibition of net FN binding to 2 3 106 C. albicans cells were determined from dose-dependent curves for the individual fragments or peptides in DPBS– 0.1% BSA, pH 6.0. Results are the averages from multiple experiments performed in triplicate. b IIIcs, type III connecting sequence.

that the binding of FN requires amino groups on the Candida cell surface receptors, since modification of amino groups by an hydroxysuccinimide ester prevented the binding of 125I-FN (25). Based on this observation, ligand protection from chemical modification was used to identify surface proteins that bind to FN and to the 40-kDa recombinant cell-binding domain. Candida cells were incubated with or without FN or the 40kDa recombinant fragment to allow binding to their receptors on the cell surface. After the removal of excess ligand, unprotected amino groups of surface proteins were modified by using sulfo-SHPP in the presence of bound ligand. The bound ligands were then removed by thorough washing, and the cells

FIG. 3. Growth of C. albicans in hemoglobin enhances adhesion to the cellbinding domain of FN. Candida cells prepared from cultures with (shaded bars) or without (solid bars) hemoglobin were allowed to adhere to FN or FN fragments applied to Chamber slides as described in Materials and Methods. The numbers of cells attached per square millimeter of surface were determined in triplicate and are presented as the means 6 SD. An asterisk indicates that the adhesion induced by hemoglobin significantly differs from that induced by the respective control, with P , 0.05 by a two-tailed t test.

HEMOGLOBIN PROMOTES C. ALBICANS BINDING TO FN

VOL. 66, 1998

FIG. 4. Dose dependence of cell adhesion to immobilized FN or its fragments. FN (ƒ) and FN fragments p120 (E), p30 (F), r31 (h), and r40 (Œ) were adsorbed at 0.01, 0.1, 1, and 10 mg/ml in DPBS without Ca21 and Mg21 (pH 7.5) to Chamber slides by incubation overnight at 4°C. Unbound protein was removed by washing the wells three times with DPBS. Candida cells prepared from cultures with hemoglobin in YNB medium were added to each well. Attached Candida cells were fixed, stained, and counted (n 5 3).

were biotinylated. Biotinylation under these conditions was limited to those amino groups that were protected by bound ligand. Cell surface proteins were extracted, and biotinylated proteins were visualized by streptavidin blotting (Fig. 5). When cells were prebound with FN or the r40 fragment derived from FN, protection by 100 mg of either ligand per ml resulted in prominent labeling of a 55-kDa protein (lanes a and c) that was only faintly visible in the cells without protection (lane b) and increased labeling of a 30-kDa protein. This specific labeling contrasted with the prominent band at 110 kDa and diffuse background bands that were equally labeled in the cells with or without ligands. DISCUSSION Previous work in this laboratory showed that growth of C. albicans in Sabouraud broth induces multiple classes of receptors for FN (18). A p30 fragment from the collagen-binding domain near the N terminus of FN had the highest affinity for binding to FN, which was similar to the affinity of native FN. However, fragments from the cell-binding domain were also active as inhibitors of radiolabeled-FN binding to C. albicans grown in Sabouraud broth. The biphasic Scatchard plot obtained under these conditions suggested that lower-affinity receptors may recognize other domains of FN, including the cell-binding domain. In this study, we show that the hemoglobin-induced binding of FN to cultures grown in YNB broth, a defined medium, involves one class of receptors that preferentially recognize the cell-binding domain relative to the collagen-binding or fibrin I domains. Therefore, hemoglobin appears to induce increased expression or activity of a class of receptors specific for the cell-binding domain of FN. Based on the ligand protection data, this recognition of the cell-binding domain may be mediated by 55- and/or 30-kDa proteins on the surface of C. albicans. Because these putative receptors were identified by ligand protection, we cannot determine whether the proteins are newly expressed on the surface or are simply activated to bind FN following growth with hemoglobin. The induced cultures also show an increased adhesion to immobilized FN and to endothelial cells (24), suggesting that such enhancement may have significance in the pathogenesis of systemic candidiasis. The present data demonstrate that this hemoglo-

1907

bin-induced increase in adhesion is also mediated by the cellbinding domain of FN, but not by the collagen-binding or fibrin I domain. FN is a dimeric protein and contains multiple ligand-binding domains. Recombinant and proteolytic fragments of FN provide powerful tools for locating the domain responsible for these interactions (18, 23). Unlike cells grown in Sabouraud medium, where the collagen-binding domain was recognized a higher affinity than that of the cell-binding domain, growth with hemoglobin in defined medium enhanced FN binding preferentially through the cell-binding domain. This finding is similar to the results of Penn and Klotz obtained with Sabouraud medium (19). As reported previously, however, FNbinding results with the complex Sabouraud medium vary from batch to batch and from laboratory to laboratory (13, 18, 24). Therefore, the binding data generated from cultures grown in complex medium may depend on multiple factors. The presence of multiple binding sites for FN in C. albicans may be characteristic of a successful pathogen, since the pathogenic bacterium Staphylococcus aureus also expresses receptors that interact with several domains of FN (5). Hemoglobin in defined medium consistently induces the expression of a specific receptor in C. albicans that interacts with the cell-binding domain of FN. Therefore, these defined growth conditions will provide a reproducible method to study the role of this domain in the interaction of C. albicans with the extracellular matrix. It has been proposed that C. albicans interactions with FN are mediated by an integrin-like molecule on the cell surface (14). Lack of a requirement for the RGD sequence in the cell-binding domain for interactions of FN with C. albicans is clearly demonstrated by the equal activities of the 33-kDa FN fragments with and without the RGD sequences. The inability of RGD peptides to influence FN binding supports this conclusion. Although the a5b1 integrin in mammalian cells requires the RGD sequence for FN binding, other FN-binding integrins such as a4b1 do not require this sequence (23). Several mammalian b1 integrins also promote RGD-independent adhesion to FN type III repeats (8). In addition, specific sequences flanking the RGD sequence, such as the PHSRN

FIG. 5. Identification of FN-binding proteins by ligand protection. C. albicans cells at late-exponential phase were preincubated without ligand (lane b), with the r40 fragment of FN (lane a), or with FN (lane c) at 100 mg/ml for 2 h, and exposed amino groups were blocked by reacting them with sulfo-SHPP at a concentration of 0.5 mg/ml in 50 mM sodium bicarbonate, pH 7.8, for 1 h with shaking. The bound FN was removed by thorough washing and lysyl residues of surface proteins that were masked by FN binding were then labeled by biotinylation. Cell surface proteins were then extracted and separated, and biotinylated proteins were visualized as described in Materials and Methods.

1908

YAN ET AL.

sequence in the ninth type III repeat, are required for highaffinity interactions of the a5b1 integrin with FN (1) and may also directly interact with the integrin (16). Activity of the 28-kDa recombinant fragment to inhibit FN binding to C. albicans localizes a primary recognition sequence in repeats 10 to 12, although this recognition does not require the RGD sequence in repeat 10. The higher level of activity of the r40 fragment relative to the 28-kDa fragment, however, suggests that a synergy site for C. albicans may be present in repeats 12 and 13, amino acid residues 1723 to 1851. This putative synergy site is different from that identified for a5b1 integrin binding (1), since the known synergy sequence is missing in the r28 fragment and the r33 fragment containing this sequence is less active. Folding of a given fragment of FN may also be important for recognition by C. albicans in addition to the possession of a specific peptide sequence. Alternatively, self-association of some of these fragments with the labeled intact FN may contribute to the apparent differences in their inhibitory activities (11, 15). Further work will be needed to identify the specific sequences responsible for hemoglobin-induced binding of FN to C. albicans. The receptors interacting with the cell-binding domain induced by hemoglobin may also be different from those induced by Sabouraud medium. In the complex medium, r28 (type III10-12) and p120 had equal potencies as inhibitors, whereas the r40 fragment showed very weak activity. In defined medium supplemented with hemoglobin, however, the r40 fragment is a stronger inhibitor than the r28 fragment. The reason for this discrepancy is not clear and may indicate the involvement of multiple receptors that are differentially induced by the different growth conditions. The role of RGD peptides in the interaction between FN and C. albicans has been controversial. Comparing epithelial cell adhesion by C. albicans and Candida tropicalis, Bendel et al. (3) found that purified FN or RGD peptides failed to block C. albicans adhesion but that the epithelial adhesion of C. tropicalis was significantly inhibited by these peptides. They concluded that the pathogenic yeasts C. albicans and C. tropicalis recognize distinct RGD ligands containing distinct flanking sequences (2, 9). Other studies demonstrated that radiolabeled FN binding was inhibited by RGD, GRGESP, and GRGDTP, but not by GRGDSP (13). A separate study found that the binding of the p120 cell-binding-domain fragment to C. albicans was inhibited by RGD and an RGD-containing 23-mer FN peptide (19). Using complex medium, however, we failed to demonstrate any significant inhibition by RGD peptides, including an iC3b peptide (18). Furthermore, the deletion of the RGD sequence from the recombinant cell-binding domain did not alter its activity either in defined or complex medium (18). The specific induction of FN binding to C. albicans by hemoglobin in defined medium provides a unique model to clarify the role of RGD peptides and the cell-binding domain in the recognition of FN. As we expected, none of the RGD peptides used inhibited FN binding to C. albicans, and the recombinant domain with the RGD sequence deleted had the same activity as the native sequence for inhibiting FN binding. The nature of the FN-binding receptors of C. albicans remains unclear. The putative receptor protein protected from chemical modification by both FN and the r40 fragment has a molecular mass lower than those of proteins that were identified as potential FN receptors in uninduced Candida cells by affinity chromatography on immobilized FN (14) or of the recently identified ALA1 adhesin (10). This difference could result from proteolytic degradation during extraction. Thus, the native molecular mass of the receptor identified with FN

INFECT. IMMUN.

and the r40 fragment may be larger than 55 kDa. Purification on an FN affinity column also identified the 55-kDa protein as a potential FN receptor (25). Further characterization of these surface proteins at the molecular level should be helpful in defining their functions. ACKNOWLEDGMENT We thank Tikva Vogel of Biotechnology General Ltd. for providing the recombinant constructs of fibronectin. REFERENCES 1. Aota, S., M. Nomizu, and K. M. Yamada. 1994. The short amino acid sequence Pro-His-Ser-Arg-Asn in human fibronectin enhances cell-adhesive function. J. Biol. Chem. 269:24756–24761. 2. Bendel, C. M., and M. K. Hostetter. 1993. Distinct mechanisms of epithelial adhesion for Candida albicans and Candida tropicalis. Identification of the participating ligands and development of inhibitory peptides. J. Clin. Invest. 92:1840–1849. 3. Bendel, C. M., J. St. Sauver, S. Carlson, and M. K. Hostetter. 1995. Epithelial adhesion in yeast species: correlation with surface expression of the integrin analog. J. Infect. Dis. 171:1660–1663. 4. Bouchara, J., G. Tronchin, V. Annaix, R. Robert, and J. Senet. 1990. Laminin receptors on Candida albicans germ tubes. Infect. Immun. 58:48–54. 5. Bozzini, S., L. Visa, P. Pignatti, T. Petersen, and P. Speziale. 1992. Multiple binding sites in fibronectin and the staphylococcal fibronectin receptor. Eur. J. Biochem. 207:327–333. 6. Calderone, R. A., and W. Scheld. 1987. Role of fibronectin in the pathogenesis of candida infections. Rev. Infect. Dis. 9:S400–S403. 7. Calderone, R. A. 1993. Molecular interactions at the interface of Candida albicans and host cells. Arch. Med. Res. 24:275–279. 8. Chi-Rosso, G., P. J. Gotwals, J. Yang, L. Ling, K. Jiang, B. Chao, D. P. Baker, L. C. Burkly, S. E. Fawell, and V. E. Koteliansky. 1997. Fibronectin type III repeats mediate RGD-independent adhesion and signaling through activated beta1 integrins. J. Biol. Chem. 272:31447–31452. 9. DeMuri, G. P., and M. K. Hostetter. 1996. Evidence for a beta 1 integrin fibronectin receptor in Candida tropicalis. J. Infect. Dis. 174:127–132. 10. Gaur, N. K., and S. A. Klotz. 1997. Expression, cloning, and characterization of a Candida albicans gene, ALA1, that confers adherence properties upon Saccharomyces cerevisiae for extracellular matrix proteins. Infect. Immun. 65: 5289–5294. 11. Hocking, D. C., R. K. Smith, and P. J. McKeown-Longo. 1996. A novel role for the integrin-binding III-10 module in fibronectin matrix assembly. J. Cell. Biol. 133:431–444. 12. Jakab, E., M. Paulsson, F. Ascecio, and A. Ljunch. 1992. Expression of vitronectin and fibronectin binding by Candida albicans yeast cells. APMIS 101:187–193. 13. Klotz, S. A., and R. Smith. 1991. A fibronectin receptor on Candida albicans mediates adherence of the fungus to extracellular matrix. J. Infect. Dis. 163: 604–610. 14. Klotz, S. A., R. C. Hein, R. L. Smith, and J. B. Rouse. 1994. The fibronectin adhesin of Candida albicans. Infect. Immun. 62:4679–4681. 15. Morla, A., and E. Ruoslahti. 1992. A fibronectin self-assembly site involved in fibronectin matrix assembly: reconstruction in a synthetic peptide. J. Cell. Biol. 118:421–429. 16. Mould, A. P., J. A. Askari, S. Aota, K. M. Yamada, A. Irie, Y. Takada, H. J. Mardon, and M. J. Humphries. 1997. Defining the topology of integrin alpha5beta1-fibronectin interactions using inhibitory anti-alpha5 and antibeta1 monoclonal antibodies. Evidence that the synergy sequence of fibronectin is recognized by the amino-terminal repeats of the alpha5 subunit. J. Biol. Chem. 272:17283–17292. 17. Munson, P. J., and D. Rodbard. 1980. Ligand: a versatile computerized approach for characterization of ligand-binding systems. Anal. Biochem. 107:220–239. 18. Negre, E., T. Vogel, A. Levanon, R. Guy, T. J. Walsh, and D. D. Roberts. 1994. The collagen binding domain of fibronectin contains a high affinity binding site for Candida albicans. J. Biol. Chem. 269:22039–22045. 19. Penn, C., and S. A. Klotz. 1994. Binding of plasma fibronectin to Candida albicans occurs through the cell binding domain. Microb. Pathog. 17:387–393. 20. Sela, S., A. Aviv, A. Tovi, I. Burstein, M. G. Caparon, and E. Hanski. 1993. Protein F: an adhesin of Streptococcus pyogenes binds fibronectin via two distinct domains. Mol. Microbiol. 10:1049–1055. 21. Thakur, A. K., P. J. Munson, D. L. Hunston, and D. Rodbard. 1980. Characterization of ligand-binding systems by continuous affinity distributions of arbitrary shape. Anal. Biochem. 103:240–254. 22. Vogel, T., M. M. Werber, R. Guy, A. Levanon, A. Nimrod, C. Legrand, M. Gorecki, A. Eldor, and A. Panet. 1993. Studies on fibronectin and its domains. I. Novel recombinant cell-binding domain of fibronectin—a

VOL. 66, 1998

HEMOGLOBIN PROMOTES C. ALBICANS BINDING TO FN

modulator of human platelet functions. Arch. Biochem. Biophys. 300:501– 509. 23. Yamada, K. 1991. Adhesive recognition sequences. J. Biol. Chem. 266: 12809–12812. 24. Yan, S., E. Ne`gre, J. A. Cashel, N. Guo, C. A. Lyman, T. J. Walsh, and D. D. Roberts. 1996. Specific induction of fibronectin binding activity by hemoglo-

Editor: T. R. Kozel

1909

bin in Candida albicans grown in defined media. Infect. Immun. 64:2930– 2935. 25. Yan, S., R. G. Rodrigues, D. Cahn-Hidalgo, T. J. Walsh, and D. D. Roberts. 1998. Hemoglobin induces binding of several extracellular matrix proteins to Candida albicans. Identification of a common receptor for fibronectin, fibrinogen and laminin. J. Biol. Chem. 273:5638–5644.