Role for Dendritic Cells in Immunoregulation during Experimental Vaginal Candidiasis

Abstract

Vulvovaginal candidiasis (VVC) caused by the commensal organism Candida albicans remains a significant problem among women of childbearing age, with protection against and susceptibility to infection still poorly understood. While cell-mediated immunity by CD4⁺ Th1-type cells is protective against most forms of mucosal candidiasis, no protective role for adaptive immunity has been identified against VVC. This is postulated to be due to immunoregulation that prohibits a more profound Candida-specific CD4⁺ T-cell response against infection. The purpose of this study was to examine the role of dendritic cells (DCs) in the induction phase of the immune response as a means to understand the initiation of the immunoregulatory events. Immunostaining of DCs in sectioned murine lymph nodes draining the vagina revealed a profound cellular reorganization with DCs becoming concentrated in the T-cell zone throughout the course of experimental vaginal Candida infection consistent with cell-mediated immune responsiveness. However, analysis of draining lymph node DC subsets revealed a predominance of immunoregulation-associated CD11c⁺ B220⁺ plasmacytoid DCs (pDCs) under both uninfected and infected conditions. Staining of vaginal DCs showed the presence of both DEC-205⁺ and pDCs, with extension of dendrites into the vaginal lumen of infected mice in close contact with Candida. Flow cytometric analysis of draining lymph node DC costimulatory molecules and activation markers from infected mice indicated a lack of upregulation of major histocompatibility complex class II, CD80, CD86, and CD40 during infection, consistent with a tolerizing condition. Together, the results suggest that DCs are involved in the immunoregulatory events manifested during a vaginal Candida infection and potentially through the action of pDCs.

Vulvovaginal candidiasis (VVC) is a significant problem affecting 75% of normal healthy women at least once during their reproductive years (39, 40). Candida albicans, the causative agent of VVC, is a dimorphic fungal organism and both a commensal and opportunistic pathogen of the genital and gastrointestinal tracts (11, 16, 25, 27, 32, 36). Symptoms of VVC include itching, burning, soreness, abnormal discharge, dyspareunia, and vaginal and vulvar erythema and edema (40). Factors contributing to the development of disease include antibiotic and high estrogen contraceptive use, hormone replacement therapy, pregnancy, and uncontrolled diabetes mellitus (15, 22, 39, 40). Although most women experience infrequent occurrences of VVC, an additional 5 to 10% of women suffer from recurrent VVC, defined as three or more episodes of infection per annum without any recognizable predisposing factors (38, 39).

While CD4⁺ Th1-type cell-mediated immunity (CMI) is the dominant host defense mechanism for most Candida infections of mucosal tissues (4, 33-35), no protective role for local or systemic CMI has been demonstrated for vaginal Candida infections. Instead, it is postulated that a protective adaptive immune response to vaginal Candida infection is prohibited by a local tolerizing condition or immunoregulatory environment at the vaginal mucosa and the draining lymph nodes. This is supported by the presence of CD4⁺ CD25⁺ regulatory T cells in the draining lymph nodes (43), γ/δ T cells in the vagina (42), and immunoregulatory cytokines (interleukin-10 [IL-10] and transforming growth factor β [TGF-β]) in the lymph nodes and vaginal tissue (42). Furthermore, there is little evidence for changes in the percentage or composition of vaginal T-cell populations during experimental vaginal candidiasis (12), suggesting a lack of systemic T-cell infiltration or local T-cell expansion. Immunoregulation at a reproductive site may reflect a means for the host to avoid chronic inflammation to a commensal organism such as Candida. Yet some form of host response is required to keep Candida in a commensal state and out of a pathogenic condition. Current evidence suggests this involves innate immune mechanisms by epithelial cells (2, 10, 44).

Dendritic cells (DCs) are considered the major antigen-presenting cell (APC) responsible for activation of the adaptive immune response. Recent studies also suggest a role for DCs in the initiation of immunoregulatory events and maintenance of tolerogenic conditions. Different DC subsets appear responsible for each condition: CD11c⁺ B220⁺ plasmacytoid DCs (pDCs) are involved in tolerance induction, whereas CD11c⁺ B220⁻ myeloid DCs are involved in the induction of classical inflammatory responses. pDCs have been shown to be the predominant DC population in lymph nodes under tolerizing conditions (28), express low levels of major histocompatibility complex (MHC) class II and costimulatory molecules (24, 28), and can induce the differentiation of IL-10-producing T_reg cells (5, 17, 24, 37). This is in contrast to the myeloid DCs that express high levels of MHC class II and costimulatory molecules in their activated state and are associated with Th cell activation (24, 28). Finally, DCs are known to be present in the vaginal mucosa (18-20, 41, 45), and studies on the interaction of Candida with bone marrow-derived DCs have been reported (1, 6, 26, 31).

The purpose of this study was to investigate the presence of DCs in the vagina during infection with C. albicans and to examine the potential role of DCs in the induction phase of the vaginal immune response to Candida in the vagina that includes the immunoregulatory events, using the well-established murine model of experimental vaginal candidiasis.

MATERIALS AND METHODS

Mice.

Female CBA/J (H-2^k) mice, 8 to 10 weeks of age (purchased from the National Cancer Institute; Frederick, MD), were used throughout these studies. All animals were housed and handled according to institutionally recommended guidelines.

Vaginal Candida infection.

Mice were injected subcutaneously with 100 μl of 0.2 mg/ml estradiol (Sigma Chemical Co., St. Louis, MO) in sesame oil 72 h prior to inoculation. Intravaginal inoculation was performed by introducing 20 μl of phosphate-buffered saline (PBS) containing 5 × 10⁴ C. albicans blastoconidia from a stationary-phase culture (overnight culture at 25°C in Phytone-peptone medium plus 0.1% glucose) into the vagina. Control animals were estrogenized and inoculated with PBS only or PBS containing 5 × 10⁴ heat-killed C. albicans blastoconidia (incubated at 60°C for 2 h, washed, and cultured for verification). Separate groups of mice were sacrificed at days 4 and 7 postinoculation. Vaginal lavages were conducted using 100 μl of sterile PBS with repeated aspiration for 30 to 40 s, and the fluid was cultured at 1:10 serial dilutions on Sabouraud-dextrose agar plates (Becton Dickinson and Co., Cockeysville, MD) supplemented with gentamicin (Sigma) as previously described (13). CFU were enumerated after incubation at 35°C for 48 h and expressed as CFU/100 μl.

Specimen collection and processing.

Vaginal draining lumbar lymph nodes were identified on the posterior abdominal wall lateral to the inferior vena cava and abdominal aorta, respectively. These lymph nodes and/or the vagina were excised and embedded in optimum-cutting-temperature (OCT) medium (Tissue-Tek, Torrance, CA) within cryomolds in an orientation that allowed for cross-sectional cutting. Tissue sections were cut (8 to 12 μm), fixed in acetone (10 min), and then washed in PBS (5 min). For dendritic cell enrichment, lumbar lymph nodes were removed and placed in RPMI 1640 medium (Sigma). Single-cell suspensions were prepared by digesting the nodes in collagenase D (400 U/ml) (Sigma) for 25 min at 37°C and 5% CO₂. The collagenase-cell suspension was incubated for 1 min at room temperature in 0.1 M EDTA, filtered (40-μm pore size), and centrifuged, and the cell pellet was resuspended in PBS supplemented with 0.5% fetal bovine serum (FBS). Cells were then enumerated with a hemocytometer. CD11c⁺ DCs were enriched from the total lymph node cell population by magnetic bead selection (Miltenyi Biotec, Auburn, CA). Briefly, nonspecific protein sites on the cells were blocked by incubation with anti-Fc receptor antibodies. Cell suspensions were next incubated with anti-CD11c-antibody (clone N418)-conjugated microbeads (100 μl) and passed over a magnetic selection column (Miltenyi Biotec). After washing the column with PBS-0.5% FBS to remove unlabeled cells, the column was demagnetized and positively selected cells were eluted and collected by centrifugation. Pelleted cells were resuspended in PBS-0.5% FBS and enumerated with a hemocytometer. DC enrichment was confirmed by flow cytometry, with the final enrichment ranging from 60 to 80% CD11c⁺ cells.

Immunohistochemistry. (i) Chromophore.

Vaginal and lymph node tissues were stained using the R&D Systems tissue staining kit (Minneapolis, MN). Briefly, tissues were blocked with peroxidase (5 min), goat serum (15 min), avidin (15 min), and biotin (15 min) blocking buffers and rinsed in PBS after each step. Tissues were then incubated with purified rat anti-mouse DEC-205 (0.2 μg/ml) (BMA Biomedicals; Rheinstrasse, Switzerland) or plasmacytoid DC marker (1 μg/ml) (Miltenyi Biotec) at room temperature for 1 h. Rat immunoglobulin G2a (IgG2a) isotype-specific antibody staining was used as a negative control. Following incubation, the slides were washed three times (15 min/wash) in PBS and then incubated with biotinylated goat anti-rat IgG antibody (5 μg/ml) (BD Biosciences; San Jose, CA) for 30 min. Washed slides were then incubated with high-sensitivity streptavidin-horseradish peroxidase conjugate (30 min), washed, and incubated with the substrate 3,3′ diaminobenzidine (20 min). Slides were counterstained with Mayer's hematoxylin and preserved using aqueous mounting medium (R&D Systems).

(ii) Fluorescence.

Vaginal tissues were blocked with goat serum (20 min), incubated with purified rat anti-mouse DEC-205 (0.2 μg/ml) (BMA Biomedicals) for 1 h, washed in PBS (5 min/wash), and then incubated with biotinylated goat anti-rat IgG antibody (5 μg/ml) (BD Biosciences) for 30 min. Rat IgG2a isotype-specific antibody staining was used as a negative control. Slides were washed in PBS (5 min/wash) then incubated with streptavidin-conjugated phycoerythrin (PE; 25 μg/ml). Slides were washed in PBS (5 min/wash) and then incubated with calcofluor white M2R (Molecular Probes; Eugene, OR) for 10 min to stain C. albicans. Slides were washed and visualized by fluorescent microscopy under the tetramethyl rhodamine isocyanate (TRITC) and 4′,6′-diamidino-2-phenylindole (DAPI) filters.

Flow cytometry.

Analysis of cell surface molecules on lymph node cells enriched for DCs was performed by flow cytometry using standard methodology for direct and indirect immunofluorescence. Briefly, cells were incubated with combinations of PE-labeled anti-plasmacytoid DC marker (Miltenyi Biotec), fluorescein isothiocyanate (FITC)-labeled anti-B220, PE- or biotin-labeled anti-CD11c (clone HL3), biotin-labeled anti-I-A^k, or purified anti-CD80, anti-CD86, or anti-CD40 antibodies (BD Biosciences) at optimal concentrations (2.0 to 2.5 μg/ml) for 20 min at 4°C. Cells were washed in buffer, and those labeled with purified antibodies were then incubated with a secondary biotinylated anti-rat IgG antibody (2.5 μg/ml) for 20 min at 4°C. Cells were washed and incubated in streptavidin-conjugated Cy-Chrome for 20 min at 4°C. After incubation, the cells were washed and fixed in Poly-Lem fixative (200 μl) (Polysciences, Inc., Warrington, PA). Samples were analyzed on a FACS Vantage SE flow cytometer (Becton Dickinson; Franklin Lakes, NJ) using Cellquest software at the LSU Health Sciences Center Core Labs Facility. Cells incubated with either buffer alone or fluorochrome-conjugated isotype control antibodies (rat IgG2a) were used to determine background staining. For data analyses, cells were evaluated from a predominantly DC population identified by gating on large, forward-scattering cells with isotype-specific antibody staining as a negative control. Compensation for each fluorochrome was determined by parallel single-color analysis of cells labeled with each fluorochrome-conjugated antibody.

Statistical analysis.

The unpaired Student's t test was used to analyze data. Significant differences were defined as a confidence level at which P was <0.05, using a two-tailed test. All statistics were evaluated using GraphPad Prism (GraphPad Software, San Diego, CA).

RESULTS

Effects of vaginal candidiasis on draining lymph node architecture.

In initial studies, we sought to evaluate the DC presence and distribution in vaginal draining lumbar lymph nodes in uninfected mice, mice inoculated with killed Candida blastoconidia, and mice inoculated with live Candida albicans blastoconidia. For this, lymph node tissue sections were stained for the lymphoid DC marker DEC-205 and visualized microscopically. Lymph nodes from mice given PBS intravaginally showed DCs evenly dispersed throughout the lymphoid tissue (Fig. 1A). The presence of live Candida cells in the vagina induced a profound cellular reorganization in draining lymph nodes within 4 days, with DCs becoming concentrated in the T-cell zones of the lymph nodes (Fig. 1B). Lymph nodes from mice given killed Candida intravaginally showed no signs of lymph node cellular migration and rearrangement, similar to PBS-treated mice (Fig. 1C). Tissue stained with isotype control antibodies (rat IgG2a) showed no staining (Fig. 1D). Vaginal infection with live Candida blastoconidia was verified based on a vaginal lavage and quantitative plate counts. There was no fungal growth from lavages of mice treated with PBS or killed Candida, consistent with the lack of yeast colonization in mice (data not shown). Henceforth, PBS was used as the negative control.

FIG. 1.

Effects of vaginal candidiasis on draining lymph node architecture. Frozen lumbar lymph node tissue sections from mice inoculated with PBS (uninfected) (A), live C. albicans blastoconidia (B), or killed C. albicans blastoconidia (C) 7 days postinoculation were stained with anti-DEC-205 (α-DEC-205) antibody (0.2 μg/ml) or isotype control antibodies (rat IgG2a) (D). The figure shows representative images of three repeats with three mice per group. Images are shown at a magnification of ×100.

Analysis of dendritic cell subsets in vaginal draining lymph nodes.

Since immunohistochemical staining clearly indicated DC reorganization in the lymph nodes of infected mice suggestive of DC-T-cell interaction, we next sought to analyze the properties of the DC populations in the vaginal draining lymph nodes of uninfected (baseline) and infected mice using flow cytometry. Based on the differential expression of B220 on myeloid and plasmacytoid DCs as evidence of an inflammatory versus tolerizing condition (28), we first evaluated the presence of B220 on forward-scattering (large) cells and then confirmed them as CD11c⁺ MHC class II⁺ DCs. Results demonstrated that a higher percentage of the forward-scattering CD11c⁺ MHC class II⁺ DCs (Fig. 2A, left frame) were B220⁺ (plasmacytoid) DC (Fig. 2A, center frame) and were confirmed by the smaller percentage of B220⁻ (myeloid) DC (Fig. 2A, right frame). We next determined that the majority of the B220⁺ cells were putative pDC, indicated by the high percentage of B220⁺ cells that stained positive for both pDC marker and CD11c (Fig. 2B). These results did not appreciably change in mice with vaginal Candida infections through 7 days postinoculation (data not shown). Analysis of the absolute numbers of DCs per lymph node before enrichment indicated that despite increased numbers of cells per lymph node during infection (P = 0.008), the percentage of DCs, including pDCs, remained similar (P > 0.05) (Table 1).

FIG.2.

Analysis of DC subsets in vaginal draining lymph nodes. Lymphocytes were isolated from draining lumbar lymph nodes of uninfected mice and enriched for CD11c⁺ cells. (A) Cells were labeled with B220-FITC, CD11c-PE, and MHC class II-Cy-Chrome and analyzed by flow cytometry. Cells were stratified by gating on the entire forward-scattering population, forward-scattering B220⁺ cells, or forward-scattering B220⁻ populations (upper panel) and then determining the percentage of DC (CD11c⁺ MHC class II⁺) in the gated region (lower panel). (B) Cells were labeled with B220-FITC, CD11c-Cy-Chrome, and plasmacytoid DC (P-DC) marker-PE and analyzed by flow cytometry. The forward-scattering B220⁺ cells (upper panel) that were also CD11c⁺ plasmacytoid DC marker positive (lower panel) were used to determine the percentage of plasmacytoid DCs. Cells labeled with isotype control antibodies were used to set the initial gates. Panels A and B each show a representative result from three repeats using pooled cells from five mice per group.

TABLE 1.

Dendritic cells in draining lumbar lymph nodes during experimental vaginal candidiasis

Sample	No. of cells/lymph nodea	% CD11c+	No. of DCs/lymph nodeb	% B220+	No. of pDC/lymph nodec
Uninfected	8.5 × 105 ± 1.8 × 105	2.6 ± 0.43	∼2.3 × 104	1.6 ± 0.1	∼1.36 × 104
7 days postinfection	3.0 × 106 ± 4.0 × 105	3.3 ± 0.83	∼9.68 × 104	1.9 ± 0.2	∼5.7 × 104

^aValues are mean ± standard error of the mean for three experiments using lymphocytes pooled from five mice per group.

^bCalculated from percent CD11c⁺.

^cCalculated from percent B220⁺.

Dendritic cells in vaginal tissue.

We next examined the presence of DCs in the vagina before and after inoculation with Candida. Immunohistochemical staining for DEC-205 revealed the presence of DCs dispersed throughout the vaginal lamina propria and epithelium in both PBS-treated and infected mice (Fig. 3A and B). However, in infected mice (days 4 and 7 postinoculation), DCs were also observed at the outer epithelium with dendrites extending into the vaginal lumen (Fig. 3B). Tissue stained with isotype control antibodies (rat IgG2a) showed no staining (data not shown). Concurrent immunofluorescent staining of DEC-205⁺ DCs and C. albicans chitin revealed colocalization of DCs with Candida at the surface of the vaginal epithelium (Fig. 3C). Staining for vaginal pDCs in uninfected mice revealed positively stained cells similarly dispersed throughout the vaginal epithelium (Fig. 3D). Under infected conditions (days 4 and 7 postinoculation), pDCs were found accumulated at the interface of the lamina propria and the epithelium (Fig. 3E).

FIG. 3.

Dendritic cells in vaginal tissue. Frozen vaginal tissue sections from estrogenized mice intravaginally inoculated with PBS (frames A and D) or live Candida (frames B, C, and E) were labeled with anti-DEC-205 (frames A and B), anti-DEC-205-PE, and calcofluor white (frame C) or anti-plasmacytoid DC marker (frames D and E). Tissue regions are labeled. White arrows indicate DC staining in vaginal tissue. The small insert for frame C represents a standard image of tissue revealing the orientation for the confocal image. The figure shows representative images of three repeats with three mice per group. The magnification is ×400 for all images.

Effects of vaginal Candida infection on dendritic cell activation markers in the draining lymph nodes.

Next we examined the effect of vaginal Candida infection on MHC class II and costimulatory molecule expression on DCs in the draining lymph nodes. Controls included mice given PBS intravaginally. Lymph nodes were removed at days 4 and 7 postinoculation, and total lymph node cells were enriched for DCs. Figure 4A shows a representative flow cytometric result for the expression of the four markers on B220⁺ cells under uninfected and infected conditions. Cumulative results of the total forward-scattering DC population (total DC) for several experiments revealed no statistical differences in mean fluorescence intensity (MFI) for MHC class II, CD80, CD86, and CD40 with a trend toward downregulation on DCs from infected mice on both days 4 and 7, as compared to DCs from control animals given PBS and expressed as percent change (Fig. 4B). When stratified by the B220 marker specifically, it was determined that the downregulatory trend for each of the markers was associated with the B220⁺ cells rather than the B220⁻ cells (Fig. 4C and D), although only the changes for CD86 reached statistical significance (P = 0.0029).

FIG. 4.

Effects of vaginal Candida infection on DC in the draining lymph nodes. Lymphocytes were isolated from draining lymph nodes of mice given PBS or live Candida intravaginally, harvested at day 4 or 7 postinoculation, and enriched for CD11c⁺ cells. Cells were labeled with B220-FITC, CD11c-PE, and MHC class II-Cy-Chrome, CD80-Cy-Chrome, CD86-Cy-Chrome, or CD40-Cy-Chrome and analyzed by flow cytometry. (A) Representative result of MFI for each marker of three experiments performed. B220⁺ cells were identified from total DCs (CD11c⁺). (B) Percent change in MFI ± standard error of the mean (SEM) for total DCs (CD11c⁺) for three experiments. (C) Percent change in MFI ± SEM for B220⁺ DCs for three experiments. (D) Percent change in MFI ± SEM for B220⁻ DCs for three experiments. Controls were labeled with isotype control antibodies to set gates. The figure shows results from pooled cells using five mice per group.

DC subsets in nonvaginal draining lymph nodes.

Since pDCs were found to be the predominant DC subset of the vaginal draining lymph nodes and have been found by several groups to have immunoregulatory potential (5, 17, 23, 24, 28), we next sought to determine if the tolerizing DC profile was unique to vaginal draining lymph nodes or common to all lymph nodes under naive conditions. To investigate this, the axillary and brachial lymph nodes were analyzed for DC subsets and compared to DC subsets in lumbar lymph nodes of naïve mice. Flow cytometric analysis revealed an equivalent percentage of pDCs among lymph node cells from all sites (47% ± 5% for vaginal draining lymph nodes versus 55% ± 8% for nonvaginal draining lymph nodes).

DISCUSSION

Although Th1-type CMI is protective against Candida infection at most mucosal sites, there is no evidence for protection against vaginal candidiasis. Instead, considerable evidence suggests that immunoregulation compromises such protection. The aim of this study was to investigate the role of dendritic cells in the initiation of these immunoregulatory processes.

Immunohistochemical analysis of DCs in vaginal draining lymph nodes revealed a profound cellular reorganization, with DCs becoming concentrated in the T-cell zones of lymph nodes following an experimental vaginal Candida infection, but not in control animals given PBS or killed Candida. This cellular rearrangement under infected conditions was evidence for initiation of classical cellular immune activation events, as suggested by our in vitro studies (14), with significant DC involvement in the immune response to a vaginal Candida infection.

In contrast to the implications surrounding the cellular reorganization in the draining lymph nodes of infected mice, examination of specific DC subsets revealed that plasmacytoid DCs (CD11c⁺ B220⁺) predominated prior to and throughout the vaginal Candida infection. While it is common to have pDCs in the lymphoid organs under naive conditions, as confirmed in the present study, the continued presence during Candida infection is indicative of a tolerizing condition. For example, in a murine airway inflammation model, Oriss et al. (28) demonstrated that while mDCs became the predominant DC subset under inflammatory conditions, pDCs remained the predominant subset after stimulation with antigen under tolerizing conditions (28). Similar lymph node DC dynamics have been demonstrated in studies with other fungal immunogens. Bauman et al. showed that in mice immunized with an immunogen that induced a protective response to systemic infection with Cryptococcus neoformans, mDCs were the predominant DC population in lymph nodes draining the site of immunization (3). Conversely, in mice given nonprotective cryptococcal immunogen, the lymphoid DC population (now considered plasmacytoid) was predominant (3). Thus, considerable evidence suggests that the myeloid-plasmacytoid DC balance in draining lymph nodes in response to specific antigen is important for the development of protective or tolerizing CMI conditions.

Flow cytometric analysis of MHC class II and costimulatory molecules (CD80, CD86, and CD40) on lymph node DC following a vaginal Candida infection, which revealed a lack of upregulation of these molecules, is consistent with the presence of pDCs in the vaginal draining lymph nodes (known to express low levels of MHC class II and costimulatory molecules). In fact, when focusing on the B220⁺ pDCs, there was evidence for substantial downregulation of the respective markers, although statistical significance was only obtained for CD86 at one time point (day 7). This failure to upregulate MHC class II and costimulatory molecules and trend toward a downregulation further suggest a deviation from events associated with a classical inflammatory response and the establishment of a tolerizing condition.

Analysis of DCs in the vaginal tissue revealed the presence of DEC-205⁺ DCs at the edge of the epithelium of infected mice with dendrites extending into the lumen. Evidence for the extension of dendrites through the epithelium for the purpose of sampling the microenvironment has been shown previously in the gastrointestinal tract (29). Furthermore, pDCs that are not DEC-205⁺ were seen to aggregate at the interface of the lamina propria and the epithelium. DC positioning in the vaginal tissue during Candida infection suggested that DEC-205⁺ DCs were in close proximity to Candida in the vaginal lumen and that pDCs at least migrated towards the epithelium. Although it is not known why the pDCs failed to reach the outer epithelium or what role the interaction of DEC-205⁺ DCs with Candida may ultimately play in the inflammatory versus immunoregulatory outcome, it is possible that the pDCs will or have had contact with Candida through processed antigen in the tissues or from Candida hyphae that may have invaded into the epithelium. Alternatively, the tissues may not have been sampled at the optimal time to observe complete migration of pDCs to the outer epithelium.

As additional evidence for a direct interaction of DCs and Candida in the vagina, immunofluorescent confocal microscopy demonstrated the close proximity of DCs and Candida hyphae at the epithelial surface. This putative interaction provides strong evidence of a role for DCs in the initiation of cellular manifestations observed in the draining lymph nodes in the present study. Interactions of bone marrow-derived DCs with Candida have been demonstrated previously in vitro and are shown to be receptor mediated, depending on fungal morphology (i.e., yeast versus hyphae) (6, 26, 31). We postulate that after interaction with Candida in the vagina, vaginal DCs take up Candida antigen and travel to the draining lymph nodes where they are able to induce activation of CD4⁺ CD25⁺ T_reg cells observed in previous studies (43) and initiate regulatory cytokine production (i.e., TGF-β and IL-10) (42), which together play a potential role in the decrease of T cells expressing homing receptors required to migrate and enter the vaginal mucosa via tissue-associated adhesion molecules (43). From this, a general state of tolerance is created, resulting in a lack of protective CMI against VVC.

A paradoxical observation as alluded to earlier is that cells isolated from the draining lymph nodes of mice inoculated intravaginally with Candida responded to Candida antigen in vitro by the production of Th1-type cytokines (IL-2 and gamma interferon [IFN-γ]) as evidence of a classical inflammatory response (9). This is in contrast to the evidence for DC-induced immunoregulation and lack of Th1-type CMI response observed in vivo (12, 42, 43). The in vitro result may be explained by the removal of the responsive cells from the local immunoregulatory cytokine milieu whereby the T cells can be stimulated to produce a more classical inflammatory response.

Other groups have also demonstrated that exposure of DCs to antigen in the vagina failed to induce DC activation. Fausch and colleagues have shown that human papillomavirus-like particles (HPV VLP) failed to induce upregulation of Langerhans cell activation markers and cytokine production despite antigen uptake, although other DCs (non-Langerhans) became activated by HPV VLP (7, 8). This is significant because Langerhans cells are found in vaginal epithelium and would be the first APCs to come in contact with a vaginal pathogen. Studies by Zhao et al. using a murine model of vaginal herpes simplex virus type 2 infection provide yet another example of the lack of epithelium-associated Langerhans cell activation in response to a vaginal pathogen, although submucosal DCs exposed to antigen were able to induce an inflammatory response in the draining lymph nodes (45). These results emphasize the tolerogenic nature of Langerhans cells in response to vaginal pathogens associated with the epithelial layer, whereas submucosal DCs may not be inherently tolerogenic. It would be interesting to determine whether tolerance to Candida can be broken in mice by injection of Candida into the vaginal submucosa or by coinfecting mice with another vaginal pathogen that stimulates an inflammatory response. The latter is unlikely, however, based on a previous study where responses, or lack thereof, following coinfection with Candida and Chlamydia trachomatis remained independent (21).

Mucosal sites, especially the vaginal mucosa, are considered very immunotolerant environments. This is important in light of the many commensal organisms of these tissues and the need to distinguish between pathogen and commensals. A tolerizing microenvironment in the vagina protects against unwanted inflammatory responses at a site that is constantly exposed to foreign antigens and creates avoidance of states of chronic inflammation. Furthermore, immune tolerance in the reproductive tract is especially important for fertilization and pregnancy, and substantial evidence for immunoregulation involving immunoregulatory cytokines and possible cellular mediators in the upper reproductive tract has been demonstrated (reviewed in reference 30). Future work on the role of DCs in the initiation of immunoregulation during vaginal candidiasis may include studies of DC-specific cytokine production in the draining lymph nodes and studies to elucidate the intracellular signaling mechanisms involved in DC-mediated immunoregulation similar to what has been shown for human papillomavirus (8). Noteworthy for such studies is the fact that the murine model is distinct from the human condition by virtue that Candida is not a commensal in rodents. Thus, the results from the murine model may not be directly applicable to humans, although for issues related to immunoregulation at the vaginal mucosa, data from the animal model have mirrored data from humans very well (10). Thus, there is confidence that the murine model is adequate and appropriate for such studies and will eventually be confirmed in humans.

In summary, we have demonstrated that pDCs are a major subset of DCs in the vagina and draining lymph nodes and remain so throughout the course of an experimental vaginal Candida infection. Additionally, vaginal infection with C. albicans fails to induce the upregulation of MHC class II antigen and costimulatory molecules on DCs that is consistent with a predominant immunoregulatory microenvironment. We hypothesize that initiation of T-cell tolerance and immunoregulation that begins in the vagina and continues in the draining lymph nodes during vaginal candidiasis occurs through an established mechanism of antigen presentation involving DCs, and in particular pDCs, in the absence of costimulation.