Mucosal Tissue Invasion by Candida albicans Is Associated with E-Cadherin Degradation, Mediated by Transcription Factor Rim101p and Protease Sap5p

Abstract

The ability of Candida albicans to invade mucosal tissues is a major virulence determinant of this organism; however, the mechanism of invasion is not understood in detail. Proteolytic breakdown of E-cadherin, the major protein in epithelial cell junctions, has been proposed as a mechanism of invasion of certain bacteria in the oral mucosa. The objectives of this study were (i) to assess whether C. albicans degrades E-cadherin expressed by oral epithelial cells in vitro; (ii) to compare the abilities of strains with different invasive potentials to degrade this protein; and (iii) to investigate fungal virulence factors responsible for E-cadherin degradation. We found that while E-cadherin gene expression was not altered, E-cadherin was proteolytically degraded during the interaction of oral epithelial cells with C. albicans. Moreover, C. albicans-mediated degradation of E-cadherin was completely inhibited in the presence of protease inhibitors. Using a three-dimensional model of the human oral mucosa, we found that E-cadherin was degraded in localized areas of tissue invasion by C. albicans. An invasion-deficient rim101⁻/rim101⁻ strain was deficient in degradation of E-cadherin, and this finding suggested that proteases may depend on Rim101p for expression. Indeed, reverse transcription-PCR data indicated that expression of the SAP4, SAP5, and SAP6 genes is severely reduced in the rim101⁻/rim101⁻ mutant. These SAP genes are functional Rim101p targets, because engineered expression of SAP5 in the rim101⁻/rim101⁻ strain restored E-cadherin degradation and invasion in the mucosal model. Our data support the hypothesis that there is a mechanism by which C. albicans invades mucosal tissues by promoting the proteolytic degradation of E-cadherin in epithelial adherens junctions.

Candida albicans employs virulence strategies that contribute to its ability to colonize and invade host cells and tissues and to cause common mucosal infections, such as oropharyngeal candidiasis in immunocompromised patients (36). These mechanisms include expression of host recognition molecules or adhesins (13, 42, 51), the ability to exist as reversible morphotypes (5), the ability to undergo phenotypic switching (29, 44), and expression of tissue invasion-facilitating enzymes, such as phospholipases and secreted aspartyl proteinases (12, 15, 22, 32, 40, 41).

At certain mucosal sites, such as the esophageal mucosa, demonstration of fungal invasion is required for definitive diagnosis of infection, since C. albicans is also a commensal colonizer of mucous membranes (27). Moreover, at these sites the extent of fungal invasion has been shown to correlate well with the severity of infection (1, 2). Fungal invasion of the superficial layers of the oral epithelium is found in human cases of advanced immunosuppression and in animal models of oropharyngeal candidiasis (11, 17, 18). Moreover, we have shown that the tissue invasion capacity of C. albicans correlates with its ability to stimulate a strong inflammatory response by oral mucosal cells (48). Although the role of invasion in the virulence of C. albicans has been demonstrated, the mechanism by which C. albicans invades the oroesophageal mucosa is not understood.

Epithelial cells are the first barrier against microbial mucosal invasion. Adhesion complexes, known as adherens junctions, contribute to the integrity of this barrier. An adherens junction is a specialized region of the plasma membranes of two adjacent cells in which cadherins act as adhesion molecules, linking together the actin cytoskeletons of the cells (47). Proteolytic breakdown of E-cadherin, the predominant protein in epithelial adherens junctions (47), has been proposed to be a mechanism of invasion of Bacteroides fragilis and Porphyromonas gingivalis in the intestinal mucosa and the oral mucosa, respectively (19, 20, 35, 50). Although studies have shown that C. albicans is able to invade oral epithelial cells intracellularly by inducing its own endocytosis (37), its effect on the oral epithelial intercellular junctions is not known.

In this study we hypothesized that one mechanism used by C. albicans to invade the oroesophageal mucosa is to degrade E-cadherin in epithelial adherens junctions. To investigate this hypothesis, we analyzed the ability of C. albicans to degrade E-cadherin expressed by oral epithelial cells in vitro. Moreover, we compared the abilities of strains with different invasive potentials to degrade this protein. We found that E-cadherin from epithelial adherens junctions is degraded during coculture with C. albicans and that the invasive capacities of C. albicans strains are commensurate with their capacities to degrade E-cadherin in vitro. Furthermore, we obtained evidence that Sap5p is responsible for E-cadherin degradation in vitro.

MATERIALS AND METHODS

Organisms.

The C. albicans strains used in this study included SC5314 and its derivatives BWP17 (49), CAI4 (4), and VIC18 (8) (Table 1). Clinical strain SC5314, which was originally isolated from a patient with invasive disseminated candidiasis (16) and has a strong invasive phenotype in several oral mucosal models (48), was used in some experiments to study E-cadherin degradation. A homozygous deletion of the RIM101 gene from C. albicans has been shown to severely curtail the ability of this organism to invade tissues both in vivo and in vitro (7, 48). Therefore, in the present study we used a RIM101 gene knockout strain (DAY25) to represent the reduced invasive phenotype of the microorganism. RIM101-complemented strain DAY44 and reference strain DAY185, which had a strong invasive phenotype in vivo and in vitro (7, 48), were included as control strains.

TABLE 1.

C. albicans strains used in this study

Strain	Genotype	Reference
DAY25	ura3Δ::λimm434/ura3Δ::λimm434 his1::hisG::pHIS1/his1::hisG arg4::hisG/arg4::hisG rim101::URA3/rim101::ARG4	7
DAY44	ura3Δ::λimm434/ura3Δ::λimm434 his1::hisG::pHIS1::RIM101/his1::hisG arg4::hisG/arg4::hisG rim101::ARG4/rim101::URA3	7
DAY185	ura3Δ::λimm434/ura3Δ::λimm434 his1::hisG::pHIS1/his1::hisG arg4::hisG::pARG4::URA3/arg4::hisG	7
CJN759	ura3Δ::λimm434/ura3Δ::λimm434 arg4::hisG/arg4::hisG his1::hisG/his1::hisG rim101::dpl200/rim101::dpl200 nrg1::ARG4/nrg1::URA3	This study
CJN793	ura3Δ::λimm434/ura3Δ::λimm434 arg4::hisG/arg4::hisG his1::hisG::pHIS1-NRG1/his1::hisG rim101::dpl200/rim101::dpl200 nrg1::ARG4/nrg1::URA3	This study
CJN1111	ura3Δ::λimm434/ura3Δ::λimm434 arg4::hisG/arg4::hisG his1::hisG::pHIS1-NRG1/his1::hisG rim101::dpl200/rim101::dpl200 nrg1::ARG4/nrg1::URA3 TEF1-SAP5::NAT1/SAP5	This study
DSY452	sap5Δ::hisG/sap5Δ::hisG-URA3-hisG	39
DSY459	sap6Δ::hisG/sap6Δ::hisG sap4Δ::hisG/sap4Δ::hisG sap5Δ::hisG/sap5Δ::hisG-URA3-hisG	39
CAI4	ura3Δ::λimm434/ura3Δ::λimm434	4
BWP17	ura3Δ::λimm434/ura3Δ::λimm434 his1::hisG/his1::hisG arg4::hisG/arg4::hisG	49
VIC18	ura3Δ::λimm434/ura3Δ::λimm434 his1::hisG/his1::hisG arg4::hisG/arg4::hisG	8
SC5314	Clinical isolate	12

A rim101⁻/rim101⁻ strain expressing SAP5 (CJN1111) was used to investigate the role of the Sap5p secreted aspartyl proteinase (regulated by the Rim101p transcription factor [Table 2 ]) in E-cadherin degradation, using an otherwise isogenic rim101⁻/rim101⁻ strain (CJN793) as a control. Strain CJN759 was used to create CJN793. CJN759 was constructed by PCR product-directed gene deletion (49) with 100-mer oligonucleotides NRG1null-5DR (5′-CTAGGATTTGTTATTACAATTTGCTTTACAAGAGAACCTCATTGT TCAAGTATCTTCCAGAATCTGAAACAGGTATTATATATTTCCCAGTCA CGACGTT-3′) and NRG1null-3DR (5′-GTAGCACCAAAAGGGGTAGAGACTAGGCAGTACTATATTTGGAGGTAAAAGTATATTAGAAAAAGAAC CTATACATGAAC CGTGGAATTGTGAGCGGATA-3′) in C. albicans strain VIC18 (8). The NRG1 complementing plasmid was constructed using PCR to produce a fragment for NRG1 from 1,000 bp upstream of the ATG to 500 bp downstream of the stop codon. This fragment was inserted into the pGEMT-Easy vector (Promega), digested with NotI to excise the fragment, and then ligated into the NotI-digested and alkaline phosphatase-treated HIS1 vector pDDB78 (45) to obtain plasmid pCJN104. Strain CJN793 was constructed by transforming CJN759 with NruI-digested plasmid pCJN104; the unique NruI site directed integration to the HIS1 locus. The NAT1-TEF1-SAP5-expressing C. albicans strain CJN1111 was constructed by transforming CJN793 using PCR products obtained with template plasmid pCJN498 (34) and primers SAP5-F-OE-Ag-NAT-Ag-TEF1p (5′-TATGTGGAAAGAAAAAGCGTTCTGAACAATTTCGATTTCAATGCTGAACATCATAACCATCATCAACATTTTTAAGACACCAAGGTATGCATCAAGCTTGCCTCGTCCCC-3′) and SAP5-R-OE-Ag-NAT-Ag-TEF1p (5′-ATTAAAGTCTAAGGTAACAAACCCTGGAGATCTTTTAACTGGAGCAGCATCATTATAAAGCAAAAGCAAGAACACTCAAGATATTTTTCAAAACATTATAAATTTACTTAGAA-3′), as described previously (34).

TABLE 2.

Analysis of expression of SAP4, SAP5, and SAP6 genes by quantitative RT-PCR: comparison of SAP4, SAP5, and SAP6 gene expression in rim101⁻/rim101⁻ mutant strain CJN793, RIM101 gene knockout strain CJN1111 expressing SAP5, and the reference strain DAY185^a

Strain	Relevant genotype	Transcript	Ct (mean ± SD)	RNA level (% of wild type)
DAY185	RIM101/RIM101	TEF1	32.25 ± 2.41	Control
		SAP6	6.25 ± 0.73	100
		SAP5	5.44 ± 2.25	100
		SAP4	8.76 ± 1.41	100
CJN793	rim101/rim101	TEF1	20.93 ± 1.34	Control
		SAP6	32.72 ± 1.34	4.2 × 10−10
		SAP5	32.56 ± 3.11	2.7 × 10−10
		SAP4	40.78 ± 0.50	9.0 × 10−12
CJN1111	rim101/rim101 PTEF1-SAP5	TEF1	22.84 ± 1.47	Control
		SAP6	35.25 ± 1.80	2.7 × 10−10
		SAP5	15.95 ± 2.11	1.0 × 10−4
		SAP4	45.77 ± 2.95	1.1 × 10−12

^aThe relative RNA levels were calculated using the ΔΔC_t method. All primers resulted in amplification efficiencies of at least 95%.

A SAP5 gene knockout strain (strain DSY 452 provided by B. Hube [39]) and a SAP4-SAP5-SAP6 gene knockout strain (strain DSY 459 provided by B. Hube [39]) were used to investigate the role of SAP genes in E-cadherin degradation, using the otherwise isogenic strain CAI4 (4) as a control.

All of the strains used in this study had similar growth rates in keratinocyte medium, as determined by direct counting of yeast cells or by the XTT assay (23) when filamentous organisms were tested (not shown). In addition, all strains formed true hyphae when they were cocultured with oral epithelial cells. The organisms were routinely propagated on YPD agar (Difco Laboratories, Detroit, MI) at 25°C.

Cell cultures.

The epithelial cells used included a well-differentiated squamous cell carcinoma of the ventral tongue (cell line SCC15) (25) known to express high levels of E-cadherin (14) and were obtained from D. Wong (University of California at Los Angeles). SCC15 cells were maintained in keratinocyte serum-free medium (KSFM) (Invitrogen, Carlsbad, CA) supplemented with 0.4 mM CaCl₂, 0.1 ng/ml epidermal growth factor, 50 μg/ml bovine pituitary extract (Invitrogen, Carlsbad, CA), and antibiotics (100 U/ml penicillin and 100 μg/ml streptomycin).

Coculture of C. albicans with epithelial cell monolayers.

To study the effect of C. albicans on E-cadherin protein and gene expression, SCC15 cells were seeded to near confluence on six-well polystyrene plates (4 × 10⁵ cells/well) or polystyrene culture dishes (Costar, Cambridge, MA) (4 × 10⁶ cells/culture dish). Cell monolayers were incubated in complete KSFM at 37°C in a 5% CO₂ atmosphere for up to 3 days with daily changes of the medium. Stationary-phase yeasts were prepared by growth for 18 h at room temperature in YPD broth (Difco Laboratories, Detroit, MI) supplemented with 2% (wt/vol) dextrose. The fungal cells were harvested by centrifugation, washed in phosphate-buffered saline (PBS), and counted with a hemacytometer, and the final concentration in complete KSFM was prepared before the fungal cells were added to epithelial cells. At the end of the 3-day epithelial cell culture, yeast cells were suspended in fresh KFSM at a concentration of 8 × 10⁵ cells/ml, and then 1 and 10 ml of a yeast suspension were added to cultures in six-well plates and dishes, respectively, and incubated for up to 14 h.

In order to inhibit proteolytic activity, some experiments were carried out in the presence or absence of a cocktail of protease inhibitors (Sigma, St. Louis, MO). The protease inhibitor cocktail included a mixture of aprotinin, bestatin, leupeptin, E-64, and pepstatin A and was used at a dilution of 1:200 as a supplement in KSFM throughout the infection period. The negative controls for these experiments included epithelial cell cultures and C. albicans alone. At the end of the experiment, cell lysates were prepared for RNA and protein assays and stored at −80°C until they were used.

To prepare cell lysates for protein analyses, cultures were rinsed once in cold PBS and exposed to 0.5 ml of radioimmunoprecipitation assay buffer (20 mM Tris-HCl [pH 7.4], 150 mM NaCl, 0.1% sodium dodecyl sulfate, 1% Triton X-100, 1% deoxycholic acid, 5 mM EDTA) containing inhibitors of proteases (2 mM phenylmethylsulfonyl fluoride [Sigma, St. Louis, MO], 25 μg of pepstatin [Sigma, St. Louis, MO], 25 μg of chymostatin [Sigma, St. Louis, MO], 5 μg of leupeptin [Sigma, St. Louis, MO], and 5 μg of antipain [Sigma, St. Louis, MO]) for 25 min on ice. Cells were scraped from the dishes with a rubber policeman, and the cell lysates were sedimented by centrifugation at 4°C in a microcentrifuge at the maximal speed for 10 min.

Epithelial cell-free E-cadherin degradation assays.

In order to directly assess the ability of C. albicans to degrade E-cadherin, we used an epithelial cell-free assay, in which cell membranes containing E-cadherin were fractionated from epithelial cells and were incubated with C. albicans. To prepare cell membranes, we used the protocol of Nagamatsu et al. (31). Briefly, SCC15 cells were seeded to near confluence on polystyrene culture dishes (Costar, Cambridge, MA) (4 × 10⁶ cells/culture dish) and were incubated in complete KSFM at 37°C in a 5% CO₂ atmosphere for up to 3 days. The cells were then homogenized in 1 ml of homogenization buffer (10 mM Tris-HCl [pH 7.4], 1 mM EDTA, 200 mM sucrose, 1 mM phenylmethylsulfonyl fluoride), and the nuclei and cell debris were removed from the homogenate by centrifugation at 1,000 × g for 10 min at 4°C. The resulting supernatant was centrifuged at 110,000 × g for 75 min at 4°C, and the pellet was solubilized in buffer (10 mM Tris-HCl [pH 7.4], 1 mM EDTA, 0.5% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, 1 μg/ml aprotinin) for 1 h at 4°C. The insoluble material was removed by centrifugation at 14,000 × g for 10 min at 4°C, and the resulting supernatant was stored at −80°C until it was assayed. In some experiments the ability of C. albicans to degrade E-cadherin was assessed by incubating C. albicans with immunoprecipitated E-cadherin. E-cadherin was purified from MDCK or SCC15 cells as previously described (19, 20).

Immunoprecipitated E-cadherin or cell membrane preparations were exposed to live C. albicans (1.5 × 10⁵ yeast cells) or C. albicans-conditioned medium at 37°C for up to 18 h. To prepare C. albicans-conditioned medium, yeast-phase organisms were inoculated into complete KSFM and allowed to form hyphae at 37°C for 18 h. Subsequently, the organisms were removed by centrifugation, and the medium was filter sterilized. In order to inhibit proteolytic activity, some experiments were carried out in the presence or absence of a cocktail of protease inhibitors (used at a dilution of 1:200) or 1.0 μM pepstatin A (a Sap inhibitor [41]; Sigma, St. Louis, MO). The positive and negative controls consisted of incubation with 1× trypsin-EDTA and KSFM, respectively.

Western blot analyses.

Samples were electrophoresed in sodium dodecyl sulfate-8% polyacrylamide gel electrophoresis minigels and transferred to Immuno-Blot polyvinylidene difluoride membranes (Bio-Rad, Hercules, CA). The membranes were blocked overnight at 4°C with Tris-buffered saline containing 5% milk and 0.1% Tween 20 (block buffer). The following day the membranes were probed with anti-E-cadherin monoclonal antibody (1:2,500; clone 36; BD Biosciences, San Diego, CA) or anti-β-actin monoclonal antibody (1:5,000; Sigma, St Louis, MO) for 2 h. The membranes were then washed three times (20 min each) with Tris-buffered saline containing 0.1% Tween 20 (wash solution) and probed with horseradish peroxidase-labeled goat anti-mouse antibody (1:20,000) for 60 min. Signals were developed by addition of an enhanced chemiluminescence-type solution (Pierce, Rockford, IL) and exposure of membranes to X-ray film (Pierce, Rockford, IL) for up to 1 min. Signals in the film were scanned and stored as TIF images, and their intensities were determined by densitometry analysis using the Chemilmager software (Alpha Innotech Corporation, San Leandro, CA). β-Actin was used as a loading control in all experiments with epithelial cells.

Real-time RT-PCR for E-cadherin expression.

Total RNA was extracted from epithelial cells using Tri-reagent (Molecular Research Center, Cincinnati, OH) according to the manufacturer's instructions. Expression of E-cadherin transcripts by SCC15 monolayers was assessed quantitatively by real-time reverse transcription (RT)-PCR, and data were analyzed using the iCycle iQ system software (Bio-Rad, Hercules, CA). Briefly, 3 μg total RNA of each sample was reverse transcribed into cDNA using random primer oligonucleotides (Invitrogen, Carlsbad, CA) and BD sprint PowerScript (BD Biosciences, Palo Alto, CA). Real-time PCR was performed on 96-well optical reaction plates (Bio-Rad, Hercules, CA). All PCR mixtures contained 7.5 μl of iQ SYBR green Supermix (Bio-Rad, Hercules, CA), 0.5 μl of forward primer (10 μM), 0.5 μl of reverse primer (10 μM), 4 μl of each diluted RT product, and 2.5 μl of distilled water per well. PCR amplification of the cyclophilin housekeeping gene was performed to control for sample loading and to normalize samples. The sequences of the primers used for real-time PCR were as follows: human E-cadherin forward primer, 5′-CAGCATCACTGGCCAAGGAGCTGA-3′; human E-cadherin reverse primer, 5′-GACCACACTGATGACTCCTGTGTTCC-3′; human cyclophilin forward primer, 5′-CGGGTCCTGGCATCTTGT-3′; and human cyclophilin reverse primer, 5′-GCAGATGAAAAACTGGGAACCA-3′. The negative controls included water and RNA extracted from C. albicans incubated under the same conditions in the absence of oral epithelial cells. The PCR program consisted of 95°C for 3 min, followed by 50 cycles of 30 s at 95°C, 30 s at 60°C, and 30 s at 72°C. Real-time measurements of E-cadherin expression were normalized using the expression of the cyclophilin reference gene, which resulted in the normalized amount of E-cadherin in each sample. Data were expressed as fold induction values obtained by dividing the normalized expression of E-cadherin in epithelial cells cocultured with C. albicans by the normalized expression of E-cadherin in epithelial cells alone. Only changes equal to or greater than 2-fold and equal to or less than 0.5-fold were considered to represent up- and down-regulation, respectively.

Real-time RT-PCR for C. albicans gene expression.

To ensure that SAP5 expression with a constitutive promoter did not result in transcription of far more SAP5 than was transcribed from the wild-type promoter in the reference strain, SAP expression in the strains was analyzed by RT-PCR. Briefly, overnight cultures were inoculated into 5 ml of YPD broth at 30°C. The following day, 100 ml of RPMI medium was inoculated with the YPD broth overnight culture to obtain a starting optical density at 600 nm of 0.05 and was grown at 37°C for 6 h, which resulted in an optical density at 600 nm of 1.0. Cells were harvested by vacuum filtration, and RNA was extracted as previously described (8). For real time RT-PCR detection of SAP4, SAP5, and SAP6 transcripts, 10 μg of total RNA was treated with DNase at 37°C for 1 h, precipitated with ethanol, and resuspended in 100 μl of nuclease-free water. cDNA was synthesized using the StrataScript RT-PCR system with oligo(dT) primers, with reverse transcriptase and without reverse transcriptase (as a control). SAP4, SAP5, and SAP6 transcript levels were determined and quantitatively assessed using a Bio-Rad iQ5 cycler and the iQ5 software, respectively. The following primers were used: SAP4F343 (5′-GTTCCAGATTCAAATGCCG-3′) and SAP4R1193 (5′-CTTGAGCCATGGAGATCTTTC-3′); SAP5F195 (5′-TGAGACTGGTAGAGATGGTG-3′) and SAP5R1075 (5′-GGTTTACCACTAGTGTAATAT-3′); SAP6F614 (5′-AAACCAACGAAGCTACCAGAA-3′) and SAP6R1197 (5′-TAACTTGAGCCATGGAGATTTT-3′); and RTTEF1F (5′-CAGATTTGAAGAAATCATCAAGGAAAC-3′) and RTTEF1R (5′-TTAGCAGCTTTTTGAGCAGCC-3′). Each PCR mixture (final volume, 50 μl) contained 1 μl of cDNA (at five dilutions to obtain an ideal standard curve), 25 μl of 2× SYBR green Supermix (Bio-Rad), 1 μl of forward primer (10 μM), 1 μl of reverse primer (10 μM), and 22 μl of nuclease-free water. The cycling conditions used were 50°C for 2 min, 95°C for 10 min, and 40 cycles of 95°C for 15 s, 55°C for 1 min, and 72°C for 30 s. Next, the samples were cooled to 55°C, and a melting curve for temperatures between 55 and 95°C with 0.5°C increments was recorded. Real-time expression measurements were normalized against expression of the reference gene TEF1. Relative RNA levels were calculated using the ΔΔC_t method (26); all primers resulted in amplification efficiencies of at least 95%.

Three-dimensional model of the human oral mucosa.

To investigate E-cadherin degradation during oral tissue invasion by C. albicans, we used a three-dimensional model of the oral mucosa described previously (9, 10, 48). This system is composed of gingival fibroblasts embedded in a biomatrix of collagen type I, overlaid by a multilayer of oral epithelial cells. To study the effect of C. albicans invasion on E-cadherin integrity in this system, the three-dimensional model of the oral mucosa was challenged with 1 × 10⁵ C. albicans yeast cells in 100 μl of airlift medium (KSFM containing 5% fetal bovine serum, 1.88 mM CaCl₂, and 0.025 mM dextrose) added to the apical surface of the cells. Seventeen to 48 h later, the cultures were fixed with 10% formaldehyde-PBS and embedded in paraffin. Formalin-fixed paraffin-embedded sections of three-dimensional epithelial cultures were heated at 96°C for 30 min with target retrieval solution (DakoCytomation, Carpinteria, CA) for antigen unmasking. After blocking, sections were incubated for 1 h with monoclonal antibodies against E-cadherin (1:300; clone 36; BD Biosciences, San Diego, CA) or cytokeratin (1:100; clone AE1/AE3; Dako) and then for 30 min with secondary antibody and avidin-biotin complex (VectaStain ABC peroxidase kit; Vector, Burlingame, CA). The appropriate isotype control antibodies were used to confirm the specificity of each antibody. Stained sections were examined with a Leica DM RB microscope connected to a digital camera.

Confocal laser scanning microscopy.

To confirm the intercellular presence of C. albicans during invasion of the three-dimensional model of the oral mucosa, formalin-fixed paraffin-embedded sections of three-dimensional epithelial cultures were analyzed by confocal microscopy. Briefly, after blocking with goat normal serum (Jackson ImmunoResearch Laboratories, Inc.) at a 20-fold dilution for 30 min, sections were incubated for 1 h with rabbit polyclonal fluorescein isothiocyanate-conjugated antibody against C. albicans (1:50; Biodesign, Maine). Next, the tissues were stained for 40 min with tetramethyl rhodamine isocyante-conjugated phalloidin (0.1 μg/ml; Sigma, Steinheim, Germany) to detect actin microfilaments. For nuclear staining, sections were stained with Hoechst 33258 (0.1 μg/ml; Invitrogen, Oregon) for 20 min. Stained sections were washed in PBS and mounted in fade-retarding mountant (Biomeda, California). Sections were viewed and analyzed by confocal laser scanning microscopy using a Zeiss LSM510 confocal microscope. The final confocal images were produced by combining multichannel recordings of optical sections taken through the z axis.

Transmission electron microscopy.

Transmission electron microscopy analyses were performed to confirm the intercellular presence of C. albicans during challenge of confluent cell monolayers. For these experiments 24-transwell polystyrene plates (Corning Incorporated, Corning, NY) were used. The upper compartment of each transwell plate was seeded at a level that was slightly above the level required for confluence (5 × 10⁴ cells/well), and epithelial cells were allowed to adhere overnight at 37°C in the presence of 5% CO₂. To confirm the integrity of the monolayer and to determine its permeability to macromolecules in vitro, the dextran permeability assay was used, as previously described (3). The negative and positive controls included membranes without cells, membranes with subconfluent SCC15 cells, and membranes with confluent MDCK cell (ATCC CCL-34) monolayers. MDCK cells are known to form tight junctions, which are relatively impermeable to dextran, under these in vitro conditions. Stationary-phase viable C. albicans cells (1 × 10⁴ yeast cells suspended in 0.1 ml of KSFM) were added to the upper compartment. After 4 h of incubation, the transwell insert was removed. The cultures were washed with cold PBS and treated with fresh 0.5× Karnovsky's fixative. The cultures were postfixed in 1% osmium tetroxide and 1.25% potassium ferrocyanide and then washed with two changes of phosphate buffer for 30 min. The cultures were then dehydrated using graded concentrations of ethanol, embedded in LR White (Electron Microscopy Sciences), and polymerized for 24 h at 60°C. Blocks were cut using an ultramicrotome with a diamond knife. The resultant sections were collected on grids and stained with 2% uranyl acetate and lead citrate. Sections were examined and photographed with a transmission electron microscope (Philips CM10: Philips, Eindhoven, The Netherlands).

RESULTS

E-cadherin protein levels decreased during the interaction of oral epithelial cells with C. albicans.

To test the hypothesis that C. albicans degrades E-cadherin in oral epithelial cell junctions, oral epithelial cell-fungal cell cocultures were set up, and the E-cadherin protein levels were monitored over time by Western blotting. Kinetic experiments showed that the level of E-cadherin, but not the level of β-actin protein, was reduced in a time-dependent manner during the interaction with the fungus (Fig. 1). A reduction in the E-cadherin signal compared to the signal obtained with uninfected cells was apparent as early as 4 h after fungal challenge (Fig. 1, lanes 3 and 4). At this time we also observed that C. albicans had gained access to the intercellular spaces of the confluent monolayer (Fig. 2). Moreover, no bands corresponding to either E-cadherin or β-actin were detected in cell lysates of C. albicans alone (Fig. 1, lane 7).

FIG. 1.

Time-dependent degradation of E-cadherin during incubation of oral epithelial cells with C. albicans. SCC15 cells were exposed to C. albicans strain SC5314 for 2 h (lane 2), 4 h (lane 4), 6 h (lane 6), 8 h (lane 9), 10 h (lane 11), 12 h (lane 13), and 14 h (lane 15), and E-cadherin expression was analyzed by Western blotting. Membranes were probed with β-actin to control for sample loading. Cell lysates of C. albicans alone were analyzed to determine E-cadherin and β-actin contents (lane 7). Densitometry values (in optical density [OD] units) are indicated below the bands. The positions of the molecular mass markers are indicated on the left. The results of one of three independent experiments are shown in each panel.

FIG. 2.

Intercellular location of C. albicans in oral epithelial cell monolayers. Oral epithelial cell monolayers were incubated with C. albicans strain SC5314 for 4 h and examined by transmission electron microscopy. C. albicans (black arrow) gained access to the intercellular spaces of the confluent monolayer. Oral epithelial cells are indicated by white arrows. Bar = 1 μm.

To examine the possibility that the reduction in E-cadherin protein levels was due to a decrease in gene expression, E-cadherin gene expression was monitored during fungal challenge by quantitative RT-PCR. As shown in Fig. 3, the magnitude of the E-cadherin gene expression changes probably did not account for the reduction in the protein levels that were observed. This suggested that the reduced E-cadherin protein levels observed in epithelial cells cocultured with C. albicans were triggered by a mechanism other than reduced gene expression.

FIG. 3.

Analysis of E-cadherin gene expression during incubation of oral epithelial cells with C. albicans. SCC15 cells were exposed to C. albicans strain SC5314 for up to 14 h, and E-cadherin gene expression was analyzed by quantitative RT-PCR. Data were expressed as fold induction values obtained by dividing the relative expression value for E-cadherin in epithelial cells cocultured with C. albicans by the relative expression value for E-cadherin in epithelial cells alone. A fold induction value of 1.0 indicates that there was no induction of E-cadherin expression in epithelial cells by C. albicans, while values less than 0.5 and greater that 2.0 indicate that there was decreased and increased E-cadherin expression in epithelial cells cocultured with C. albicans. The error bars indicate one standard deviation of the mean of triplicate experiments.

E-cadherin was proteolytically degraded by C. albicans.

In order to obtain evidence that E-cadherin was proteolytically degraded by C. albicans, oral epithelial cell monolayers were cocultured with strain SC5314 in the presence or absence of a cocktail of protease inhibitors. While challenge of SCC15 cell monolayers with C. albicans strain SC5314 resulted in almost complete degradation of E-cadherin, degradation was completely inhibited when epithelial monolayers were challenged in the presence of protease inhibitors (Fig. 4A).

FIG. 4.

E-cadherin is proteolytically degraded by C. albicans. (A) Effect of inhibitors of proteinases on E-cadherin degradation during incubation of SCC15 cells with C. albicans strain SC5314. (B) Effect of C. albicans strain SC5314 and C. albicans-conditioned medium on the integrity of E-cadherin expressed on epithelial cell-derived membranes. Trypsin was used as a positive control. (C) Effect of C. albicans strain SC5314 on the integrity of immunoprecipitated E-cadherin in the presence or absence of a cocktail of protease inhibitors or pepstatin A. The protease inhibitor cocktail contained aprotinin, bestatin, leupeptin, E-64, and pepstatin A. E-cadherin integrity was analyzed by Western blotting. The positions of the molecular mass markers are indicated on the left. The results of one of three independent experiments are shown in each panel.

Because degradation of E-cadherin could theoretically be attributed to oral epithelial cell-derived proteases, E-cadherin degradation by C. albicans was also tested in an epithelial cell-free environment. In order to do this, oral epithelial cell-derived membranes containing intact E-cadherin were incubated with strain SC5314 or keratinocyte medium conditioned by this strain. We observed that while partial degradation of E-cadherin occurred after treatment of membranes with C. albicans-conditioned medium, complete degradation of E-cadherin was observed after incubation with C. albicans for 18 h (Fig. 4B). To eliminate the possibility that epithelial membrane-associated proteases mediated E-cadherin degradation, the experiments were repeated using immunoprecipitated E-cadherin, and similar results were obtained (Fig. 4C). Moreover, C. albicans-mediated degradation of immunoprecipitated E-cadherin was completely inhibited in the presence of a cocktail of protease inhibitors or pepstatin A, an inhibitor of C. albicans-derived secreted aspartyl proteinases (41) (Fig. 4C). Together, these results indicate that a protease (or proteases) secreted by C. albicans is responsible for E-cadherin degradation.

E-cadherin was degraded during C. albicans invasion in a three-dimensional model of the human oral mucosa.

Since E-cadherin degradation was accompanied by the presence of C. albicans in the intercellular spaces of oral epithelial cell monolayers (Fig. 2), we hypothesized that (i) C. albicans invades the oral mucosa by gaining access into the intercellular spaces of oral epithelial cells and (ii) this invasion of the intercellular spaces in the oral mucosa may be associated with E-cadherin degradation. To investigate this hypothesis, we used a three-dimensional tissue model of the human oral mucosa and submucosa that was recently developed in our laboratory (9, 10, 48). Confocal analysis of this tissue infection model showed that C. albicans invaded the three-dimensional model of the human oral mucosa both directly by entry into keratinocytes, as indicated by the colocalization of C. albicans with actin filaments (fluorochrome codetection, indicated by yellow fungi [Fig. 5]) and by gaining access to intercellular spaces (green fungi [Fig. 5]). When E-cadherin was analyzed during this process by immunohistochemistry, it was found to be degraded only in localized areas where there was tissue invasion by C. albicans, despite the presence of a biofilm completely covering the three-dimensional surface culture (Fig. 6B). In contrast, cytokeratin remained intact in these sites where there was localized invasion (not shown).

FIG. 5.

Intercellular location of C. albicans in the three-dimensional model of oral mucosa. The oral mucosal model was challenged with C. albicans strain SC5314 for 8 h. Paraffin sections (thickness, 5 μm) of the three-dimensional culture were stained with fluorescein isothiocyanate-conjugated antibody against C. albicans, tetramethyl rhodamine isocyante-conjugated phalloidin, and Hoechst 33258 and analyzed by confocal laser scanning microscopy. C. albicans invaded the three-dimensional model of the human oral mucosa both directly by entry into keratinocytes (keratinocytes are red, indicated by yellow fungi [white arrows]) and by gaining access to intercellular spaces (indicated by green fungi [blue arrows]). Bar = 20 μm.

FIG. 6.

Immunohistochemistry analysis of E-cadherin expression in a three-dimensional model of oral mucosa. (A) Paraffin section (thickness, 5 μm) of three-dimensional culture alone stained with periodic acid-Schiff stain and antibody specific for E-cadherin (brown). (B) Paraffin section of three-dimensional culture incubated with DAY185 and stained with periodic acid-Schiff stain and antibody specific for E-cadherin (brown). (C) Paraffin section of three-dimensional culture incubated with invasion-deficient rim101⁻/rim101⁻mutant CJN793 and stained with periodic acid-Schiff stain and antibody specific for E-cadherin (brown). (D) Paraffin section of three-dimensional culture incubated with the rim101⁻/rim101⁻ mutant strain expressing SAP5 (strain CJN1111) and stained with periodic acid-Schiff stain and antibody specific for E-cadherin (brown). Sections were counterstained with hematoxylin and eosin (blue). The oral mucosal model was challenged with C. albicans added on the apical surface. Bars = 65 μm.

In order to confirm that E-cadherin degradation is a mechanism that C. albicans uses to invade the oral mucosa, we tested the integrity of E-cadherin in three-dimensional tissue cultures infected with an invasion-deficient rim101⁻/rim101⁻ mutant (CJN793) for up to 48 h. As previously reported (48), invasion-deficient rim101⁻/rim101⁻ mutant CJN793 was not able to invade the mucosa and submucosa vertically. Although prolonged incubation of three-dimensional cultures with this mutant strain resulted in biofilm formation in the outer layer of the epithelium, intact E-cadherin was found in all epithelial cell layers (Fig. 6C).

Invasive phenotype of C. albicans is associated with its ability to induce Sap5p-mediated E-cadherin degradation.

To confirm the finding that E-cadherin was degraded in the three-dimensional model, we compared the abilities of the invasion-deficient rim101⁻/rim101⁻ mutant DAY25, the congenic RIM101 complemented strain DAY44, and reference strain DAY185 to degrade immunoprecipitated E-cadherin. As expected, strain DAY25 was extremely deficient for E-cadherin degradation after 4 and 6 h of incubation (Fig. 7, lanes 2 and 6) compared to the complemented strain (Fig. 7, lane 7) and the reference strain (Fig. 7, lane 8), which resulted in almost complete degradation after 6 h of incubation.

FIG. 7.

Effect of C. albicans invasion-deficient mutant DAY25, complemented strain DAY44, and reference strain DAY185 on the integrity of E-cadherin. Immunoprecipitated E-cadherin was incubated with C. albicans for 4 and 6 h, and the protein integrity was analyzed by Western blotting. The positions of the molecular mass markers are indicated on the left. The results of one of three independent experiments in which similar data were obtained are shown.

Secreted aspartyl proteinases have been shown to play an important role in the penetration of C. albicans into host tissues (6, 30). Therefore, we investigated whether the inability of the rim101 mutant strain to degrade E-cadherin was due to down-regulation of these proteinases. We observed a severe reduction in SAP4, SAP5, and SAP6 gene expression in rim101⁻/rim101⁻ mutant strain CJN793 compared to the expression in reference strain DAY185 (Table 2). Since SAP5 has been shown to play an important role in oropharyngeal candidiasis in animals and humans (33, 38, 46), we focused on the role of SAP5 in E-cadherin degradation. In order to demonstrate that SAP5 expression was capable of rescuing the rim101⁻/rim101⁻ mutant from its invasion- and E-cadherin degradation-deficient phenotype, we compared the abilities of strains CJN793 (rim101⁻/rim101⁻), CJN1111 (rim101⁻/rim101⁻ TEF1-SAP5/SAP5), and reference strain DAY185 to (i) degrade immunoprecipitated E-cadherin and (ii) trigger E-cadherin degradation and invade a three-dimensional model of the human oral mucosa. The SAP5-expressing rim101⁻/rim101⁻ strain CJN1111 exhibited a specific 4 × 10⁵-fold increase in SAP5 RNA levels (Table 2) compared to the levels in rim101⁻/rim101⁻ strain CJN793, but the SAP5 expression levels were 10⁴-fold lower than the levels in RIM101/RIM101 strain DAY185. These low levels of SAP5 expression nonetheless permitted degradation of immunoprecipitated E-cadherin with an efficiency similar to that of the invasive control strain DAY185 (Fig. 8A, lanes 3 and 4, respectively). Moreover, incubation of three-dimensional cultures with SAP5-expressing rim101⁻/rim101⁻ strain CJN1111 resulted in degradation of E-cadherin in localized areas of fungal invasion (Fig. 6D). These results indicated that increased SAP5 expression was sufficient to restore the E-cadherin degradation and invasion phenotypes of the rim101⁻/rim101⁻ strain.

FIG. 8.

Effect of secreted aspartyl proteinases on the integrity of E-cadherin. (A) Role of SAP5 expression in the RIM101 requirement for E-cadherin degradation. Immunoprecipitated E-cadherin was incubated with rim101⁻/rim101⁻ mutant strain CJN793, rim101⁻/rim101⁻mutant strain CJN1111 expressing SAP5, or reference strain DAY185 for 4 and 6 h, and the protein integrity was analyzed by Western blotting. The positions of the molecular mass markers are indicated on the left. The results of one of two independent experiments in which similar data were obtained are shown. (B) Effect of loss of SAP5 and loss of SAP4, SAP5, and SAP6 on the integrity of E-cadherin. Immunoprecipitated E-cadherin was incubated with SAP5 gene knockout strain DSY 452, SAP4-SAP5-SAP6 gene knockout strain DSY 459, or reference strain CAI4 for 12 h, and the protein integrity was analyzed by Western blotting. The results of one of two independent experiments in which similar data were obtained are shown. The positions of the molecular mass markers are indicated on the left.

Finally, to strengthen our conclusions concerning the role of secreted aspartyl proteinases in the degradation of E-cadherin, we compared the abilities of sap5 homozygous mutant strain DSY452, sap4-sap5-sap6 triple mutant strain DSY459, and reference strain CAI4 to degrade immunoprecipitated E-cadherin. Both the sap5 and sap4-sap5-sap6 mutant strains were deficient for E-cadherin degradation after 12 h of incubation (Fig. 8B, lanes 2 and 3, respectively) compared to parent strain CAI4 (Fig. 8B, lane 4).

DISCUSSION

C. albicans invasion of the superficial layers of the oral epithelium has been demonstrated using human and animal models of oropharyngeal candidiasis (11, 17, 18) and is one of the factors that determine the intensity of the host proinflammatory response to this infection (48). Intercellular contacts between epithelial cells are mediated through specialized adhesion complexes, such as adherens junctions. In these junctions, E-cadherin acts as a transmembrane calcium-dependent homophilic adhesion molecule, linking the actin cytoskeletons of adjacent cells (47). Disruption of E-cadherin has a profound effect on the structural integrity of the epithelium (24, 43) and might facilitate microbial invasion (19, 20, 35, 50). In this study, using both monolayer and three-dimensional multilayer cultures, we demonstrated for the first time that E-cadherin is degraded by C. albicans. Furthermore, we obtained evidence which supports the hypothesis that E-cadherin degradation plays a role in the invasion of oral mucosa by this pathogen.

Our results also showed that degradation of E-cadherin during incubation of oral epithelial cells with C. albicans is mediated by Sap5p, a secreted aspartyl proteinase that plays a role in C. albicans virulence (32). Although this is the first time that an aspartyl proteinase has been shown to be responsible for E-cadherin degradation, other secreted aspartyl proteinases have been shown to degrade gastrointestinal mucin and endothelial cell basement membrane proteins in vitro (6, 30), suggesting that proteinases may play important roles in the penetration of C. albicans into host tissues. The cleavage site for each secreted aspartyl proteinase is not exclusive, and the sites frequently overlap (21). Although secreted aspartyl proteinases have been found to have broad substrate specificity, structural studies of Sap1p, Sap2p, Sap3p, and Sap6p have shown that the differences are sufficient that there can be distinct substrate specificities (21). Cleavage site preferences for Sap1p to Sap3p and Sap6p have been reported previously (21); however, the substrate specificity of Sap5p remains unknown (32).

Our data were confirmed by the finding that SAP5 is strongly activated during invasion of the esophageal mucosa by C. albicans in a mouse model of infection (36), where SAP5 is one of the early genes activated during hyphal penetration. Moreover, SAP5 is the most commonly expressed SAP gene in samples from patients with oral candidiasis (33). In the reconstituted human epithelium vaginal model, SAP5 expression precedes expression of other SAP genes and is associated with the development of lesions (40). Surprisingly, although SAP5 has been shown to be strongly expressed in a reconstituted human epithelium oral model of infection, it did not appear to play a role in tissue damage or invasion when the expression of SAP1, SAP3, and SAP4 was down-regulated (22). Differences in culture conditions and types of cells used in three-dimensional models may affect such host cell interactions with the fungus; moreover, nutrient availability or levels of E-cadherin expression in different systems may provide strong local pressures that influence the proteolytic activity of Sap5p. In addition, the cellular compositions of different tissues may also favor the roles of specific SAP gene products. For example, recent studies using an intraperitoneal mouse model of infection demonstrated that while disruption of SAP5 did not attenuate invasion of parenchymal organs, deletion of SAP6 significantly reduced tissue invasion (12).

Although evidence indicates that SAP5 plays a key role in oropharyngeal infection, the specific function of SAP5 during oral infection has not been described previously. Our study provided an explanation for the high level of expression of SAP5 in oroesophageal infection by identifying an oral epithelial cell-specific target for the enzymatic activity. Our study also explained the decrease in the level of the E-cadherin protein observed in the oral mucosa of patients with oropharyngeal candidiasis (28).

We found that C. albicans gains access to both intracellular and intercellular spaces of the three-dimensional model of the human oral mucosa, suggesting that this fungus uses more than one mechanism to invade tissues. In addition, the degradation of E-cadherin in localized areas where there is invasion of C. albicans in a three-dimensional model of the human oral mucosa further supports the hypothesis that there is a mechanism of tissue invasion by C. albicans via an intercellular pathway that involves degradation of this protein. This hypothesis is consistent with reports in which proteolytic breakdown of E-cadherin has been proposed as a mechanism of microbial invasion (19, 20, 35, 50). The enteropathogen B. fragilis has been shown to secrete a zinc metalloproteinase which specifically degrades E-cadherin on the basolateral surfaces of intestinal epithelial cells (35, 50). Furthermore, it has been demonstrated that lysine-specific gingipains derived from P. gingivalis, a gram-negative anaerobe implicated in the etiology of periodontal disease, also degrade the adherens junction protein E-cadherin from oral epithelial cells (19, 20).

In conclusion, our data support the hypothesis that there is a mechanism by which C. albicans invades mucosal tissues via an intercellular pathway by promoting the degradation of E-cadherin in epithelial adherens junctions. Here we present novel information that has a direct impact on our understanding of the pathogenesis of this mucosal infection.