The aim of this review is to examine the role and mechanisms whereby Candida albicans, the largely prevalent etiological agent of vulvovaginal candidiasis (VVC), causes the disease, and how the host responds to the transformation of a vaginal commensal into an aggressive pathogen. With the aim of generating a safe and efficacious anti-Candida vaccine to prevent or cure the most severe and chronic condition known as recurrent vulvovaginal candidiasis (RVVC). For a more general insight into the epidemiology, diagnosis and current treatments of the infection, the reader can consult a number of authoritative reviews.[1-5]
The disease: VVC and RVVC
VVC is an acute inflammatory disease and a frequent reason for gynaecological consultation as it can affect up to 75% of women of child-bearing age. Clinical signs and symptoms include intense pruritus, vaginal discharge, an erythematous vulva and dyspareunia. A number of predisposing factors such as oral contraceptive usage, pregnancy, uncontrolled diabetes mellitus and long-term broad spectrum antibiotic treatment have been identified as predisposing factors. Many others are suspected, including changes in the composition and function of the vaginal microbiota. With the exception of rare cases of resistance to antifungals, usually caused by azole-resistant non-albicans species of Candida,[2, 5] VVC is usually readily treated, provided that risk factors are removed or controlled, and VVC remains for many women an infrequent experience. RVVC is a much more serious clinical condition due to the recurrences of symptoms (four or more episodes per year) and for its refractoriness to successful treatment. Long-term maintenance therapy with fluconazole may help lengthen the asymptomatic periods between recurrences but does not provide a long-lasting cure. Recent epidemiological investigations have suggested that the prevalence of RVVC may be higher than previously estimated and can be as high as 7–8% of women who experience a first episode. This would translate into an estimated global annual incidence ranging of 1–2% of all women. The discomfort associated with RVVC is intense. It markedly diminishes the quality of life in young women, with a strong negative impact on both work and social life. Moreover, affected women frequently are attracted by advertisements to purchase over-the-counter formulations, prebiotics and probiotics which, besides being ineffective, can aggravate their symptoms.
The factors that determine which women will undergo the transition from sporadic VVC to RVVC are still undefined. A small fraction of recurrences may be due to the persistence of those predisposing factors that underlie VVC itself, but in the majority of cases RVVC is idiopathic, as it occurs in women without any known risk factors. This is suggestive of a genetic predisposition, similar to what has been observed for other Candida infections such as chronic mucocutaneous candidiasis. The high incidence of RVVC suggests that a genetic predisposition would probably involve a number of different genes in individual patients, and an interaction with environmental triggers. Several polymorphisms predisposing to RVVC have been reported, including single nucleotide polymorphisms (SNP) in genes coding for mannose-binding lectin, interleukin (IL)-4, Dectin-1 receptor, CARD 9, a subunit of the NLPR3 inflammasome complex and, more recently, gene variants of IL-22 and the enzyme indoleamine 2,3-dioxygenase in regulatory T-cells.[11-15] Overall, current evidence supports the assumption that gene polymorphisms are to be included as a predisposing factor which puts women at risk for VVC and RVVC.
The pathogen: Dr Yeast and Mr Hypha
Candida albicans is a eukaryotic microorganism with an extraordinary capacity to adapt to different environmental and host niches. These unique properties allow for the dual lifestyle of C. albicans as both a commensal and an opportunistic pathogen for humans and other mammals. This duality has a morphological correspondence in the capacity of C. albicans to undergo a morphogenetic change from a round-ovoid typical yeast cell (Y) to a hyphal mycelial growing organism (H) (dimorphic transition, Figure 1). This transition is of utmost relevance for C. albicans pathogenicity. There is sufficient evidence that the Y form is predominantly associated with commensalism, while the H form is associated with pathogenicity. In the Y form, C. albicans can be found in the intestine and vagina in >50% of healthy asymptomatic subjects, whereas the H form is invariably found in pathologic specimens obtained from invaded tissues, including those of women with VVC or RVVC (Figure 1). Commensal Y cells are tolerated by the host and are maintained at low numbers on the vaginal epithelial surface by a variety of mechanisms that inhibit transition to the H form. It remains to be determined whether the presence of commensal C. albicans conveys a benefit to the host in terms of balanced microbiota composition and maintenance of local immune homeostasis. However, when tolerance mechanisms become defective, the Y form changes into the H form and expresses its virulence traits. These hyphae form a robust biofilm layer that strongly adheres to, and then invades, the outermost layer of the vaginal epithelium (Figure 1).[16-18] Detachment of the H form from the epithelium, together with recruited inflammatory cells, debris from lysed cells and vaginal fluid make up the vaginal discharge which is one of the clinical signs and symptoms that are typical of VVC. Strains of C. albicans that lack the capacity to undergo the dimorphic transition do not produce a biofilm and are typically non-pathogenic.[19, 20] A possible exception is in cases where the VVC or RVVC is due to an allergic response to a C. albicans component present in the Y form. Similar to the evil transformation narrated in the novel Dr Jekyll and Mr Hyde, C. albicans changes its aspect and behaviour from a possibly beneficial member of the vaginal microbial community (Y form) to an aggressive pathogen (H form).
Animal models versus human infection
As with other microbial infections, the pathogenicity of and immune responses to C. albicans have been studied in in vitro and ex vivo systems, including the vaginal reconstituted human epithelium, as well as in animal models. Both rat and mouse models have been rather extensively utilised to investigate Candida vaginitis, recently with the application of minimally invasive bioluminescence-based imaging techniques (Figure S1). Both animal models have generated useful information but the differences from the human vaginal Candida infection, particularly the RVVC, remain substantial, as outlined in Table 1. Of relevance, both rodent models and human infection are stringently estrogen-dependent. Estrogens exert a multifunctional permissive role for vaginal candidiasis through a number of both host- and Candida-directed activities (Table 2).[22-25] However, it is not known which of the activities is more prevalent or is a real determinant in pathogenesis.
|Condition||Candida vaginal infection|
|Protective immunisation after resolution of primary infection||Noa||Yes||Yes|
|Role of vaginal T-cells||Unknown||Protective||No|
|Role of vaginal antibodies||Uncertainb||Yes||No|
|Role of neutrophils in disease||Yesc||No||Yes|
|Role of Sap and adhesins in disease||Postulated||Yes||Yesd|
|Role of hyphal transition in disease||Yes||Yes||Yes|
|Host-directed activities||Pathogen-directed activities|
|1. Enhances glycogen level of EC and favours fungus growth through its metabolism to glucose. Induces partial keratinisation of epithelial cell surface, favouring fungus adherence and biofilm[22, 23]||1. Interacts with a cytosolic receptor of C. albicans|
|2. Regulates receptors expression on vaginal epithelial cells relevant for anti-Candida defense[46, 51]||2. Influence Y to H transition of the fungus|
|3. Modulates expression of nuclear factor (NF)-κB-related inflammation||3. Increases the expression of drug efflux-related Candida homologue genes of cancer cells|
|4. Impairs Th17 immune response against C. albicans[53, 81]||4. Up-regulates HSP90 fungus chaperonin, favouring its growth and stress resistance|
|5. Regulates the expression of Fc. neonatal receptor for IgG transport. For other details see text and quoted references|
The most important difference between the human clinical infection and rodent models stems from the fact that C. albicans is a commensal of human vagina but not rodent vagina and hence there is a pre-existing anti-Candida immunity in women that is absent in rats and mice. The corollary is that, in women, either this immunity is ineffective or, more likely, factors modulating its activity come into play for symptomatic candidiasis to occur. Another strong difference from animal models is that in women the sources of infection are endogenous (from the intestine, the cervico-vaginal areas or skin), whereas in animals the infection requires the introduction of rather large (from >1 million cells) exogenous intravaginal inocula. In estrogen-treated mice the infection is associated with induction of a robust inflammatory immune response, essentially unrelated to Candida growth. In women, symptoms of vaginal candidiasis, particularly RVVC, have been associated with both a high and a low Candida burden, as well as with the presence of both high and low numbers of inflammatory cells.[1-3, 24] Fidel and collaborators investigated human volunteers intravaginally challenged with live fungal cells and found a correlation between fungal intravaginal burden and inflammation, as shown by leucocyte infiltration but, again, this protocol does not reflect the natural source of infection.
In both animal models, recovery from a primary infection provides protection against a subsequent infection.[28, 29] Although this typically does not occur in women, the animal models may be useful to study anti-candidal immunity following vaccination (see below).
How H cells convey fungus virulence
In its pathogenic H form, C. albicans expresses virulence factors that have been the subject of multiple genetic and biochemical investigations. As reviewed elsewhere, two main features characterise the elaboration of virulence factors: the redundant nature of these factors and a focus on avoidance of and resistance to host immune responses. To fully appreciate how successful these factors are, we should remember that RVVC mostly occurs in otherwise healthy women, with no apparent immunological deficits or underlying severe illnesses that predispose to other mucosal (oral) or invasive Candida infections. In the past, there has been a major emphasis on hyphae as ‘mechanical’ factors that prevent Candida elimination. It was postulated that polymorphonuclear leucocytes (PMN, neutrophils) and macrophages were unable to engulf and kill the long fungal threads that were intertwined into robust biofilms. However, recent research has shown that most of the H cell virulence is due to factors that facilitate the escape of the fungus from local host response. A list of these factors is provided in Table 3. The redundant nature of virulence traits enables the fungus to select alternative pathogenic traits when one or more of them are neutralised. Among the various virulence traits, secretory apartyl proteinases (SAP) have received a great deal of attention in the study of vaginal candidiasis.
|Secreted aspartyl proteinases (Sap family)||Inflammation and enzymatic degradation of tissue barrier proteins (e.g. E-cadherin), innate (e.g. complement) and adaptive (e.g. antibodies) immunity|
|Factor H and plasminogen-binding mannoprotein complexes of the cell wall||Interfere with complement function as opsonin|
|Hypha-associated proteins||Inhibit phagocytosis and interfere with candidacidal activity of phagocytes|
|Other factors||Inhibit cytokine production and/or function|
Secretory aspartyl proteinases and candidal vaginitis
There is substantial evidence that Sap has multiple roles in both animal models of Candida vaginal infection and human VVC and RVVC. Sap is a family of at least ten genes with differential expression in different host niches and coding for proteins (Sap1–Sap10) with potential multiple and distinct roles in disease.[32, 33] Women with VVC and RVVC have a significantly higher amount of one or more enzymatically active Sap in their vaginal fluid, and the fungal strains isolated from these women produce more Sap in vitro, as compared with those who carry the organism asymptomatically. These data are fully in keeping with reports of investigations in experimental models using Sap gene-deleted strains of C. albicans.[35-37] A differential expression of the C. albicans Sap genes in women with active vaginitis and asymptomatic carriers has been reported. Sap gene expression appears to depend on many factors, including the environmental pH, and stage and type of C. albicans growth in vivo, as some Sap genes are mostly expressed when Candida grows in the Y form and others mostly when the organism is in the H form.[32, 37, 38] Although the data point collectively to a role of Sap proteins in VVC and RVVC, there is no clear demonstration of the mechanisms involved. A widely held, and logical, explanation rests upon the capacity of Sap enzymatically to degrade stability factors of epithelial cells (e.g. E-cadherin) as well as innate and adaptive immunity factors (e.g. complement, histatin, antibodies), allowing C. albicans to escape the host immune system.[32, 33, 39] Some past[40, 41] and more recent data suggest that Sap may induce inflammation and thus contribute to fungal pathogenicity by fostering, rather than evading, the host immune systems (see below).
Interestingly, in women with RVVC, no or very low levels of anti-Sap antibodies are found in the vaginal fluid and the circulation, suggesting that Sap degrade antibodies and/or that these proteins are of low intrinsic immunogenicity. However, it is possible to induce anti-Sap antibodies (Abs) by local or parenteral administration of different vaccine formulations that incorporate a dominant member of the Sap family, Sap2. These antibodies have been shown to be protective in animal models of candidal vaginitis, constituting a rationale for the proposal of the recombinant Sap2 protein fragment as an anticandidal vaccine.[43-45]
Immune responses and inflammation
Similar to other body sites exposed to potential pathogens, the cervico-vaginal environment must be prepared to activate innate immunity and mount adaptive immune responses to control, if not eliminate, the pathogens. Consequently, the vagina is well equipped with many cellular and humoral factors, including dendritic cells (DC), as well as T-helper, regulatory and cytotoxic lymphocytes, B-lymphocytes and natural killer cells producing protective cytokines and chemokines that can help recruit additional defensive factors from distant body sites.[45-50] The activity of this rather solid immunological armamentarium is centred upon the vaginal epithelial cells (VEC), a lead actor in vaginal antimicrobial defense.[51, 52] Vaginal epithelial cells not only constitute a mechanical and trapping barrier with their mucin and keratin-like surface material, but are also able to ‘sense’ the danger constituted by the pathogen and to respond by cell activation and secretion of immune mediators driving inflammation and immune responses. Intracellular multiprotein complexes, called the ‘inflammasomes’, translate the danger signals associated to the pathogen or its products into immune activation. Recruitment of polymorphonuclear cells to the vagina, cytokine (particularly IL-1β and IL-18) production and activation of the lymphocyte subsets T-helper 1 and (with some controversy) T-helper 17 have been attributed a role in anti-Candida protection (Figure 2).[53-57] Notably, the same mechanisms, if inappropriately or excessively stimulated, cause damage through stimulation of an overwhelming inflammation, hence impairing the anti-Candida defense.
The vagina is also a special tolerogenic site as it must accept both constant (the vaginal microbiota) and occasional (semen and fetal) non-self-material. Hence, a delicate balance is maintained that must take into account two equally essential needs for health of the vagina—immune defense and immune tolerance. Paraphrasing Sansonetti, the vagina is permanently obligated to discriminate between commensals in need of tolerance and pathogens that need to be quickly identified and eliminated. Infection may occur both because of the loss of immunosurveillance and the loss of tolerance. For C. albicans it is therefore intuitive that VEC must have mechanisms to distinguish Dr. Yeast, whose presence (at low density) is tolerated if not welcomed, from Mr. Hypha, who can damage them. Part of these mechanisms have recently been highlighted by Naglik and collaborators, who showed that the two forms of growth are discriminated by activation of distinct MAPK kinase pathways. In reality, a low, ‘positive’ degree of inflammation is present in the healthy vagina and is consistent with the tolerance of low number of C. albicans cells, but a high ‘negative’ degree of inflammation follows fungus transformation into the pathogenic H form. An inflammatory ‘threshold’ would therefore separate safe, if not beneficial, co-existence with Candida in the vagina, from the presence of pathogenic fungus and clinical disease. On this basis, it is a reasonable hypothesis that in women with RVVC, this inflammatory ‘threshold’ is lower and more easily reached than in other women.
The vaginal microbiota
The vaginal microbiota is characterised by relatively few microbial communities compared with the intestinal microbiota, among which Lactobacillus species (the classical Doderlein's bacilli) such as Lactobacillus iners, Lactobacillus crispatus, Lactobacillus gasseri and Lactobacillus jensenii are dominant. The stability of this microbial ecosystem is greatly affected in some women by the physiological variations of the menstrual cycle and hormones, and changes dramatically after menopause. It has long been suggested that the vaginal microbiota contributes to protection against microbial pathogens. Microbiota disruption by antibiotic usage is often followed by VVC. There is, however, insufficient evidence and contradictory reports for establishing the magnitude of this protective contribution compared with other protective factors.[61, 63] In addition, mechanisms of putative protection are unclear. Lactobacilli produce lactic acid, rendering the vaginal microenvironment sufficiently acidic (average pH around 3.5) to inhibit the C. albicans Y to H transition. However, Candida has refined mechanisms to oppose drastic pH variations and has been shown to be capable of neutralising excess acid by ammonia derived from amino acid metabolism. Interestingly, this mechanism autoinduces hyphal development. More likely, lactic acid influences the anti-pathogen response of the VEC.
Zhou et al. reported the absence of a clear difference between the vaginal microbiota of women with a history of RVVC and healthy controls, and suggested that the vaginal microbiota does not protect against infection. However, studies comparing microbiota composition do not generally incorporate functional aspects of the microbiota that can be very relevant for the differential production of lactic acid and other anti-Candida metabolites impacting mucosal reactivity. More recent studies with refined genomic approaches suggest a higher diversity of vaginal microbiome than previously suspected, yet no consistent vaginal microbiome differences between healthy and Candida-vaginitis patients have been reported. An interesting recent study shows that growth of C. albicans on lactate renders these cells less of an immunostimulant than those grown on glucose. Lactate-grown fungal cells induce production of the inhibitory cytokine IL-10 and decrease the production of the pro-inflammatory cytokine IL-17, and are more virulent than glucose-grown cells in an experimental vaginal infection model. Overall, there still is no clear evidence to suggest that women with VVC or RVVC have a vaginal microbiota definitely different from that of normal women. Studies at genomic or sub-genomic level are requested to identify fine, possibly strain or type-related differences in the microbiota that could affect the vaginal pathology.
Vaccines against RVVC
The high incidence of RVVC and the difficulties controlling its occurrence with conventional chemotherapy constitute a strong medical need and driving force for the development of immunological treatments adding to, if not replacing, the current antifungal armamentarium. The use of exogenous cytokines, antibodies and immunomodulators is of potential interest[71, 72] but the cornerstone for immunotherapy would be the development of a safe and efficacious anti-Candida vaccine. Ideally, this vaccine should be capable of inducing immune responses against fungal virulence traits, without alteration of the delicate tolerance/inflammation balance of the vaginal environment discussed above. At variance with immunocompromised subjects at risk of invasive Candida infection, women with RVVC are healthy and fully immune-competent to mount vaccine-induced immune responses capable of controlling if not eliminating the fungus. After decades of neglect, activity in this area is now increasing and there are several research teams addressing anti-Candida vaccination, with RVVC as the major if not the sole indication (reviewed in Refs [30, 73, 74]). Two such vaccines have passed phase 1 clinical trials for safety and immunogenicity, and one of them has entered a phase 2 clinical trial. Not surprisingly, the two vaccines are based on a recombinant product or fragments of proteins largely studied in experimental models, i.e. the Als3 adhesin, formulated with alum as adjuvant, and a dominant member of the aspartyl proteinase enzyme family, the Sap2, in a virosomal formulation.[39, 75] Both vaccines show solid evidence of protection in rat and mouse models of vaginal infection by C. albicans although with a likely different mechanism of immunological protection.[30, 74] Other anti-Candida vaccines, from attenuated strains of C. albicans to a number of glycoconjugate of cell wall polysaccharides, have also been shown to be immunogenic and protective in experimental animal models, although these, to my knowledge, have not yet been entered into clinical trials in humans.
Table 4 summarises the state of the art of major anti-Candida vaccines currently under study.
|Vaccine and adjuvant/carrier||Antigen biological function||Present clinical state||Putative protection mechanism|
|Als-3 alum||Member of major adhesin family of C. albicans. Involved in adherence to epithelia and biofilm formation. Hypha-associated||Phase II||Activation of Th1/Th17 cells Possible role of pre-existent antibodies boosted at high titer in vaccine recipients|
|Virosomal Sap2||Degradation/deviation of host immune factors. Inflammation Consistently involved in virulence in mucosal pre-clinical models||Closed at Phase 1||Virulence-neutralizing, inflammation-neutralizing antibodies|
|β-glucan-CRM–conjugate/MF59||A main pathogen-associated molecular pattern, involved in fungal cell stability. biofilm formation and activatory and inhibitory immune mechanisms||Preclinical||Protection due to universal anti-beta-glucan-antibodies, growth- and adherence inhibitory|
|Beta-mannan- and Βeta-mannoside conjugates||Adhesin, involved in host immune-deviation||Preclinical||Opsonizing antibodies|
|HyR-1 (no adjuvant defined, probably alum)||Hypha-associated virulence protein involved in protecting the fungus from PMN activity.||Preclinical||Protective antibodies|
Of interest, almost all of the experimental vaccines owe their protection to the generation of specific neutralising antibodies, in apparent contrast to the presumed absence of a role for antibodies in the natural anti-Candida defense. As a matter of fact, this is not a contradiction. In the normal situation of the commensal Y form, there appears to be no need for a protective antibody in the presence of an active immune surveillance system dominated by innate immunity. However, this innate defensive system evidently fails to protect from RVVC. Inhibitory antibodies, some of which may directly block growth of the H form and/or dampen inflammation caused by pro-inflammatory Candida components,[37, 39, 54, 78] may be generated by both mucosal and parenteral administration of a vaccine. Serum IgG translocates to the vagina from the circulation by diffusion as well as by an active transport system mediated by a neonatal Fc receptor, the same transport system ensuring transplacental passage of maternal IgG antibodies to the fetus. The neonatal Fc receptor is estrogen-sensitive, is expressed by the vaginal epithelium, and has been shown to confer protection against herpes simplex type 2 vaginal infection. This system constitutes a solid rationale for parenteral (e.g. intramuscular) anti-Candida vaccination, thus avoiding administration of an antigen in a tolerant if not immune-inhibitory environment, rich in potentially interfering bacteria and enzymes, such as the vagina. Both Sap2 and Als3 vaccines are intended for parenteral administration and have already been shown to generate neutralising, presumably protective, vaginal antibodies in women following i.m. administration in the phase 1 clinical trial (other unpublished data).
One of the problems with the two most advanced vaccines described above is that both constituent antigens belong to families with several members, with limited similarity and immunological cross-reactivity, so that in principle the protective immune response raised by the vaccination could not be broad enough to protect against the expression of similar but non-cross-reactive members. As noted above, the virulence program of C. albicans is redundant. One mechanism to address this potential risk of vaccine failure is by making a combined vaccine, composed of more than one antigen derived from different virulence traits, likely broadening the spectrum of protective immune responses from blockade of inflammation to inhibition of fungus growth under hyphal form and adherence/damage to the VEC. Potential immunologically synergistic antigen combinations have been extensively dealt with elsewhere. Alternatively, inactivated C. albicans cells or virulence-attenuated Candida strains could be used as vaccines. However, the use of these whole cell vaccines in humans which are normally colonised by C. albicans would raise several immunological and regulatory issues.