Novel Mechanism behind the Immunopathogenesis of Vulvovaginal Candidiasis: “Neutrophil Anergy”

ABSTRACT

For over 3 decades, investigators have studied the pathogenesis of vulvovaginal candidiasis (VVC) and recurrent VVC (RVVC) through clinical studies and animal models. While there was considerable consensus that susceptibility was not associated with any apparent deficiencies in adaptive immunity, protective immune mechanisms and the role of innate immunity remained elusive. It was not until an innovative live-challenge design was conducted in women that a fuller understanding of the natural history of infection/disease was achieved. These studies revealed that symptomatic infection is associated with recruitment of polymorphonuclear neutrophils (PMNs) into the vaginal lumen. Subsequent studies in the established mouse model demonstrated that infiltrating PMNs were incapable of reducing the fungal burden, which supported the hypothesis that VVC/RVVC was an immunopathology, whereby Candida and the host response drive symptomatic disease. Further studies in mice revealed the requirement for C. albicans hyphae and identified pattern recognition receptors (PRRs) and proinflammatory mediators responsible for the PMN response, all of which are critical pieces of the immunopathogenesis. However, a mechanism explaining PMN dysfunction at the vaginal mucosa remained an enigma. Ultimately, by employing mouse strains resistant or susceptible to chronic VVC, it was determined that heparan sulfate (HS) in the vaginal environment of susceptible mice serves as a competitive ligand for Mac-1 on PMNs, which effectively renders the PMNs incapable of binding to Candida to initiate killing. Hence, the outcome of symptomatic VVC/RVVC is postulated to be dependent on a PMN-mediated immunopathogenic response involving HS that effectively places the neutrophils in a state of functional anergy.

INTRODUCTION

Vulvovaginal candidiasis (VVC) is an extremely common mucosal infection, caused primarily by Candida albicans and characterized by itching, burning, pain, and redness of the vulva and vaginal mucosa, often accompanied by vaginal discharge (1). It is estimated that 75% of otherwise healthy immunocompetent women of childbearing age will experience primary (episodic) VVC at least once in their lifetime (2). Although treatment of primary VVC with antifungals is usually successful, approximately 5 to 8% of afflicted women will suffer from recurrent VVC (RVVC), characterized by four or more symptomatic episodes per year often requiring continual (maintenance) antifungal therapy (1). Predisposing factors for primary VVC include high-estrogen oral contraceptive use, hormone replacement therapy, antibiotic usage, and uncontrolled diabetes mellitus (2). Importantly, disruption of the vaginal microbiota can also contribute to this complex disease (3–5). RVVC is considered idiopathic with no identified predisposing factors, although the mechanisms of VVC and RVVC pathogenesis are likely identical (1). Importantly, VVC and RVVC are not associated with immunodeficiency but instead are associated with a vigorous local acute inflammatory response. This minireview covers the history of the investigations of host defense in VVC that ultimately led to the elusive mechanism of the immunopathogenic response.

HISTORY OF INVESTIGATIONS TO IDENTIFY PROTECTIVE OR DEFICIENT IMMUNE MECHANISMS IN VVC

Despite its high prevalence worldwide and ongoing investigations for more than 3 decades using clinical studies as well as animal models (mouse, rat, and nonhuman primates [6, 7]), knowledge about host defense mechanisms against VVC/RVVC has been largely elusive. Among all models employed, a well-established estrogen-dependent mouse model has proven to be most useful based on the fact that the experimental infection closely parallels the chronic nature of the disease in women. Accordingly, the mouse model has been a valuable tool to dissect the host response and obtain information that is translatable to the human disease. Following the current dogma at the time, susceptibility to Candida vaginitis was long believed to result from defects in the adaptive immune response similar to those in other forms of mucosal candidiasis (oral, chronic mucocutaneous, or gastrointestinal) in which susceptibility has been shown to be T-cell dependent (8–14). However, numerous clinical studies examining women with RVVC showed no humoral or cell-mediated immune deficiencies (15–17). This was supported further by there being no increased prevalence of VVC/RVVC in HIV-positive women with reduced numbers of CD4⁺ T cells (1, 18–20). This clinical evidence was consistent with results from the mouse model that revealed no obvious protective roles for local or systemic adaptive immunity (cell mediated or humoral) (21–26), although certain epitope-specific antibodies characterized as “protective antibodies” have been efficacious in animal models (27–31). Even attempts at using gene therapy that induced Th1-type cytokines (a protective response in most other forms of candidiasis) at the vaginal mucosa failed to provide any level of protection (32). This was explained in part by relatively high production of immunoregulatory factors (e.g., transforming growth factor β [TGF-β], regulatory T cells [Tregs], and Υ/δ T cells) in the vagina, which may limit local cell-mediated immunity as a tolerance mechanism (33, 34).

In line with inherently suppressed immune reactivity in the female reproductive tract, the lack of a strong role for adaptive immunity during Candida vaginitis is not surprising yet is very distinct from other forms of candidiasis. On the other hand, innate immunity, not generally considered a major component of immune tolerance, was a formidable alternative prospect as a protective mechanism. However, studies revealed that classical innate immune cells, such as macrophages, dendritic cells (DCs), and polymorphonuclear neutrophils (PMNs), were either only scarcely present during infection (DCs and macrophages) (35) or, if present (PMNs), had no apparent effect on fungal burden (36–38). Rather, a more unconventional yet highly prominent innate immune cell type at mucosal sites, epithelial cells, became the focus of studies investigating protective mechanisms. Indeed, vaginal epithelial cells appeared to provide a modest yet important first line of protection through a growth-inhibitory activity upon direct contact with Candida (39–41). This epithelial cell-mediated antifungal activity was shown to be both fungistatic and noninflammatory and likely serves as a primary defense mechanism to avoid unnecessary tissue damage and promote asymptomatic commensalism (42, 43). The precise mechanism for epithelial restraint is still somewhat unclear but is presumed to be associated with annexin A1, a multifactorial immunoregulatory receptor on epithelial cells, which was confirmed as the primary mediator in antifungal activity by oral epithelial cells in vitro (44). Considerable clinical evidence demonstrating greater inhibition of Candida growth by epithelial cells from women without RVVC than by epithelial cells from those with RVVC also suggested a role for vaginal epithelial cells in host defense (41, 45, 46).

Initially, the presence of PMNs in the animal model seemed perplexing. More than 70% of inoculated mice showed high levels of PMN infiltration, but this appeared to have no role in reducing the vaginal fungal burden (37). At the time, it was unclear what was initiating the ineffective innate response, but both estrogen and the organism were being considered independently and dependently. Interestingly, multiple mouse strains demonstrated the same natural history following inoculation (37). If PMNs were observed clinically in women with VVC/RVVC, this was quite dynamic as well with no obvious pattern. It was not until a pioneering study in woman volunteers challenged intravaginally with live C. albicans, who demonstrated a similar presence of vaginal PMNs commensurate with experiencing signs/symptoms of vaginitis, that a likely link between the PMNs and the infection was determined. Women with no VVC symptoms following the C. albicans challenge showed no evidence of PMN migration (36). These studies suggested for the first time that vaginitis symptoms were associated with PMN recruitment into the vagina and that organism burden alone was not predictive of disease. This prompted a number of additional studies focused on elucidating the mechanisms associated with the vaginal PMN response in the mouse model. Historically, in mice the density of C. albicans burden following inoculation was used as evidence of infection/disease because disease symptomatology (e.g., vulvar edema, irritation, and scratching) was difficult to assess or quantify. Subsequent studies indicated that a strong correlate of a symptomatic condition (disease), as opposed to asymptomatic colonization (commensalism), was the presence of PMNs (37). This revelation also provided the strongest evidence to date that VVC/RVVC was the result of an immunopathogenic response in which the host response, and not the fungal burden alone, was driving disease symptoms.

Despite the discovery of the immunopathogenic host response, the virulence determinants that govern C. albicans pathogenesis are equally important in the disease process. C. albicans expresses several virulence traits that contribute to VVC/RVVC pathogenesis, the most prominent of which is the morphological transition from yeast to hyphae. Most notably, strains of C. albicans defective in Efg1-driven hyphal formation displayed significantly reduced vaginal PMN migration, indicating a requirement for hyphae in the pathogenesis (47). Other hypha-associated virulence factors studied in vaginitis include the agglutinin-like sequence (ALS) genes that regulate mucosal tissue attachment (48–50) and biofilm formation on vaginal mucosa (51), which seemingly play minor roles in the immunopathogenic response. However, a newly identified secreted cytolytic peptide toxin (Candidalysin) encoded by C. albicans ECE1 (extent of cell elongation 1) plays a major role in disease pathogenesis (52). ECE1 was among the most highly expressed genes during murine vaginitis (53), and use of deletion mutants completely ablated the PMN-mediated immunopathogenic response, despite colonization equivalent to that for isogenic controls (54). Collectively, these studies highlight the importance of Candidalysin in the pathogenesis of mucosal C. albicans infections.

In addition to the strict requirement of C. albicans morphological transition from yeast to hyphae in the vaginal immunopathological response during murine VVC, there are several reports implicating C. albicans secreted aspartyl proteases (Saps) as potential inflammatory inducers during vaginitis. However, the role of specific Saps in vaginal immunopathogenesis has varied depending on the model systems (e.g., animal strains, infection stages, or reconstituted human epithelium [RHE] culture) (55–58). Similarly, in the most recent reports, one study showed a role for Escherichia coli-derived recombinant Sap2 and Sap6 in migration of human PMNs in vitro and vaginal inflammation in uninoculated mice (59). However, this was countered by a study which showed that overexpression of Sap2 or Sap5 in a hypha-defective C. albicans mutant failed to exacerbate the immunopathogenic response (60). Hence, the role of Saps in vaginal pathogenesis remains controversial.

The rat model of VVC has also contributed to the investigations on host defense. The rat model is similar to the mouse model in many ways, including a lack of C. albicans as part of the vaginal microbiota, neutral vaginal pH, and dependence on exogenous estrogen to initiate colonization (6). A major difference, however, is that the rats spontaneously clear the infection over a 2- to 3-week period, and prolonged colonization requires multiple inoculations (61–64). Moreover, histological analyses in these studies revealed abundant mononuclear infiltrates into the vaginal mucosa and showed no sign of PMN migration throughout infection. Hence, the clinical pattern of chronic vaginitis is not well simulated. Due to these limitations, rats may not be an ideal model for dissecting the immune response during VVC. Despite these issues, there has been a variety of reports identifying various putative protective host defense mechanisms using the rat model. These include protective humoral immunity via several antibody isotypes (29, 64), protective CD4⁺ T-cell responses (62, 63), roles for dendritic cells (65), and protection via a Sap2 antigenic vaccine delivered by virosomes (27). Thus, there has been a diverse array of studies in the rat model that largely counter the clinically paralleled observations in the mouse model.

THE PMN-MEDIATED IMMUNOPATHOGENIC RESPONSE IN VVC

Considerable information is now known regarding the mechanism associated with the aggressive vaginal PMN migration triggered by C. albicans hyphae. From a series of studies using the mouse model, the PMN response was shown to be initiated by C. albicans hypha-epithelial cell interactions, including production of Candidalysin and fungal recognition mediated via the pattern recognition receptors (PRRs) Toll-like receptor 4 (TLR4) and SIGNR1. Downstream signaling events lead to production of key inflammatory mediators, including S100A8 alarmin and interleukin-1β (IL-1β), via the NLRP3 inflammasome. These mediators initiate vigorous PMN chemotaxis to the vaginal cavity, where their activated state serves to amplify the response by increasing the local production of these same and other effectors (positive feedback loop) (37, 47, 53, 54, 66, 67). Failure of the migrating activated PMNs to clear Candida from the vaginal mucosa and subsequent death with release of granules, combined with the pseudoestrus environment that promotes Candida growth, confines the vaginal mucosa to a chronic inflammatory state. Figure 1 illustrates these events. This has been shown for a variety of mouse strains (22, 37, 68) that can collectively represent a chronic VVC-susceptible (CVVC-S) phenotype. Hence, the presence of PMNs is the hallmark of the inflammatory condition. This robust yet nonclearing PMN response is a unique and intriguing attribute of Candida vaginitis that counters the potent antimicrobial properties and protective roles for PMNs in other forms of candidiasis. Host damage, however, during the immunopathogenic response appears to be mediated primarily by Candida (hypha) rather than the infiltrating PMNs, as tissue damage measured by levels of lactate dehydrogenase (LDH) is similar in mice with PMNs or depleted thereof (neutropenic) (47).

FIG 1.

Proposed host-pathogen interactions leading to the immunopathogenic response in mice susceptible to chronic vulvovaginal candidiasis (CVVC). (A) Initiation of the PMN-mediated immunopathological response. Following inoculation with C. albicans, the hypha-inducing environment (i.e., elevated pH, increased estrogen) promotes epithelial signaling by Candidalysin and, together with fungal-epithelial cell recognition via the pattern recognition receptors (PRRs) TLR4 and SIGNR1, results in epithelial cell activation. Downstream signaling events lead to production of key inflammatory mediators, including S100A8 alarmin and IL-1β, via the NLRP3 inflammasome. Due to their potent chemotactic properties, the increased presence of the inflammatory mediators leads to robust recruitment of PMNs to the vaginal cavity, resulting in an acute inflammatory state. (B) Symptomatic/immunopathological state of the vaginal epithelium in a CVVC-susceptible environment. Recruited activated PMNs lose their antifungal killing capacity once reaching the vaginal lumen and fail to clear C. albicans. The ensuing fungal overgrowth continues to elicit the initial triggers by the vaginal epithelium, while the production of the inflammatory mediators is amplified (feedback loop) from the activated PMNs. Insufficient regulation of the inflammatory response enhances the immunopathogenic processes and, together with tissue damage by C. albicans, results in the symptomatic environment.

Another interesting caveat is that while type 17 responses are critical to neutrophil responses in Candida infections at other anatomical sites (oral and bloodstream), the role for type 17 responses in the vaginal neutrophil response remains controversial (66, 69). The strongest evidence to date has shown that in mice, deficiency in several cytokines encompassing the Th17 axis had no effect on the Candida-induced neutrophil response during pseudoestrus, with the mice exhibiting wild-type levels of vaginal PMNs and fungal burden. Hence, there appears to be little to no pathogenic or protective role of the type 17 response in modulating the PMN response under commonly used experimental VVC conditions (66). Together, these findings suggested that PMN dysfunction in the vaginal cavity was due either to inherent killing defects in the migrating population of PMNs or to factors present in the vaginal environment that inhibit their ability to function normally.

Clinically, the immunopathogenic response is hypothesized to be triggered by the sensitivity of the vaginal epithelium to Candida hyphae; epithelial cells of women susceptible to VVC or RVVC are considered inherently sensitive to Candida, resulting in robust inflammation, whereas cells of women with no history of VVC are less sensitive to Candida and therefore exhibit a lessened or absent response. Observations from the clinical live-challenge studies support this theory; extremely high inocula (>10⁸) given to women with no history of disease either failed to colonize or colonized at low levels (36). Ultimately, the epithelial cell-triggered inflammatory response in susceptible women is considered to be dependent on a threshold level of Candida. Accordingly, it is proposed that in susceptible women, Candida stimulates epithelial cells to produce alarmins and proinflammatory cytokines to induce neutrophil migration, resulting in an inflammatory environment and accompanying symptomatic condition (37, 53). On the other hand, the epithelial cells in resistant women are not prone to the alarmin/cytokine response but instead respond in a noninflammatory manner (perhaps via interaction with annexin A1) that promotes static growth inhibition of C. albicans to maintain commensalism (37, 44).

ELUCIDATING THE MECHANISM OF PMN DYSFUNCTION IN VVC

PMNs are one of the primary components of the host innate immune defense against C. albicans infection (70–72). Antifungal activity by PMNs involves the interaction of Mac-1 (also termed α_Mβ₂, CD11b/CD18, or complement receptor 3) with C. albicans pH-regulated antigen 1 protein (Pra1p), which is present in both yeast and hyphal forms but expressed at a higher density on hyphae (73, 74). Mac-1-mediated PMN activation has been shown to promote fungal killing by formation of neutrophil extracellular traps (NETs) (75). Although this killing mechanism appears to be absent during VVC, the requirement for hyphae in the immunopathogenic response (47) suggests that Pra1p should be sufficiently expressed for optimal fungal recognition by PMNs during infection.

As a counter to the many strains of mice with a chronic VVC-susceptible phenotype (CVVC-S), CD-1 mice are uniquely incapable of maintaining C. albicans colonization following inoculation using commonly used inocula for CVVC-S mice (76, 77). The mechanism of fungal clearance that occurs in this strain had not been specifically identified but was thought to result from their nonresponsiveness to estrogen, which negatively impacted C. albicans adherence and colonization. On the other hand, recognizing the progression of experimental VVC in CVVC-S mice, it was equally possible that clearance was immune based and represented a chronic VVC-resistant (CVVC-R) phenotype. This translated into a major opportunity to begin to dissect the immunopathogenic response by exploiting mouse strains susceptible and resistant to chronic vaginal C. albicans colonization.

The first studies using the two phenotypically distinct mouse strains focused on the progression of the experimental vaginal infection relative to putative immune reactivity. Accordingly, the C3H/HeN and CD-1 strains were used to broadly define CVVC-S and CVVC-R phenotypes, respectively. Following inoculation with C. albicans, the two strains showed comparable colonization over the first 48 h, with similar migration of PMNs into the vaginal cavity. In contrast to the C3H/HeN mice, which maintained the typical vaginal fungal burden over a 2-week period with high numbers of vaginal PMNs, the CD-1 mice showed an incremental reduction in fungal burden over the same period, commensurate with reduced presence of PMNs. Microscopic examination of vaginal smears from CD-1 mice showed PMNs in close physical contact with Candida hyphae, whereas smears from C3H/HeN mice showed qualitatively less fungus-neutrophil contact. Fungal viability staining showed mainly nonviable Candida in the CD-1 mice, while most Candida cells were viable in C3H/HeN mice (78). Thus, there was clear evidence for the first time that the reduction in vaginal fungal burden in CVVC-R mice was due to the PMN response and clearance by adequate antifungal activity. This was confirmed by PMN depletion studies in CD-1 mice that resulted in a steady-state fungal burden (78).

These data prompted a series of studies designed to investigate whether manipulation of host or microbial factors could promote PMN antifungal activity in C3H/HeN CVVC-S mice. Accordingly, the vaginal fungal burden was assessed when PMNs were elicited into the vaginal cavity pre-Candida challenge (via vaginal administration of recombinant chemotactic S100A8 alarmin) versus the typical Candida-initiated PMN migration, as well as employing different morphological forms of C. albicans (hypha deficient, hypha locked, and wild type) at standard and lowered inocula. In all cases, the vaginal fungal burden was unaffected by the presence of elicited PMNs (78). Hence, there was a clear dysfunction of the PMNs in the vaginal environment that could not be explained by the capacity to form hyphae or fungal load. These observations demonstrate that the presence of PMNs has no effect on C. albicans growth and colonization in the vaginas of CVVC-S mice. Together, these findings further strengthened the hypothesis that PMNs, irrespective of temporal or quantitative recruitment, are rendered defective in fungal clearance once localized to the vaginal cavity.

To further investigate the mechanism of PMN dysfunction observed in CVVC-S mice, studies were conducted to examine whether the PMN antifungal activity rendered seemingly defective in the vagina could be restored in an extravaginal environment. Accordingly, an in vitro killing assay was used with both elicited peritoneal and vaginal PMNs in standard killing assay medium (RPMI). Surprisingly, the elicited vaginal PMNs challenged with C. albicans yeast and hyphae exhibited a level of antifungal capacity equivalent to that of the elicited peritoneal PMNs. The reverse experiment was also conducted, where the two types of elicited PMNs were evaluated for fungal killing in vaginal-conditioned medium (VCM) (sterile preparations of vaginal lavage fluid from C3H/HeN [CVVC-S] mice lavaged with standard RPMI culture medium). VCM served as a useful tool to simulate the CVVC-S vaginal environment and further examine PMN function in vitro. In stark contrast to the results in standard medium, neither elicited peritoneal nor vaginal PMNs could exert adequate killing capacity against C. albicans yeast and hyphae in VCM. These key pieces of information suggested that there was an inhibitory factor(s) in the vaginal environment of CVVC-S mice that was impeding the antifungal activity of PMNs. As a confirmation that the factor was not unique to the C3H/HeN mouse strain, similar results were observed in C57BL/6 mice, in line with previous reports showing high susceptibility to vaginal colonization (37, 76, 77). In contrast, VCM from CVVC-R (CD-1) mice had no inhibitory activity on either elicited peritoneal or vaginal PMNs, indicating a lack of the inhibitory factor in the vaginal environment of CVVC-R mice. Together, these observations support the hypothesis that an inhibitory factor present in the vaginal environment of CVVC-S mouse strains, but absent in the CVVC-R environment, is responsible for the susceptible phenotype.

Studies that focused on identification of the elusive inhibitory factor in the CVVC-S mice confirmed that the factor was a protein or contained protein-associated moieties, based on the sensitivity of the inhibitory VCM to heat and protease treatment. Considering the requirement for the interaction of Mac-1 on PMNs and Pra1p on C. albicans for effective killing (73, 74), it was hypothesized that the inhibitory factor present in VCM may be a competitive ligand for Mac-1. To investigate this possibility, VCM was pretreated with Mac-1^+/+ and Mac-1^−/− PMNs to deplete the putative competitive Mac-1 ligands. Indeed, VCM pretreated with Mac-1^+/+ PMNs eliminated the inhibitory effect, while pretreatment with Mac-1^−/− PMNs retained the inhibitory activity. A similar effect was observed with soluble recombinant Mac-1, which rescued the killing capacity of PMNs in VCM. These results provided strong evidence that the putative inhibitor present in VCM was a competitive Mac-1 ligand that blocks the interaction of PMNs with C. albicans, resulting in reduced antifungal activity.

The proteoglycan heparan sulfate (HS) was identified as a strong candidate for the inhibitory factor based on the ability to interact with Mac-1 (79, 80) and evidence for constitutive expression by mouse vaginal epithelium and human vaginal epithelial cell cultures (81–83). Indeed, PMN activity in VCM collected from several CVVC-S strains of mice was rescued by pretreatment with heparinase III (an enzyme that specifically cleaves HS, also termed heparanase). Conversely, PMN killing was reduced by the addition of HS to standard medium in a dose-dependent manner. Hence, there was strong evidence that HS was the putative competitive ligand for Mac-1 responsible for inhibiting PMN antifungal activity in the vaginal environment of CVVC-S mice. This environment-dependent abrogation of killing activity by activated PMNs is indicative of an anergic condition (defined as “absence of the normal immune response to an immunogen”) and thus is being referred to as “neutrophil anergy.” Figure 2 shows a diagram of how resistance and susceptibility to VVC via the presence/action of HS are envisioned. Table 1 highlights the key findings in VVC/RVVC pathogenesis investigations that have led to the immunopathogenic mechanism of HS-mediated PMN anergy.

FIG 2.

Schematic model representing normal PMN function in CVVC-resistant mice and the mechanism for “neutrophil anergy” in CVVC-susceptible mice. (A) Effective fungal recognition and clearance by PMNs in a CVVC-resistant mouse strain. Production of heparan sulfate (HS) by the vaginal epithelium is minimal in estrogen-hyporesponsive CD-1 mice. Following intravaginal inoculation with C. albicans, PMNs are recruited to the vaginal lumen and further activated via Mac-1 receptor binding to C. albicans surface protein Pra1. The interaction of Mac-1 with Pra1 is sufficient to mediate fungal killing by PMNs, leading to rapid fungal clearance and timely resolution of acute inflammation. (B) Lack of fungal clearance by PMNs in a CVVC-susceptible mouse strain. Production of HS by the vaginal epithelium is elevated in mice with normal estrogen responsiveness. Increased levels of HS in the vaginal lumen block recognition of C. albicans Pra1 by PMNs by acting as a competitive ligand for Mac-1. The failure of activated PMNs to adequately recognize C. albicans results in the inhibition of PMN killing (PMN anergy) despite a sustained acute inflammatory response, placing the vaginal mucosa in a chronic immunopathological state.

TABLE 1.

Key findings, in chronological order, from studies reported between 1990 and 2017 investigating the pathogenesis of VVC/RVVC

Category	Finding(s)
Lack of adaptive immunity in protection	Local rather than systemic immunity is critical for host defense against infection
	Immunoregulatory mechanisms are responsible for the lack of adaptive immunity in the vagina
Innate resistance by epithelial cells	Epithelial cell anti-Candida activity via annexin A1 is predominant innate protective response
Immunopathology by PMNs	Paradigm change: innate responses by neutrophils rather than adaptive immune deficiency contribute to immunopathogenesis
	S100 alarmins and IL-1β produced by epithelial cells in response to Candida via TLR4 and SIGNR1 and involving the NLRP3 inflammasome are key to initiation of the immunopathogenic response associated with symptomatic infection
	Morphogenesis by C. albicans is required for the immunopathological response
	Candidalysin is predominant molecule invoking epithelial cell damage leading to the epithelial cell activation and subsequent immunopathogenic response
	Mechanism of immunopathogenic response uncovered: heparan sulfate in the vaginal environment serves as a competitive inhibitor of neutrophil-Candida interaction required for killing, resulting in a state of neutrophil anergy

HS is a member of the glycosaminoglycan family of polysaccharides ubiquitously expressed on a wide range of cell surfaces and throughout the extracellular matrix in all mammalian tissues, including the vaginal epithelium (81–84). HS proteoglycans, a biologically relevant form of HS, occur as a macromolecule in which one or more HS chains are attached to a cell surface or extracellular matrix protein with a remarkable structural diversity due to variations in chain length (85). HS proteoglycans are found in a variety of physiological activities and are known to interact with both host and certain microbial components during inflammation and infection (85). In epithelial tissues, HS proteoglycans are expressed on the entire surface of stratified squamous epithelial cells (an estrogen-responsive cell type in the vagina) at much higher levels than on columnar epithelial cells (82). While it has been shown that HS is estrogen inducible in mice (81), it may be that the basal levels of circulating estrogen present in CVVC-S mice, due to their responsiveness to estrogen, are sufficient to promote constitutive expression of HS. This is supported by similar inhibitory properties of VCM collected from naive and oophorectomized mice. In CVVC-R CD-1 mice, on the other hand, the lack of the inhibitory environment may be linked to reduced estrogen responsiveness such that circulating levels of estrogen are unable to increase expression of HS. Interestingly, a recent microarray-based approach using human vaginal epithelial cells infected with C. albicans revealed considerable upregulation of two-heparin related genes, those for heparin binding epidermal growth factor (EGF)-like growth factor (HBEGF), which is involved in tissue healing, and for heparan sulfate glucosamine 3-O-sulfotransferase 1 (HS3ST1), which catalyzes the biosynthesis of tissue heparan sulfate and anticoagulant heparin (54). These unbiased global transcriptional approaches further support a role for heparan sulfate and potential HS-mediated downstream signaling (including tissue repair and wound healing) at the vaginal mucosa during VVC. Future studies will address what estrogen-associated factors (e.g., physiological estrogen concentrations, intensity of responsiveness, or estrogen receptor expression levels) are most critical to regulating HS levels in the vaginal environment, which ultimately determines the CVVC-S and CVVC-R phenotypes. Of note, a recent report described a similar PMN-inhibitory factor in lavage fluid of CVVC-R CD-1 mice that was restricted to during or following infection with C. albicans (59). Thus, there may be inducible PMN-inhibitory factors during infection that function during subsequent infections, though likely via mechanisms independent of HS.

FUTURE DIRECTIONS

Building upon the discovery of the newly defined role of HS in the immunopathogenic neutrophil anergic response in experimental VVC, there are several directions that can be initiated, both in the mouse model and clinically. First, HS may represent a novel biomarker to focus on as a therapeutic target. Accordingly, future studies in the animal model can include HS neutralization by intravaginal administration of soluble recombinant Mac-1 or heparanase into CVVC-S mice during infection to alter the anergic environment. Additionally, intravaginal administration of HS in CVVC-R CD-1 mice to inhibit PMN killing may also shed light on HS as a target biomarker. However, we recognize that caution should be taken with exploiting HS as a therapeutic target, based on its multifunctional properties; neutralization or cleavage of HS may interfere with other protective or nonprotective processes in the vaginal environment. In the clinical setting, it will be important to begin testing these hypotheses. Based on the similarity between mice and humans in the immunopathological response (36, 37), we predict that vaginal secretions from women susceptible to VVC/RVVC will have considerable levels of HS and similarly promote neutrophil anergy by the same competitive inhibition of Mac-1. In contrast, we predict that women with no history of VVC (resistant) either have low HS, express different isoforms of HS, or fail to trigger the immunopathogenic PMN migration in the presence of similar HS levels. Together with the identification of a major mechanism underlying the immunopathogenesis of experimental VVC, these clinical hypotheses and related follow-up studies in the mouse model will be useful in developing therapeutic strategies for clinical application targeting the immunopathogenic response in women susceptible to VVC/RVVC.