NLRP3 inflammasome is a key player in human vulvovaginal disease caused by Candida albicans

The expression of host inflammatory and Candida albicans putative virulence factors was studied in women with vulvovaginal candidiasis (VVC; twenty) or colonized by the fungus but asymptomatic (carriers; fifteen) or non-colonized asymptomatic (ten subjects). Overexpression of genes encoding NLRP3 and caspase-1 inflammasome components sharply differentiated VVC patients from asymptomatic colonized or non-colonized women. Inflammasome expression was coupled with neutrophils recruitment in the vagina of VVC women and IL-1β and IL-8 production. Both cytokines were present, though to a lower concentration, also in the vaginal fluid of colonized and non-colonized women. Secretory aspartyl proteinases (SAPs) and hyphae associated genes HWP1 and ECE1 were upregulated in VVC but with some differences among infected women. The most overexpressed SAP gene was SAP2, that correlated with neutrophils accumulation. Our data provide clinical evidence that the intracytoplasmic activation of NLRP3 inflammasome complex plays a critical, pathogenesis-relevant role in human VVC.


Vaginal inflammation. The distinctive presence of abundant PMN in the vaginal samples of subjects with
VVC suggested for an active vaginal inflammatory process. To corroborate this hypothesis, we first explored whether pro-inflammatory cytokines, such as IL-8 and IL-1β, which are critically involved in PMN recruitment and activation, were commonly present in the vaginal samples of symptomatic women. Both cytokines were indeed detected in the vaginal fluid of all women. However, their concentration was remarkably higher in symptomatic respect to asymptomatic subjects (mean concentrations of symptomatic and asymptomatic were 378.6 ± 38.2 vs 36.3 ± 9.6, p = 2.1 × 10 −9 for IL-1β and 416.7 ± 49.5 vs 78.3 ± 8.4, p = 4.2 × 10 −8 for IL-8, respectively) (Fig. 2). Very low to undetectable levels of IL-1β and IL-8 were found in the vaginal samples of the ten asymptomatic, non Candida-carriers subjects (Fig. 2) where no PMN were found (Fig. 1).
The above cytokines are strongly involved in the inflammatory response which mostly relies upon activation of NLRP3, a critical component of inflammasome complex that initiates the inflammatory response through the activation of caspase-1. Activated caspase-1 cleaves the pro-IL-1β to active secreted IL-1β, a key inflammatory mediator driving the host response to infection 14 . We therefore verified the expression of NLRP3 and caspase-1 in our vaginal samples. As shown in Fig. 3, both NLRP3 and CASP1 were found to be expressed in the vaginal samples of all symptomatic subjects, to a different degree depending on the subject tested, but not in those of asymptomatic, colonized women ( Fig. 3a and b). No expression of the above inflammasome components was either found in asymptomatic, non-colonized women (data not shown).

Secretory aspartyl proteinases.
It has long been suggested that one or more proteins of ten member of Saps family of C. albicans exerted a role in the immune-pathogenesis of VVC and RVVC 6,10,[15][16][17][18][19] . More recently, some Saps, particularly Sap2, Sap5 and Sap6 have been reported as inducers of inflammation in murine models, suggesting that they could mediate the pathogenesis of human vaginal candidiasis 5,6,8 . Therefore SAP1-10 expression was evaluated in all vaginal samples of symptomatic and asymptomatic subjects. No SAP1-10 expression was detected in vaginal samples which were negative for C. albicans (n = 10). Moreover, SAP1-10 were scarcely overexpressed and generally only in few of the asymptomatic Candida-carriers (Figs 4 and 5). In contrast, expression of each SAP was upregulated in at least some of symptomatic women (Figs 4 and 5). Overall, of the SAPs most studied in the immune-pathogenesis and inflammation context, SAP2 was upregulated in all patients, and to high level, in some of them (Fig. 4a). SAP1 and SAP3 were expressed in roughly one half of symptomatic women (Fig. 4a). The hypha-associated SAP such as SAP6 was overexpressed in a major part of symptomatic women; SAP4 and SAP5 were moderately overexpressed in roughly one half of the subjects (Fig. 4b). Of note, the cell wall SAP9 and SAP10, called yapsins 20 , and also SAP7 and SAP8 resulted overexpressed in symptomatic subjects (Fig. 5).
Hyphae-associated virulence genes. Hyphal cells of C. albicans appeared to be only a relatively minor component of the fungal forms present in the vaginal samples of women with VVC (see Fig. 1), and a typical hyphae-associated SAP gene, SAP5, was not the one mostly expressed in our patient cohort, while SAP6 was expressed in about 75% of symptomatic women (see Fig. 4b). Nonetheless, we were interested in the expression of hyphae-associated genes for the role usually assigned to the hyphal development in C. albicans pathogenesis and particularly in vaginal infection and inflammation 12,21 . Therefore, we examined the expression of two pathogenesis-relevant hyphae associated genes such as ECE1 and HWP1 in our vaginal samples. In particular, ECE1 was selected because it encoded a precursor of the recently identified candidalysin toxin 13 . As shown in Fig. 6 the expression of both the above C. albicans genes was negligible in the asymptomatic Candida-carriers. In contrast, both ECE1 and HWP1 were expressed to substantial levels in all vaginal samples of tested symptomatic women.

Discussion
VVC is a common infection in most of otherwise healthy women of childbearing-age, and in a quite remarkable proportion of them it develops as highly distressing chronic, recurrent forms (RVVC) 1,2 . These are usually caused by C. albicans but can occasionally be caused by other non albicans species such as C. krusei, C. tropicalis, C. glabrata 22 . C. albicans is a common type of fungus, often found in small amounts in the vagina, mouth, digestive tract, and on the skin without causing infection or symptoms, but can switch from the commensal state to a pathogenic one. In mouse models, experimental vaginal infection by C. albicans results in a strong inflammatory response with a marked leukocyte infiltrate essentially consisting of polymorphonuclear cells (neutrophils) and activation of caspase-1/inflammasome components 5,6 . These studies have also shown that vaginally recruited neutrophils are unable to eliminate or substantially restrict the fungus growth in the vagina, likely not due to the loss of intrinsic candidacidal capacity but to the presence of Candida-or host-derived inhibitors 11,23 . Eventually the constant presence of high numbers of fungal cells, the release of inflammation-inducing constituents, such as, for instance, Sap 6 and candidalysin 13,24 , together with the inflammatory cytokines actually released by the leukocytes and epithelial cells cause an acute inflammatory state. However, the extent to which mouse and other animal models of vaginal candidiasis are faithful representatives of human infection is unclear 4 .  and symptomatic (n = 20, identified by dots with different colors) women were centrifuged at 3000 rpm for 10 min, then cellular fractions were lysed and total RNA was extracted and retro-transcribed in cDNA. The expression levels of NLRP3 (a) and CASP1 (b) genes in asymptomatic and symptomatic women were calculated by comparative Ct method (2 −∆Ct formula) after normalization with GADPH gene. The results are from triplicates samples of each subject and the statistical significance was determined with Mann Whitney U test. The dashed line denotes the cutoff at which the overexpression of NLRP3 and CASP1 genes was defined as >2 times the value of GADPH expression and the number of women expressing genes is indicated in the brackets. *p = 3 × 10 −10 symptomatic vs asymptomatic women for NLRP3 and CASP1.
Unlike C. albicans, C. glabrata, unable to make hyphae and produce Sap in vitro 9 , did not elicit an inflammatory immunopathogenic response in a murine model of vaginitis 25 .
In this paper we addressed vaginal inflammation in women with VVC in comparison with asymptomatic C. albicans vaginal carriers and non-colonized, healthy women. Inflammation was defined by the presence of neutrophils, pro-inflammatory cytokines and expression of inflammasome components. The NLRP3 inflammasome is an intracellular receptor complex that plays a key role in most inflammatory diseases via activation of caspase-1 that leads to cleavage of proIL-1β to the biologically active IL-1β 26 . It has been recently reported that it could be also activated by other non albicans species such as C. parapsilosis 27 .
Polymorphisms of, and presence of variable number tandem repeats in, the NLRP3 gene have been variably associated to VVC or RVVC 28,29 but to our knowledge, in none of the previous studies the expression of NLRP3 inflammasome and caspase-1 components in women with VVC has been detected and quantified. We here report a rather sharp difference between asymptomatic carriers and symptomatic infected women in that only the latter showed evidence for consistent NLRP3 inflammasome and CASP1 gene overexpression in their vaginal cells. This finding rather than the fungal burden or even the presence of cytokines themselves appears to be a true landmark of VVC. In fact, the difference between the CFUs detected in the vagina of VVC or asymptomatic carriers was relatively low (only roughly three times) and cytokines such as IL-1β and IL-8 were also consistently present, though to a lower concentration, also in the vaginal fluid of asymptomatic vaginal carriers. Low levels of IL-1β and IL-8 were also present in the vaginal fluid of healthy, non-colonized women who, as the asymptomatic carriers, had no expression at all of CASP1 and NLRP3 inflammasome. These data provide evidence that NLRP3 inflammasome expression is also a consistent marker of VVC in humans as it appears to be in two different murine models 5,6 . It is unclear whether vaginal epithelial cells are those uniquely expressing NLRP3 or the neutrophils could contribute to that expression, also in view of their resistance to apoptosis in an inflamed, cytokine-rich medium 30 . Further studies will address this interesting aspect. Our recent investigations have shown that Sap2 and Sap6 are likely involved in inducing inflammatory response by human monocytes, macrophages, and dendritic cells in vitro, through activation of NLRP3 inflammasome and induction of different caspases 31,32 . In a more recent paper, by using a murine model 6 we demonstrated that several Saps, particularly Sap2, cause a sort of aseptic inflammatory response in the mouse vagina in the absence of fungal cells. Since anti-Sap2 antibodies inhibited vaginal inflammation caused by C. albicans, it was proposed that one or more Saps could be the inflammation mediator of fungal disease. For the above reasons, we further studied the expression of the ten genes coding for Saps family in women with VVC and in fungus carriers. While these latter had very faint expression of some of the above genes, a variable but in some VVC women elevated overexpression of some of SAPs genes was detected. In particular, SAP2 expression was variably though consistently upregulated in the vaginal samples of all VVC patients. Other most frequently upregulated SAPs were SAP6, SAP4, SAP5 and SAP9 in this ranking order. SAP6 was overexpressed in 15 samples over 20 tested, SAP4 in 12 samples, SAP5 and SAP9 in 11 samples. Comparison with quantitative SAPs genes expression to other studies is impossible because of the differences in women recruitment, different case definition and times of VVC disease. Nonetheless, our data are qualitatively similar to those reported by Naglik and collaborators 10 with the possible exception of SAP5 and SAP10 genes expression that was detected only in some of VVC patients.
The relatively infrequent overexpression of the SAP5 gene, that is known to be hyphae associated, could be due to the relatively minor abundance of hyphal cells in the vaginal samples of our VVC patients. The fungal morphology of our samples showed a consistent presence of yeast with a lower proportion of some hyphae and pseudohyphae in symptomatic patients, with never massive presence of hyphae. This could be due to the exclusion of pregnant women from our cohort. In other studies (data not shown here) we noted that hyphae were much more consistently present in vaginal samples from pregnant VVC and RVVC subjects (unpublished data). In agreement with data by others 10 , no or limited expression of any SAPs gene was detected in non-VVC asymptomatic patients.
Hyphal growth of C. albicans is recurrently advocated to play an important role in mucosal infection by this fungus 12,21 . Although pseudohyphal and hyphal cells were less abundant than yeast cells in the vaginal samples of our symptomatic patients, we were interested in measuring the expression of putative hyphae-associated genes in our patients. To this purpose, among all the hyphae-associated genes we selected two genes, the HWP1 which codes for a fungal cell wall protein, required for hyphal development 33 , and ECE1 which is highly expressed by hyphae during infection of epithelial cells 13,34 . Hwp1 and Ece1 displayed a key role in two phases of C. albicans pathogenesis: adhesion and tissue damage. During C. albicans mucosal infection the fungal interaction to epithelial cells promoted the formation of hyphae which strengthen fungal adhesion. This link was mediated by adhesins expressed on hyphae such as Hwp1 35 .
We found that both HWP1 and ECE1 were overexpressed in symptomatic patients, but not or very low so in asymptomatic carriers. This is in apparent contrast with previous studies reporting rather similar expression of HWP1 gene in both symptomatic and in asymptomatic women 12 . ECE1 is one of hyphal genes in C. albicans which has been identified in 1990 36 . This gene codes for the recently discovered toxin, named candidalysin 13 . The candidalysin is reported to directly damage host epithelial membranes, to trigger a danger response signaling pathway and activate epithelial immunity 13 . We found upregulated expression of ECE1 in all VVC patients tested, while not in asymptomatic carriers. To our knowledge, this is the first demonstration that ECE1 is overexpressed during human vaginal candidiasis.
This study has some limitations concerning the non elevated number of patients examined, the exclusion of non-C. albicans infected VVC subjects and the lack of quantitative distinction of the different C. albicans forms of growth in the vaginal samples of VVC subjects. Despite this, the data overall suggest that NLRP3 inflammasome expression and the consequent pro-inflammatory cytokine cascade play a central role in VVC. They also indirectly confirm that the overexpression of one or more SAPs and possibly other fungal components, including the recently discovered candidalysin peptide from the Ece1 protein 13 , rather than the fungal burden per se, could play an important role in triggering the inflammatory cascade. These data strengthen some mechanistic similarity between VVC and vaginal inflammation in mouse models 37 . However, if this similarity is also applicable to the chronic, recurrent forms of vaginal candidiasis remains an open question.

Methods
Subjects. Fortyfive non-pregnant, non-diabetic women, 19 to 53 years old, attending the microbiological diagnostic service of the University Hospital Santa Maria della Misericordia, Perugia (Italy) over the period from February 2016 to April 2017 were consecutively enrolled in this study. This cohort included C. albicans-colonized symptomatic (n = 20), C. albicans-colonized asymptomatic (n = 15) and C. albicans-non colonized asymptomatic (n = 10) women. Prior to enrollment, each subject answered a questionnaire indicating their health status and current symptoms of vaginal disease. A case of VVC due to C. albicans was defined by the isolation of C. albicans from the vaginal sample and the presence of at least two of the following signs and symptoms: vaginal discharge, itching, burning and dyspareunia. None of the recruited women had RVVC as indicated by the absence of documented or woman-reported, repeated VVC episodes per year. All women signed an informed consent in accordance with the Declaration of Helsinki. Local Ethical Committee CEAS (Comitato Etico delle Aziende Sanitarie, Umbria, Italy) approval was received for the whole study (VAG1 n. 2652/15). All methods were performed in accordance with the relevant guidelines and regulations. Samples collection. A vaginal swab was taken from each participant and soaked in 1 ml of saline. After sampling, the vaginal swab was plated on CHROMagar TM Candida (VWR International p.b.i., Milan, Italy) at 37 °C for 48 h to evaluate the vaginal colonization by C. albicans. The presence of C. albicans was also confirmed by MALDI-TOF test (Biomérieux S.A. France). The vaginal fluid was serially diluted and plated on CHROMagar TM Candida (VWR International p.b.i.) to quantify the vaginal fungal burden 7 . Subsequently, the vaginal fluid was centrifuged at 3000 rpm for 10 min, then cellular fraction was used for gene expression analysis and supernatant for cytokine production.

Neutrophil infiltration analysis. Vaginal samples were examined under light microscope (Olimpus,
Milan, Italy) to evaluate the presence of C. albicans in yeast or hyphal form and neutrophils (PMN) by their morphology after staining with Papanicolaou technique 7 . The number of PMN was counted in four fields at x400 magnification and expressed as average number of PMN/field. The PMN score was graded on a scale from 0 to 2: 0, 0 PMN/field; 1, 1 to 10 PMN/field; 2, 11 to 40 PMN/field. The C. albicans-colonized symptomatic women showed at least two of the following symptoms: vaginal discharge, itching, burning, dyspareunia, associated with PMN infiltration score 2 and C. albicans isolation from the vaginal sample.
The C. albicans-colonized asymptomatic women showed C. albicans isolation from the vaginal sample, PMN score 1 but not clinical symptoms.
The C. albicans-non colonized women did not show C. albicans isolation from the vaginal sample, PMN infiltration (score 0) and clinical symptoms.
Cytokine Production. The supernatants of vaginal samples were collected and tested for IL-1β and IL-8 production by specific ELISA assays (all from eBioscence, San Diego, CA). Cytokine titers were calculated relative to standard curves. Total RNA was extracted and retro-transcribed by using the Moloney murine leukemia virus reverse transcriptase reaction (M-MLV RT), as described in the manufacturer's instructions. cDNA concentration was determined using a spectrophotometer. Human GADPH, NLRP3, CASPASE1 (CASP1) and C. albicans ACT1, SAP1, SAP2, SAP3, SAP4, SAP5, SAP6, SAP7, SAP8, SAP9, SAP10, ECE1 and HWP1 gene expression was detected by using primers reported elsewhere and showing similar capacity and efficiency in detecting expression of the above genes 10,12,13,38,39 . Real-time PCR (quantitative PCR) was performed in 96-well PCR plates (Thermo Scientific, Waltham, MA USA) using SYBR green (BioRad, Milan, Italy). For real-time PCR reaction 200 ng of cDNA was used. All samples were measured in triplicates. The expression levels of SAP1-10, ECE1, HWP1, NLRP3 and CASP1 genes in asymptomatic and symptomatic women were calculated by comparative Ct method (2 −∆Ct formula) after normalization with ACT1 for C. albicans genes and GADPH for human genes [40][41][42] . Amplification conditions were the same used for ACT1, SAP1-10, ECE1, HWP1, NLRP3, CASP1 and GADPH genes: 3 min at 95 °C, 40 cycles of 10 sec at 95 °C and 30 sec at primer specific temperature. The experiments were performed using Applied Biosystems 7300 (Thermo Scientific). Overexpression of relevant genes was defined as >2 times the value of housekeeping gene expression.

Quantitative analysis of
Statistical Analysis. Results reported in the dot plot graphs were from triplicate samples of all vaginal fluids; results reported in the bar graphs were the mean ± SEM from triplicates samples of 5 asymptomatic or symptomatic vaginal fluids.
Quantitative variables were tested for normal distribution by using SigmaPlot 12.5 program. For CFU count the quantitative variables were compared by means of Student's two-tailed t test. For other determinations the quantitative variables were compared by Mann Whitney U test.
Values of p < 0.05 were considered significant.