Candida biofilm formation and recurrent vulvovaginal candidiasis

Introduction

Like many other opportunistic pathogens, Candida species inhabit the mucosal surfaces of the human gastrointestinal tract, genitourinary tract and skin as harmless commensals that become pathogenic in hosts with elevated sensitivity to fungal colonisation.¹ Candida has gained considerable importance as an infective agent in women’s health, causing vulvovaginal candidiasis (VVC) in females of reproductive age.² According to the US Centers for Disease Control and Prevention, VVC can be broadly categorised as uncomplicated VVC that is sporadic and mild-to-moderate infections caused by Candida albicans in immunocompetent women and complicated VVC that includes recurrent VVC (RVVC), severe VVC, VVC caused by non-albicans Candida, and VVC in immunocompromised patients (see https://www.cdc.gov/std/treatment-guidelines/candidiasis.htm). The terms sporadic VVC and RVVC are widely used in clinical diagnosis of VVC cases; the former refers to infections of less than three episodes per year, and the latter includes no less than three relapses or re-infections in a 12-month period, with entirely asymptomatic intervals between these episodes.³ It has been estimated that up to 75% of reproductive women experience VVC and ~5–8% suffer from RVVC.⁴ C. albicans remains the leading causative agent for both sporadic VVC and RVVC.⁵ Other non-albicans Candida species account for 10–20% of RVVC, with a reported upward trend and an emerging importance.⁵

Several studies evaluated antifungal susceptibility of Candida yeast cells isolated from VVC patients and reported very low percentage of azole resistance among these clinical isolates.^6,7 Despite the fact that most Candida clinical isolates from VVC patients, including those with RVVC remain sensitive to azoles, long-term maintenance therapy using fluconazole often fails to eradicate Candida cells or completely cure the infection, suggesting that microbial strategies other than intrinsic resistance, such as biofilm formation may contribute to the suboptimal treatment outcome and infection recurrence.^8,9 This review will discuss insights into the roles of biofilm formation of Candida, including but not limited to the albicans species, in the pathogenesis and persistence of RVVC.

The current debate: with or without Candida biofilms in the human vagina?

Biofilm formation has been proposed as one of the most important virulence factors of Candida spp. contributing to the occurrence of RVVC and its tolerance to antifungal treatments.^9,10 Supporting this theory are the clinical pattern of RVVC that coincides with the model of relapsing infections proposed by Lewis, in which biofilm formation is the main culprit,¹¹ and experimental findings from in vivo studies using mice as the host.^9,10 Murine models of this disease, however, differ from human VVC and RVVC in several physiological aspects of the vaginal environment, including the absence of C. albicans in the vaginal microbiota, neutral vaginal pH, and dependence on exogenous estrogen to initiate fungal colonisation.^10,12 Such differences have led researchers to question whether the importance of Candida biofilms in the pathogenesis of VVC and RVVC established in rodents could be transferred to humans.¹² Recently, Swidsinski et al. examined vaginal tissues from patients with sporadic VVC or RVVC using fluorescence in situ hybridisation (FISH) and Candida-specific 18S rRNA probes. Based on cytogenetic findings, these authors disputed the presence of Candida biofilms in the human vagina, and concluded that VVC and RVVC was a result of polymicrobial invasion of vaginal tissues.¹³ In our preliminary clinical study, we observed biofilm-like Candida growths on the vaginal epithelium of RVVC patients, using scanning electron microscopy to visualise adherent Candida cells (Fig. 1a), and confocal laser scanning microscopy in combination with Concanavalin A (ConA) staining for the presence of Candida biofilm extracellular polymeric substances (EPS) (Fig. 1b). Such high-resolution visual evidence supports the presence of Candida biofilms in the vagina of RVVC patients and its possible association with the disease.

Fig. 1. 

(a) Scanning electron microscopy (SEM) of human vaginal tissue showing Candida albicans monolayer biofilms grown on the vaginal epithelium of a recurrent vulvovaginal candidiasis (RVVC) patient. (b) Confocal laser scanning microscopy (CLSM) of a vaginal biopsy section stained with ConA supported the presence of Candida biofilm extracellular polymeric substance (EPS) matrix (green).

Biotic and abiotic Candida biofilms associated with RVVC

Two types of clinical biofilms may be linked to RVVC depending on their supporting substratum, including abiotic biofilms that grow on the surface of intrauterine devices (IUDs), and tissue-based biotic biofilms that are formed directly on the vaginal epithelium.^10,14 IUD-based biofilms resemble in vitro biofilms cultured in 96-well microplate or on silicone pads and present either as a complex three-dimensional structure comprising of multilayer aggregates of yeast cells encased within a matrix of EPS,¹⁴ or yeast-dominated adherent monolayers.¹⁵ Many molecular regulators of Candida biofilm formation on abiotic surfaces have been identified, comprising mainly transcription factors such as Bcr1, Tec1, Efg1, Ndt80, Rob1, Med31 and Ccr4.^16,17 Epithelium-based biotic Candida biofilms differ significantly from abiotic biofilms in morphology, often presenting as monolayers or microcolony biofilms in the vagina.^9,10 Epithelium-based biofilms may play a more important role in RVVC than IUD-based abiotic biofilms, as many patients still suffer from RVVC without an IUD or after removal of the infected IUD.¹⁸

The possible important role of biofilm formation in the pathogenesis of RVVC

Biofilm formation of Candida has been long proposed as an important microbial strategy that contributes to the pathogenesis of VCC and RVVC. Most Candida clinical isolates from VVC patients were found to be able to form in vitro biofilms on abiotic surfaces,¹⁹ though individual clinical isolates may vary greatly in their abilities to produce biofilms in 96-well microplates.²⁰ In the general model of Candida mucosal infections, C. albicans express multiple surface proteins (adhesins) upon immediate contact with the host epithelium, mediating microbial adherence to host cells and formation of biotic biofilms. These adhesins include the ALS family, Ece1, Hwp1, Phr1 and Phr2.^21,22 The importance of Als3, Hwp1 and Phr1 in the initiation of VVC remains to be questioned. The unique acidic vaginal environment (pH < 4.2) of VVC patients represses the expression of phr1²³ and does not support filamentous growth of Candida that is essential for the expression of hwp1 and als3.²⁴ The importance of Candida biofilm formation in the pathogenesis of VVC was also supported by our recent work that compared the histopathological changes of vaginal mucosa infected with wild type Candida and their bcr1 and med31 mutants.⁹ These mutant strains were deficient in early fungal adherence in vitro and in vivo and caused less severe vaginal infections in mice, characterised by lower fungal colonisation of the vaginal mucosa and milder host histopathological responses.⁹

Antifungal resistance of epithelium-associated biofilms contributes to the relapse of RVVC

In addition to its involvement in the pathogenesis of RVVC, more recently, biofilm formation of Candida has been implicated in the persistence or relapses of RVVC. Numerous in vitro studies have showed increased resistance of Candida biofilms formed in 96-well microplates to antifungal agents, in particular azoles.^9,25 The elevated antifungal resistance of in vitro Candida biofilms is at least partially due to the retarded penetration of azoles into the biofilm EPS matrix.²⁶ Biofilm-associated Candida cells contain more β-1,3-glucans and other carbohydrates than their planktonic counterpart and bind to 4–5-fold more fluconazole. Another possible mechanism of biofilms resistance is the up-regulation of efflux pumps in biofilm-embedded Candida cells, including the ABC (ATP binding cassette) transporters encoded by Cdr genes and MFS (major facilitator superfamily) transporters encoded by Mdr genes.²⁷ Overexpression of these efflux pumps induce Candida resistance to various antifungal drugs such as azoles, terbinafine and amorolfine.^28–30 The high antifungal resistance and underlying mechanisms found in abiotic Candida biofilms may not be applicable to biotic biofilms formed on the vaginal epithelium, as short-term application of antifungal drugs such as azoles remains highly effective in managing acute episodes of RVVC, suggesting transiently effective killing of biofilm cells at the infection site.³¹ Clinical studies evaluating the efficacy of prolonged antifungal maintenance therapies against RVVC also found a persister-based biphasic killing kinetics, supported by the relapse of infection once the treatment was discontinued.⁸ High tolerance of Candida biofilms to fluconazole in RVVC patients may explain why suppressive maintenance therapies, even used for a prolonged period, are effective in controlling symptomatic RVVC but are rarely curative, despite the general in vitro sensitivity of causative Candida to fluconazole.⁸ Our recent animal studies showed the presence of persister cells in epithelium-based biofilms that might result in a new episode after resuscitation.^9,25 Persister cells are a small population equipped with delicate metabolic control and a coordinated stress response, showing subdued major energy-generating pathways, down-regulated glycolysis and protein synthesis, and tolerance to numerous antifungal drugs at high concentrations.³²

Limitations of the current VVC animal model

The current murine model of VVC has been widely used to study in vivo interactions between pathogenic Candida strains and the vagina.¹⁰ The disease state of infected animals is characterised by two histopathological changes that are often associated with acute mucosal infections, sub-epithelial micro-abscesses and neutrophil infiltration.^9,10 It is deduced that the current mouse VVC model may adequately represent the acute phase of VVC and RVVC, but not the chronic nature of RVVC. A new animal model is urgently needed for the comprehensive understanding of the disease pathogenesis and development of more effective therapies. Using vaginal isolates rather than laboratory reference strain with a bloodstream origin such C. albicans DAY185 or C. albicans SC5314, extending the use of exogenous estrogen and the incubation period from 48–72 h to 1–2 weeks may re-direct the infection from neutrophil-dominated acute infections to persistent and asymptomatic chronic infections.

Conclusions and future perspectives

A comprehensive understanding of the important role of Candida biofilm formation in RVVC, is crucial for the discovery of more effective antifungal therapies for this troublesome disease. Current experimental and clinical data, including that from our preliminary study, suggest a possible association between Candida biofilms and the pathogenesis of RVVC. A clinical study that recruits a substantial number of patients with accurately diagnosed RVVC or sporadic VVC, collecting vaginal tissue samples from patients at an asymptomatic period, and quantitatively and qualitatively analysing tissue samples is required to validate the presence of Candida biofilms in human RVVC. Further investigation is also needed to confirm the important role of Candida biofilm formation in the pathogenesis and persistence of RVVC, using an optimised animal model that properly represents the chronic nature of RVVC, clinical isolates from the vagina, and their genetically manipulated strains varying in biofilm formation, such as bcr1-knockout mutant strains, complementary strains, or over-expression strains.

Data availability

The data that support this study will be shared upon reasonable request to the corresponding author.

Conflicts of interest

The authors declare that they have no conflicts of interest.

Declaration of funding

This research is supported by an Alfred Research Trust Major Grant to Y. Qu.