Candida albicans Modulates Murine and Human Beta Defensin-1 during Vaginitis

Vulvovaginal candidiasis (VVC) and recurrent vulvovaginal candidiasis (RVVC) are two forms of a disease caused by Candida spp. β-defensin (BD) is one of the most important families of antimicrobial peptides in the female genital tract and includes molecules that exert essential local functions as antimicrobial and PMN chemoattractant peptides. However, the information on their role during murine and human VVC and RVVC is limited. Thus, we analyzed the behavior and contribution of BD1 to the local response in a VVC mice model and the local cytokine profile and human BD1 and BD3 expression in cervicovaginal lavage from patients with VVC and RVVC. We demonstrated that, in patients with RVVC BD1, mRNA and protein expression were severely diminished and that the aspartate proteinase and lipase secreted by C. albicans are involved in that decrease. This study provides novel information about the pathogenesis of VVC and describes a highly efficient C. albicans escape strategy for perpetuating the infection; these results may contribute to the development of new or combined treatment approaches.


Introduction
Vulvovaginal candidiasis (VVC) is a widespread disease caused by Candida species that affects a significant number of women of reproductive age. Up to 75% of women experience acute VVC (AVVC) at least once in their lifetime, which is associated with risk factors such as pregnancy, oral contraceptives, diabetes mellitus, long-term broad-spectrum antibiotic treatment, and steroid and immunosuppressive therapies [1]. Moreover, up to 9% of women experience recurrent vulvovaginal candidiasis (RVVC), a more distressing condition characterized by at least four episodes of AVVC per year [2]; however, the great majority of patients with RVVC do not display any of the associated risk factors. In their systemic review, Denning et al. reported that RVVC affects more than 130 million women in any given year, with a global annual prevalence of 3871 cases per 100,000 individuals, highlighting the significant global burden that this infection represents [3,4].
Candida species coexist with bacterial communities of the vaginal microbiome in a harmonious balance and under the strict control of the local immune response. The dis-antibody (Vector Labs, Burlingame, CA, USA) in PBS with 1% BSA [23,27] and finally with the VECTASTAIN ® Elite ABC-HRP Kit, Peroxidase (Vector Labs, Burlingame, CA, USA). Diaminobenzidine (Sigma-Aldrich, St. Louis, MO, USA) was used as chromogen substrate. The histological sections were observed with a NIKON ECLIPSE microscope (Nikon, Tokyo, Japan).

Flow Cytometry
Vaginal lavage (VL) cells were collected from 3-4 mice per group, resuspended in PBS-FCS 2%, and permeabilized with Cytofix/Cytoperm (BD Biosciences, East Rutherford, NJ, USA). The VL cells were then incubated with a rabbit antimouse BD1 antibody (Santa Cruz Biotechnology, Dallas, TX, USA) and finally with a secondary goat antirabbit AlexaFluor 647 antibody (Abcam, Waltham, MA, USA). All the staining steps were performed at 4 • C for 30 min. In another set of experiments, Gr-1+ and mBD1+ expression in VL cells were analyzed. First, VL cells were stained with a goat antimouse Gr-1PE conjugated antibody (BD Bioscience) and subsequently fixed and permeabilized with Cytofix/Cytoperm. Then the cells were incubated with a rabbit antimouse BD1 antibody (Santa Cruz Biotechnology, Dallas, TX, USA) and finally with a secondary goat antirabbit AlexaFluor 647 antibody. The controls used were unstained cells (negative control) and isotype control (Sigma-Aldrich, St. Louis, MO, USA). Analysis was performed on living cells by excluding dead cells by FSC vs. SSC and Singlets gating, as shown in the gating strategy Figure S1. The absolute number of mBD1+ cells in VL was calculated based on the percentage of mBD1+ cells and the total cell count of VL cells obtained using a Neubauer chamber. For the evaluation the expression of hBD1 in HeLa cells, rabbit anti-human BD1 antibody (Santa Cruz Biotechnology) and goat anti-rabbit AlexaFluor 647 secondary antibody (Abcam) were used. The assays were performed as described above. Stained cells were acquired (300,000 events) using a FACSCanto II cytometer (BD Biosciences), and the data were analyzed using the FlowJo software (Tree Star, Inc. Becton, Dickinson and Company, Ashland, OR, USA).

Study Participants
The patients included in this study were attending to the Hospital Privado Universitario de Córdoba and Sanatorio Allende, in Córdoba, Argentina. A total of 105 women were enrolled, and they were categorized into three groups: RVVC, presented with four episodes of VVC per year (n = 59); AVVC, presented with ongoing vaginal Candida infection and no history of previous episodes (n = 20); and Control, healthy individuals, were negative for Candida infection, and matched in age with both groups of patients (n = 26). All the individuals were 18-52 years old, nondiabetic, negative for HIV, sexually active, and nonpregnant. Patients with vaginal infection other than Candida, patients with any coinfection, menopausal women, and patients using or receiving antimicrobial therapy within the previous two weeks were excluded from the study.
Cervicovaginal lavage (CVL) was collected by washing the cervix and vagina with 3 mL of sterile saline. The samples were centrifuged at 2000 rpm, and the cells were spun down onto slides to be used in immunofluorescence assays or preserved in TRIzol Reagent (Thermo Fisher, Waltham, MA, USA) RNA extraction. Supernatants were aliquoted and stored at −80 • C.

Cytokine Immunoassays
Cytokines in the CVL were quantified by ELISA assay using TNF-α, IL-6, IL-1β, TGFβ, IL-22, IL-23, and IL-17 detection kits (R&D System, McKinley Place, MN, USA) [19]. The samples were previously concentrated with a Vivaspin-6 centrifugal concentrator (GE Healthcare, Chicago, IL, USA). In the murine AVVC model, the level of mouse IL-1 β in the vaginal lavage was quantified using ELISA tests (R&D System, McKinley Place, MN, USA). Absorbance was measured using a BIO-RAD microplate reader, and the concentrations were extrapolated from standard curves. The data were expressed as picogram/milliliter (pg/mL).

Immunofluorescence Analysis
CVL cells attached to coverslides were fixed with 4% paraformaldehyde (Sigma-Aldrich) for 40 min at RT, washed with PBS, blocked, and permeabilized with PBS containing 10% BSA (Sigma-Aldrich) 0.3% Triton X-100 (Biopack, Buenos Aires, Argentina) for 1 h at RT. Then, the CVL cells were incubated overnight at 4 • C with rabbit antihuman BD1 or rabbit antihuman BD3 antibodies (Santa Cruz Biotechnology) and next with goat antirabbit AlexaFluor 647 secondary antibody (BD Biosciences) for 30 min at RT. After DAPI counterstaining, the cells were mounted with FluorSave Reagent (Merck-Millipore), and immunofluorescence was evaluated using a LEICA-DMi8 microscope (Leica Microsystems, Wetzlar, Germany). hBD1 and hBD3 semiquantification was performed using the ImageJ software. Corrected total fluorescence was calculated as [Integrated Density-(Epithelial cell area × Mean background fluorescence)]. Mean background fluorescence was calculated using three random areas of each image that were considered negative.

Real-Time PCR
Real-time PCR was used to evaluate the levels of hBD mRNA in CVL recovered cells [24,28,29] using the StepOne System (Life Technologies, San Francisco, CA, USA) and SYBR Select Master Mix (Life Technologies). RNA was obtained using TRIzol (Thermo Fisher) following the manufacturer's instructions. DNA contamination was removed with a DNA removal kit (Thermo Fisher), and cDNA was obtained using the following reagents: Reverse Transcriptase M-MLV (Promega, Madison, WI, USA), Oligo dT (Promega), dNTPs (Invitrogen, Waltham, MA, USA), and RNAsin (Promega). The program used for cDNA obtention was: 5 min at 65 • C, 1 h at 42 • C, and 10 min at 70 • C. hBD mRNA expression was evaluated with the following primers: hBD1 Fw: 5 -ATGGCCTCAGGTGGTAACTTTC, Rev: 5 -TCGGGCAGGCAGAATAGAGA; hBD2 Fw: 5 -CCTGGAACAAAATGCTGCAA, Rev: 5 -ACATGTCGCACGTCTCTGATG; hBD3 Fw: 5 -GTGAAGCCTAGCAGCTATGAGGAT, Rev: 5 -TGATTCCTCCATGACCTGGAA; and β-actin Fw: 5 -CTGGAACGGTGAAGGTGACA, Rev: 5 -TGCGTTGTTACAGGAAGTCCCTT, as housekeeping gene. The PCR program used was: 15 s at 95 • C, 30 s at 60 • C, and 40 cycles of 30 s at 72 • C. For normalization, we obtained the Ct for the gene of interest and the β-actin gene for each patient. The expression of the gene of interest relative to β-actin was expressed as 2 −∆∆CT . The control patients were used as the reference group.

Human Cell Culture
HeLa cells were cultivated at 37 • C and 5% CO 2 in DMEM (Gibco, Waltham, MA, USA) supplemented with 10% FCS (Natocor, Cordoba, Argentina), 1% L-glutamine (Gibco), and 1% penicillin-streptomycin (Gibco). A total of 2.5 × 10 5 cells/well were incubated in a 24-well plate (CellStart, Greiner one, Frickenhausen, Germany). After 24 h, the cells were incubated for 4 h with different C. albicans strains at a 1:1 or 5:1 Candida: cell ratio or with the following stimuli: LPS (Sigma-Aldrich) 1 µg/mL or 10 µg/mL, Pam3CSK4 100 ng/mL (Invivogen, San Diego, CA, USA), Zymosan 20 µg/mL (Sigma-Aldrich, St Louis, MO, USA), IL-1β 10 ng/mL (R&D Systems, McKinley Place, Shorewood, MN, USA), and Heat-Killed Candida (HKC) at a 1:1 ratio. After the 4 h simulation/infection of HeLa cells, the supernatants were collected, and cold PBS was added to the monolayer for 15 min in order to detach the cells. Different strategies were used to study hBD1 regulation, including a transwell culture system (Costar, Cambridge, MA USA). C. albicans was placed in the upper compartment of the transwell and was not in contact with the HeLa cells. Another strategy used was the addition of 625 µM Acetylsalicylic Acid (ASA; Sigma-Aldrich), a lipase inhibitor, and 1 µg/mL Pepstatin A (Sigma-Aldrich), a SAP inhibitor.

Statistical Analysis
Data were expressed as means ± SEM. Prior the statistical analysis, the test of data normality (Shapiro-Wilk test) was performed. Differences between groups means or ratios were assessed using one-way or two-way ANOVA followed by the Bonferroni test for multiple comparisons. A p-value < 0.05 was considered statistically significant. GraphPad Prism version 6.0 for Windows (GraphPad Software, San Diego, CA, USA) was used to perform the statistical analysis. All experiments were repeated at least twice to test the reproducibility of the results.

mBD1 Expression Is Upregulated in the Vaginal Tract During VVC
In order to characterize the kinetic expression profile of mBD1 during VVC, we evaluated the pattern of expression of this peptide using a well-characterized murine model of acute VVC (AVVC) in which C57BL/6 mice were i.v. infected with C. albicans [23]. Briefly, a group of mice was inoculated with C. albicans ATCC-36801 (infected group) and treated with estrogen, another group received PBS and was treated with estrogen (uninfected/estrogenized group), and the third group was neither treated nor infected (unestrogenized group). At D2 and 4 pi, the fungal burden showed similar colonization, while a decrease was observed at the end of the study ( Figure 1A). At D2 pi, a significant increase in the percentage and absolute number of mBD1+ cells in VL was found in infected animals when compared to the other two groups ( Figure 1B,C). At D4 pi, the percentage of mBD1+ cells was similar to that of D2 pi, and no significant differences were observed when compared to the uninfected and unestrogenized animals. The absolute number and percentage of mBD1+ cells increased in the infected and uninfected groups at D8 pi, showing that the estrogen treatment was able to modulate mBD1 expression in the vaginal tract. In the intact mucosa of both negative control groups, a basal and homogeneous distribution of mBD1 expression was observed in the stratified epithelium, with a slight increase after exposure to estrogen treatment ( Figure 1D, left panels). In C. albicans-infected mice, progressive immunostaining in epithelial cells (EC) was observed. The PMNs infiltrate and corneal microabscesses showed a strong mBD1 positive reaction. These results indicate that the expression of this constitutive peptide is upregulated in the vaginal tract during C. albicans infection.

mBD1 Expression Is Differentially Regulated after Infection with a C. albicans Strain Recovered from an RVVC Patient
To evaluate whether the pathogenicity of the fungal strain can modulate mBD1 expression, mice were infected with a C. albicans strain isolated from an RVVC patient or with C. albicans ATCC-36801 strain, and mBD1 expression was evaluated. When their virulence profiles were compared, C. albicans RVVC proved to be more virulent than the collection strain. The fungal burden (D2 pi) ( Figure 2A) and IL-1β amounts ( Figure 2B) were higher when the mice were inoculated with the clinical isolate. Moreover, a significant increase in the percentage of mBD1+ cells was observed in VL from animals infected with C. albicans RVVC when compared with the other groups (D2 pi) ( Figure 2C). At D4 pi, the percentage and the absolute number of mBD1+ cells were significantly increased compared to both negative controls. The results showed a distinct profile of AMP induction depending on the strain origin, with an exacerbated mBD1+ expression when the infection was induced with the more virulent strain.

mBD1 Expression Is Differentially Regulated after Infection with a C. albicans Strain Recovered from an RVVC Patient
To evaluate whether the pathogenicity of the fungal strain can modulate mBD1 expression, mice were infected with a C. albicans strain isolated from an RVVC patient or with C. albicans ATCC-36801 strain, and mBD1 expression was evaluated. When their virulence profiles were compared, C. albicans RVVC proved to be more virulent than the collection strain. The fungal burden (D2 pi) ( Figure 2A) and IL-1β amounts ( Figure 2B) were higher when the mice were inoculated with the clinical isolate. Moreover, a significant increase in the percentage of mBD1+ cells was observed in VL from animals infected with C. albicans RVVC when compared with the other groups (D2 pi) ( Figure 2C). At D4 pi, the percentage and the absolute number of mBD1+ cells were significantly increased compared to both negative controls. The results showed a distinct profile of AMP induction depending on the strain origin, with an exacerbated mBD1+ expression when the infection was induced with the more virulent strain. The PMN recruited into the vaginal lumen during AVVC contributed to tissue damage and the local inflammation associated with vaginitis symptoms. The infection generated by the clinical strain did not produce variations in the percentage and absolute number of PMNs (Gr-1+ cells) recruited to the vaginal lumen when compared with the animals infected with C. albicans ATCC-36801 ( Figure 2D, left and right panels). However, the percentage of Gr-1+ mBD1+ cells was significantly higher (D2 pi) in mice infected with C. albicans RVVC strain than in the other experimental groups ( Figure 2E). A similar result was observed when the absolute number of Gr-1+ mBD1+ cells were compared. These data indicate that the number of PMN recruited to the vaginal lumen was independent of the C. albicans strain used. Nevertheless, the percentage of mBD1-producing PMNs was higher when the infection was induced with a clinical strain.
age and the local inflammation associated with vaginitis symptoms. The infection generated by the clinical strain did not produce variations in the percentage and absolute number of PMNs (Gr-1+ cells) recruited to the vaginal lumen when compared with the animals infected with C. albicans ATCC-36801 ( Figure 2D, left and right panels). However, the percentage of Gr-1+ mBD1+ cells was significantly higher (D2 pi) in mice infected with C. albicans RVVC strain than in the other experimental groups ( Figure 2E). A similar result was observed when the absolute number of Gr-1+ mBD1+ cells were compared. These data indicate that the number of PMN recruited to the vaginal lumen was independent of the C. albicans strain used. Nevertheless, the percentage of mBD1-producing PMNs was higher when the infection was induced with a clinical strain.

Intravaginal Cytokines Profile and Virulence Factors of Clinical Isolates from Patients with AVVC or RVVC
To understand the role of hBD in patients with AVVC or RVVC, we evaluated clinicalrelated parameters, characteristics of the etiological agent, and local production of cytokines in both groups of patients. We found that the percentage of risk factors was higher in the AVVC than in the RVVC population (Table 1) and that C. albicans was the prevalent species in both forms of the disease. C. albicans strains recovered from patients showed a similar frequency of SAP-producing isolates, but a higher enzyme activity was detected in the AVVC group. The number of LIP-producing strains was lower among the RVVC isolates than in AVVC, but the enzyme activity was similar in both groups. The symptoms of Candida vaginitis suggest an active vaginal inflammatory process resulting from an imbalance in the local microenvironment. IL-1β, IL-6, TNF-α, IL-17, IL-22, IL-23, and TGF-β concentrations were determined in CVL from Control, AVVC, and RVVC groups (Figure 3). Patients with RVVC showed higher levels of IL-1β and IL-6 in CVL than Control individuals and higher IL-6 levels than patients with the acute form of the mycosis. TNF-α was undetectable in all the evaluated samples. Moreover, while TGF-β levels showed some variations between the groups, these changes were not statistically significant. Furthermore, IL-23 significantly decreased in CVL from patients with RVVC compared to those with AVVC, and both IL-17 and IL-22 were undetectable in all the groups studied. These results indicate that the cytokine profile of RVVC patients is characterized by high amounts of IL-1β and IL-6, low concentration of IL-23, and undetectable levels of IL-17, IL-22, and TNF-α.

Study of AMPs of the BD Family in Patients with AVVC and RVVC
The modulation of mBD1 in the vaginal mucosa in response to Candida infect gests an active role of this AMP during AVVC. To corroborate this hypothesis plored hBD transcript and protein expression in CVL cells of the AVVC and groups. Interestingly, the hBD1 mRNA level was increased in the AVVC group co to the Control and RVVC groups ( Figure 4C). The immunostaining assay show cells from healthy individuals exhibited a basal hBD1 expression, while EC from with AVVC showed a more homogeneous hBD1 expression in cell cytoplasm ( Fig  Further, the expression of hBD1 was significantly decreased in RVVC samples co with those from healthy individuals and patients with AVVC ( Figure 4B).
We also evaluated hBD2 and hBD3 mRNA expression in CVL cells of the d groups. Although the levels of the hBD2 transcript showed no significant chan tween the tested patients (data not shown), hBD3 mRNA was upregulated in RVV ples compared to those of the Control group ( Figure 5C). Similar to that observed stitutive hBD1 expression, hBD3 level in the RVVC group was significantly de when compared to the AVVC group ( Figure 5A,B).

Study of AMPs of the BD Family in Patients with AVVC and RVVC
The modulation of mBD1 in the vaginal mucosa in response to Candida infection suggests an active role of this AMP during AVVC. To corroborate this hypothesis, we explored hBD transcript and protein expression in CVL cells of the AVVC and RVVC groups. Interestingly, the hBD1 mRNA level was increased in the AVVC group compared to the Control and RVVC groups ( Figure 4C). The immunostaining assay showed that cells from healthy individuals exhibited a basal hBD1 expression, while EC from patients with AVVC showed a more homogeneous hBD1 expression in cell cytoplasm ( Figure 4A). Further, the expression of hBD1 was significantly decreased in RVVC samples compared with those from healthy individuals and patients with AVVC ( Figure 4B).
We also evaluated hBD2 and hBD3 mRNA expression in CVL cells of the different groups. Although the levels of the hBD2 transcript showed no significant changes between the tested patients (data not shown), hBD3 mRNA was upregulated in RVVC samples compared to those of the Control group ( Figure 5C). Similar to that observed for constitutive hBD1 expression, hBD3 level in the RVVC group was significantly decreased when compared to the AVVC group ( Figure 5A,B).

Regulation of hBD1 in Epithelial Cells of Female Genital Tract
In order to elucidate the signals involved in the modulation of hBD1 expression in EC of the female genital tract, we used a HeLa cell line in vitro model [26,30] and determined hBD1+ cells percentage in basal condition and after exposure to different stimuli. Constitutive hBD1 expression was observed in the basal condition, and a significant increase in hBD1+ cells percentage was found after stimulation with 10 ug/mL of LPS (TLR4 agonist) and recombinant hIL-1β ( Figure 6A). Nevertheless, Pam3CSK4 (TLR1/2), Zymosan (TLR2 and Dectin-1 agonists), and HKC were not able to modulate hBD1 expression. Further, the ability of different viable C. albicans strains to regulate hBD1 expression was evaluated. We found a significant decrease in hBD1 expression in EC cultured with the C. albicans clinical isolates FKS1-R1361R/H and RVVC at a 5:1 ratio ( Figure 6B); however, no changes were observed after exposure to C. abicans SC5314. These results show that fungal strain origin and inoculum size are relevant factors during C. albicans interaction with host EC.
The next step was to establish whether direct fungal contact or released soluble factors were involved in the described phenomenon. Thus, HeLa cells were cultured with C. albicans RVVC using a transwell system. The results showed that, in the presence of fungal released molecules alone, the percentage of hBD1+ cells significantly decreased compared to the basal condition ( Figure 6C). Considering the role of C. albicans SAP family in vaginitis [31,32], previous results about fungal LIP [26,33], and the enzymatic activity we have detected in clinical isolates (Table 1), HeLa cells were also cultured in the presence of SAP and LIP inhibitors. Interestingly, hBD1 decreased expression was restored to its basal values after exposure to both drugs, supporting the notion that SAP and LIP are involved in the downregulation of hBD1, both together and separately.

Discussion
BDs are an important alarmin family within the AMPs molecules involved in the protection against pathogens in mucosal tracts [10,12,13]. Despite these properties, studies on their roles during C. albicans vaginal infection in animal models and especially in patients with the acute or recurrent form of the mycosis are limited. Using a murine AVVC model, we described the kinetic response of mBD1 expression in the vaginal tract during C. albicans ATCC-36801 infection. The number of mBD1+ cells in the lumen at D2 pi was significantly increased, and mBD1+ protein expression was observed in resident cells and recruited PMNs. These results were confirmed in an in vivo infection model using a C. albicans strain isolated from RVVC. We found that the fungal load and intravaginal level of IL-1β remained high, mBD1+ cells increased, and the percentage of Gr-1+ mBD1+ cells was nearly 2.8-fold higher than that found in C. albicans ATCC-36801-infected animals. These findings showed that, after acute exposition to C. albicans, constitutive mBD1 expression was early increased and that this response was correlated with the infection severity. In line with this result, we had previously reported that lack of TLR2 receptor signaling [23], which predisposes mice to massive C. albicans invasion of the vaginal tract, leads to an exacerbated expression of mBD1+ and IL-1β, indicating that there is a strong correlation between both mediators and the immunopathological response during VVC.
BDs production depends on the cellular type, stimulus, environment, and context of the cells being stimulated [10,34]. Pahl et al. [16] have reported, in an OE21 esophageal cell model, that the blockade of IL-1β receptor-dependent signaling strongly reduces the induction of hBD, and they also describe a signaling loop of IL-1β and hBD regulation. Several in vitro results have confirmed this report [35][36][37]. The IL-1β canonical pathway is regulated at the transcriptional level through fungal PAMPs recognition and protein secretion by the NLRP3 inflammasome [38]. The role of IL-1β during VVC has been well-documented in animal models [7,23,32,39], in works with Nlrp3 −/− mice [40], and in patients with NLRP3 polymorphisms [2,41]. In addition, the SAP family members are involved in IL-1β induction [31,32]. Our findings on murine AVVC and the effect of recombinant hIL-1β on hBD1 induction in in vitro cultures of epithelial cells of the genital tract agree with the available evidence about IL-1β and hBD1 induction in the vaginal tract.
Although it is accepted that the mucosal microenvironment regulates BDs expression, only a few studies have explored cytokines secretion during C. albicans infection in the female genital tract [31,[41][42][43]. Unfortunately, the groups of patients included in those studies are not comparable between them, nor with those of our study. We evaluated the concentration of different cytokines in CVL of patients with AVVC and RVVC and healthy individuals and observed that the CVL cytokine profile of women with RVVC was characterized by high levels of IL-1β and IL-6, low concentration of IL-23, and undetectable levels of TNF-β, IL-17, and IL-22. Regarding IL-23, Ryan et al. [44] have proposed that, during Chronic Mucocutaneous Candidiasis, low levels of this cytokine produce deregulation of the IL-6/IL-23/IL-17 axis and impact the ability to mount and sustain an efficient antifungal response. In our study, we observed low local production of IL-17 (<15 pg/mL) in both the acute and recurrent form of this mycosis, confirming that this cytokine is dispensable for the immunopathogenesis of VVC. These results agree with the notion that humans with inborn errors related to IL-17 immunity do not seem to be more susceptible to Candida vaginal infections [45][46][47]. Recently, Peter et al. [5] have demonstrated, in models of estrogen-induced VVC, that mice lacking IL-17RA, Act1, and IL-22 do not show increased susceptibility to Candida vaginitis. Taking together, these results highlight our understanding of the controversial role of Th17/IL-17 axis signaling in the immunopathogenesis of VVC [5,23,39,48].
Considering the inflammatory milieus observed during RVVC (higher IL-1β and IL-6), a novel and unsuspected result was the hBD1 expression observed in EC from patients with RVVC. Even though in the acute form of the infection hBD1 mRNA level was increased and its cytoplasmic expression presented homogeneous distribution, during RVVC, both hBD1 mRNA and protein expressions were severely downmodulated. Furthermore, hBD3 mRNA expression was upregulated, whereas its protein expression was significantly reduced, suggesting a phenomenon of posttranscriptional inhibition. These results demonstrated a significant reduction of constitutive and inducible members of the BD family in patients with RVVC. Host and fungal factors are involved in the outcome of vaginitis [48]. In agreement with our previous results [26,33] and those reported by other authors [32,33], we found that a high proportion of AVVC and RVVC clinical isolates showed active SAP and LIP secretion. Our in vitro VVC model approach provides data on the stimuli able to trigger hBD1 induction but also provides information on pathogen-EC interactions. The data linking strain origin and fungal burden suggest the existence of a critical threshold at which clinical strains can inhibit hBD1 expression. The use of enzyme inhibitors allowed us to determine that SAP and LIP were involved in this phenomenon. Considering the critical role of BDs on the mucosal surfaces, their decrease would constitute a highly efficient escape strategy for C. albicans. This finding could explain, at least in part, the low levels of both hBD1 and hBD3 in patients with RVVC, in a context of recurrent fungal overgrowth, invasion, active release of virulence factors [1,32,33,49], and high levels of inflammatory mediators, such as IL-1β, which in another context would stimulate BDs production. This fact and the deregulation of the IL-6/IL-23/IL-17 axis would contribute to C. albicans survival and persistence.
It has been shown that certain virulence factors released by pathogens are involved in the downregulation of hBD production; however, most of the existing evidence derives from studies on bacterial infections. In this respect, Patel et al. [37] found that gastric biopsies from patients infected with Helicobacter pylori have significantly decreased levels of hDB1 transcripts. Working with gastric epithelium cell lines and different bacteria strains, they proved that this effect depends on the pathogenicity of the strain and is mediated by H. pylori virulence factor CagA [29,35]. Chakraborty et al. [50] evaluated the modulation of hBD1 in human EC of the intestinal mucosa exposed to Vibrio cholerae and enterotoxigenic Escherichia coli. In both cases, essential virulence factors (choleric toxin and labile toxin) were responsible for hBD1 transcription inhibition. This phenomenon has also been described in virus [34]. The in vitro data obtained in our study demonstrate that fungal virulence factors can inhibit hBD1 protein expression in epithelial cells of the genital tract. Thus, our results and extend this conserved evolutionary strategy to fungus-host interaction.
This work identifies a novel role for mBD1 and hBD1 in host fungal defense in the female genital tract and in the pathogenesis of VVC. The knowledge provided on the mechanisms governing the interactions between C. albicans and its host is relevant for the proposal and development of new or combined treatment strategies against this mycosis. Informed Consent Statement: All the patients enrolled in the study provided written informed consent prior to participation.

Conflicts of Interest:
The authors declare no conflict of interest.