Introduction

The first step of any mucosal infection is ‘colonization’, whereby a microbe ‘attaches’ to a host cell or moiety. Depending on the microbe, this can lead to either a degree of mutualism in the case of commensal microbes, or invasion of the epithelial barrier followed by disease and immune activation in the case of pathogenic microbes. In commensalism, the microbe and host live in a mutually beneficial relationship, whereby the microbe obtains all necessary nutrients to survive and grow on the mucosal surface without causing harm, while the host reaps the beneficial effects in ways that are often less well understood (e.g. utilizing microbial secondary metabolites and vitamins). This kind of commensal relationship is well documented for colonizing bacteria in the gut [1]. These commensal bacteria are important in general health since perturbations in the microbial gut flora can lead to multiple downstream effects, including aberrant immune responses and initiation of gut diseases such as inflammatory bowel disease (e.g. Crohn’s disease and ulcerative colitis) [2]. The consensus of opinion now is that there is a close relationship between microbiota and health at multiple mucosal sites [3].

Microbiomes, Fungi and the Oral Cavity

The interaction between a host and its microbiome is essential to health. Until recently, little information was available regarding the composition of the microbiota on mucosal surfaces and even less information with regard to identifying resident commensal microbes that could turn ‘pathogenic’ under suitably predisposing conditions. However, technological advances have permitted the ‘microbiomes’ of these sites to be studied in detail, providing a better understanding of the full microbial composition at various mucosal sites. These studies have focused heavily on bacterial gut microbiomes [3–6], although other mucosal microbiomes are now being investigated. Together, these studies demonstrate that the gut microbiome is the most diverse, followed by the oral microbiome, and the vaginal microbiome being least diverse [6]. Although each microbiome is diverse, it is also stable between individuals for the respective mucosal site [5]. Recently, there has been increased interest in the oral microbiome, which comprises viruses, bacteria, fungi, protozoa and archaea, as the bacteria present in this microbiome are responsible for the two most common diseases of humans: dental caries (tooth decay) and periodontal (gum) disease [7]. Unfortunately, although the presence of a fungal community on mucosal surfaces is well known, the fungal component of these microbiomes, the ‘mycobiome’, has largely been overlooked, despite fungi having the potential to be highly detrimental to human health.

Recently, the genomic characterization of the murine gut mycobiome has demonstrated that, although not homogeneous, surprisingly over 97 % of the population comprises just ten fungal species [8••]. Remarkably, a single fungus, Candida tropicalis, constituted over 65 % of the entire population in one mouse colony, demonstrating that Candida species can be a major part of the mammalian gut mycobiome. Notably, however, C. albicans and C. parapsilosis are the major fungal constituents in humans [9•]. Therefore, it would appear that differences exist between the mouse and human gut mycobiomes, and although the underlying reasons for this are unclear, these differences may provide clues as to why C. albicans is not a natural colonizer of mice. To this end, new murine models are continually being developed to specifically understand Candida colonization rather than infection [10–13].

Only one study has characterized the human oral mycobiome, undertaken in 20 healthy individuals using a multitag pyrosequencing approach and the panfungal internal transcribed spacer primers [14••]. The work identified 74 culturable and 11 nonculturable fungal genera. Of the culturable genera, Candida species were the most frequently observed (isolated from 75 % of participants), followed by Cladosporium (65 %), Aureobasidium, Saccharomycetales (50 %), Aspergillus (35 %), Fusarium (30 %) and Cryptococcus (20 %). Four of these genera (Candida, Aspergillus, Fusarium and Cryptococcus) are known to be pathogenic in humans. Notably, 36 % of fungi belonged to the nonculturable category. In total, 101 fungal species were identified. While the abundance of Candida in these individuals was clearly expected, the presence of Aspergillus, Fusarium and Cryptococcus in the healthy oral cavity is rather surprising, since these fungi are not generally recognized as oral colonizers. However, although the incidence of Cryptococcus and Aspergillus oral mycoses is rare, they have been reported in both immunocompromised and immunocompetent patients (e.g. in diabetes mellitus), though their occurrences are usually secondary to a more serious systemic infection [15]. One could postulate that low-level oral colonization of these fungi may provide a reservoir from which these species can migrate to drive disease under suitable predisposing conditions. Despite these data and speculations, the clinical relevance for the presence of a diverse population of fungal species in the oral cavity is still unclear [14••].

Candida and the Oral Cavity

By far the most common fungi endogenous to humans are Candida species. Although Candida species usually cause no pathology, these fungi can become pathogenic if there are alterations in the local environment, such as modifications in the normal microbiota or compromised immune defences [16]. As such, Candida species are the most common fungal pathogens in humans, giving rise to severe morbidity in millions of individuals worldwide. Oropharyngeal and vaginal infections are the most common manifestation and predisposing factors include antibiotic, glucocorticosteroid and hormone therapies, as well as diabetes mellitus and infections such as HIV and AIDS [17, 18]. Although C. albicans causes most of the Candida infections, non-C. albicans species such as C. glabrata, C. tropicalis, C. parapsilosis, C. guilliermondii and C. krusei are also pathogenic to humans and have emerged as important opportunistic pathogens in the oral mucosa [16]. In immunocompromised patients (e.g. transplant, cancer and intensive care patients), Candida species cause systemic infections with approximately 40 % mortality and are now the fourth most common hospital-acquired bloodstream infection [19]. Given that the majority of systemic Candida infections are acquired across the mucosae (predominantly gut), it is of paramount importance that we understand the basic biological mechanisms that normally restrict Candida species to mucosal surfaces.

The most abundant Candida species in the oral cavity is C. albicans, a polymorphic fungus that resides as a commensal in approximately 40–80 % of individuals, and is represented by a mixed strain population [20]. Notably, in the mycobiome study described above C. albicans was found in 40 % of individuals, followed by C. parapsilosis (15 %), C. tropicalis (15 %), C. khmerensis (5 %) and C. metapsilosis (5 %) [14••]. Surprisingly, C. glabrata was not found, despite being commonly isolated from the healthy oral cavity [21] and being the second leading species causing hospital-acquired Candida bloodstream infection [22]. Further, although C. metapsilosis is known to be an oral colonizer [23], C. khmerensis has not previously been described as a colonizer of any mucosal surface. Whether any interdependent relationships exist between these Candida species or other members of the oral mycobiome is not known, but future studies investigating potential interactions may shed light on the pathogenicity of these organisms [14••]. Given that C. albicans is the most common human fungal pathogen, this review predominantly targets the mechanisms utilized by C. albicans to promote colonization and infection of oral mucosal surfaces.

C. albicans Colonization and Infection of the Oral Cavity

Although C. albicans is normally a harmless commensal, the specific attributes required by this fungus to promote colonization rather than infection have rarely been studied [24]. However, in all likelihood, the same attributes are probably required for both colonization and infection, which in turn will be determined by the physiological and immune status of the host. It should also be noted that epithelial cells appear to possess innate antifungal activity (via annexin-A1) to restrict C. albicans growth [25], which may potentially help maintain C. albicans in the commensal state during health. However, when local environmental and/or immune conditions are perturbed, C. albicans is able to override these inhibitory mechanisms to cause local superficial mucosal infections, which require ‘virulence’ factors to promote pathogenicity. C. albicans possesses many putative virulence attributes that contribute to general survival, fitness and persistence within the host as well as specific factors associated with epithelial adhesion, biofilm formation, invasion, cell damage and induction/evasion of host responses [26–29]. The host defences include mechanical barriers to microbial penetration (such as epithelial cells), antimicrobial factors, and innate and cellular immune mechanisms [27, 30–33].

Epithelial cells at mucosal surfaces are the first point of contact with C. albicans and are therefor the first line of defence against the fungus. A prerequisite for colonization is the ability of C. albicans to adhere to epithelial cells, while invasion into and damage of epithelial cells are generally regarded as infection-specific qualities. Since the surface composition of the C. albicans cell wall is continually altering, especially during the yeast-to-hypha switch, the precise nature of C. albicans adhesion to epithelial cells is a complex and multifactorial process that involves several types of adhesins [34]. Notably, the interaction of the C. albicans yeast cell with epithelial cells is a potent stimulator of hypha formation [35, 36], and since hyphae are considerably more adherent than yeast forms, this indicates that specific hyphal proteins promote epithelial adhesion. This is supported by studies showing that wild-type C. albicans strains unable to produce true hyphae have a reduced ability to adhere to epithelial cells [37].

C. albicans Proteins that Promote Colonization and Infection of the Oral Cavity

Recent studies have implicated a number of fungal moieties as being important for epithelial adhesion (Table 1). A key adhesin family is the Als (agglutinin-like sequence) family, particularly Als3 which is a hypha-specific protein [38]. Als3 is one of eight Als proteins linked to glycosylphosphatidylinositol (GPI) that mediate adhesion to numerous host cells in which the N-terminal region is required for substrate binding [39, 40]. During in vitro oral epithelial cell infection, ALS3 gene expression is upregulated [35] and both gene disruption and Als3 protein overexpression studies have demonstrated a direct role for Als3 in adherence [40, 41]. Roles for other members of the Als family in epithelial adherence are somewhat contradictory. However, high-resolution structures of N-terminal Als1 and Als9 adhesins demonstrate that ligand recognition relies on a motif capable of binding flexible C-termini of peptides in an extended conformation. Central to this mechanism is an invariant lysine that recognizes the C-terminal carboxylate of ligands at the end of a deep-binding cavity [42•]. Additional structural studies of the Als3 adhesin will no doubt provide highly informative data regarding host–Candida interactions.

Table 1 C. albicans adhesins and invasins required for direct interactions with epithelial cells
Full size table

Hwp1 (hyphal wall protein 1) is another GPI-linked hypha-associated protein with a major role in epithelial adhesion. The N-terminal domain of this protein acts as a substrate for epithelial transglutaminases, enabling strong covalent links with other epithelial proteins [43•]. Like ALS3, HWP1 expression is upregulated during in vitro infection of oral epithelial cells [35] and is expressed in vivo in individuals colonized and orally infected with C. albicans [44]. HWP1 is required for mucosal pathogenicity as a hwp1∆/∆ mutant poorly adhered to murine oral epithelial cells and was significantly attenuated in virulence in a murine oropharyngeal candidosis model [45]. A huge number of hypha-associated genes exist, including well-known genes such as HYR1 and ECE1. Their precise roles in epithelial adhesion/interactions are still unclear, although Hyr1 may be required for pathogenicity in a murine oral biofilm model [46].

A number of other morphology-independent genes and cell wall-associated proteins play roles in epithelial cell adherence and mucosal pathogenicity. These include: surface GPI-proteins such as Iff4, Eap1, Utr2 and Ecm33; noncovalent wall-associated proteins such as Mp65, Phr1, Sun41 and Pra1; the integrin-like Int1 protein; secreted and cell-surface-associated proteases such as Sap1–Sap3, Sap9 and Sap10; internal endoplasmic reticulum (cell wall-processing) proteins such as Big1; and protein trafficking and vesicle transport proteins such as Vps11 [27]. However, the contribution of many of these proteins to epithelial adhesion and infection is likely to be indirect given that they possess complex, multifactorial functions that contribute to cell wall integrity and hypha formation.

C. albicans Invasion and Damage to Epithelial Cells

In commensalism, C. albicans normally resides on mucosal surfaces without causing disease or activating host defence mechanisms. However, if the correct environmental conditions arise, C. albicans can proliferate to cause superficial infections even in health (e.g. vaginitis). In these circumstances C. albicans invades the apical epithelial cells and may cause local damage, thereby eliciting host immune defences that ultimately reduce burdens to commensal levels or clear the fungus. These processes of invasion and subsequent damage induction are considered pathogenic rather than commensal attributes. C. albicans utilizes two invasion mechanisms for epithelial cell entry: induced endocytosis and active penetration [29, 36, 47]. The entry mechanism used is dependent on the epithelial cell lineage and appears to be triggered by hypha-associated factors [47, 48•]. The direct link between hypha formation and epithelial cell invasion is evident when screening C. albicans hypha-deficient mutants, which often exhibit adhesion and invasion defects [35]. Notably, the hypha-associated factors that promote adhesion, epithelial cell invasion and subsequent damage appear to be controlled by the fungal cAMP/PKA/Efg1 signalling pathway but not the MAPK pathways [35].

Induced endocytosis is mediated by the epithelial cell and occurs early during epithelial–C. albicans interactions, usually within 4 h [49]. Two C. albicans ‘invasins’, Als3 [48•] and Ssa1 [50], have been shown to induce endocytosis by oral epithelial cells (Table 1), which appears to be mediated via epithelial E-cadherin, through an actin-dependent mechanism requiring clathrin [51]. Once endocytosed, C. albicans hyphae enter the vacuolar endosomal compartments [52]. C. albicans hyphae-containing endosomes transiently express the early endosomal marker EEA1 but show marked defects in expression of the late endosomal marker LAMP1 and the lysosomal marker cathepsin D. Thus, C. albicans hyphae may prevent maturation of the endolysosome, potentially promoting intracellular fungal growth [52]. This survival strategy may allow C. albicans to survive and develop for extended periods within the epithelial cell, enabling the hypha to emerge out of one epithelial cell to infect a neighbouring cell. This will result in cell damage and immune activation, thus leading towards an infection phenotype. Similar survival strategy mechanisms might be employed by C. albicans to evade killing by macrophages [53, 54]. However, the direct role of Als3-mediated induced endocytosis in the induction of damage and immune activation of epithelial cells may be limited [55].

Active penetration occurs at later time points than induced endocytosis and requires the fungus to be viable. This process results in hypha invasion in between or through epithelial cells [35, 47] rather than being taken up by the cell. It is still unclear which C. albicans factors contribute to active penetration but roles for the secreted aspartic proteinases (Sap), lipases (Lip) and phospholipases (Plb) have been proposed [56–58] (Table 1). Of note, Sap5 activity appears to be required for E-cadherin degradation [59], which is present at high concentrations at interepithelial junctions. E-cadherin degradation may disrupt epithelial barrier function enabling hyphae to penetrate mucosal tissues. C. albicans hyphae that have directly invaded epithelial cells or the submucosal layers will induce tissue damage via necrosis and/or apoptosis [27, 35, 49]. Although apoptosis can be beneficial to the host, necrosis is nearly always detrimental. The precise contribution of each mechanism to damage induction is still unclear but maintenance of hypha formation appears critical for eliciting damage via necrosis [35, 36, 60], whilst cell surface N- and O-mannans may promote damage via apoptosis [61].

Non-C. albicans Colonization and Infection of the Oral Cavity

Historically, C. albicans accounts for approximately 80 % of all human candidoses [62], but over the past few decades non-C. albicans species, namely C. glabrata, C. tropicalis and C. parapsilosis, have emerged as important human pathogens [63]. C. glabrata is the second most prevalent human yeast pathogen after C. albicans but has distinguishing features, in that C. glabrata is haploid rather than diploid, does not form hyphae, and has an innate resistance to azole antifungals. Likewise, C. parapsilosis and C. tropicalis, the third and fourth most prevalent Candida pathogens, are diploid but do not form true hyphae. All three species fail to filament on epithelial surfaces and this probably explains their poor ability to infect epithelial cells compared with C. albicans [60, 64–66].

A key virulence factor of C. glabrata appears to be a major group of adhesins encoded by a large family of EPA (epithelial adhesin) genes [67], which have homology to the ALS genes of C. albicans. Although few studies have been performed on the C. glabrata Epa proteins, it appears that Epa1p is the main member of this family required for epithelial adhesion through its binding to N-acetyl lactosamine-containing glycoconjugates [68•]. Likewise, both C. parapsilosis [69] and C. tropicalis [70] possess multiple Als-like cell wall proteins, but limited data currently exist regarding their role in epithelial adhesion, colonization or mucosal infection. C. glabrata, C. tropicalis and C. parapsilosis all exhibit hydrolytic enzyme activity (thought to promote adhesion and invasion in C. albicans) to some degree, with all three species possessing SAP-like genes [71••, 72–74]. As with C. albicans [75–78], SAP gene expression and activity in C. parapsilosis [65] and C. tropicalis [64] is present at epithelial surfaces, where it is thought to contribute to mucosal colonization and infection. It should be noted, however, that the role of the C. glabrata SAP-like genes remains unclear [79]. C. glabrata, C. tropicalis and C. parapsilosis also exhibit Plb activity [63, 80, 81], but again their roles in mucosal colonization and infection are undetermined. Finally, unlike C. glabrata, C. parapsilosis possesses Lip activity, which appears to contribute to epithelial colonization/infection [82], and although the C. tropicalis genome contains LIP genes [71••], the role of these proteins in C. tropicalis colonization/infection is currently unknown.

Epithelial Discrimination Between Commensal and Pathogenic Candida

Given the apparent importance of hypha formation in epithelial invasion and pathogenicity, determining how epithelial cells are activated by the yeast and hyphal form of C. albicans might provide valuable information regarding how mucosal tissues discriminate between the commensal and pathogenic states of this fungus [83, 84]. Recent advances in this area have demonstrated that in oral epithelial cells C. albicans appears to activate two main signalling pathways, the mitogen-activated protein kinase (MAPK) and the nuclear factor-kappaB (NF-κB) pathways [85••]. The first, immediate but transient epithelial recognition event is independent of morphology and appears to constitute recognition of the fungal cell wall polysaccharides (mannans and β-glucans) via the NF-κB and all three MAPK pathways including p38, JNK (c-Jun N-terminal kinase) and ERK1/2 (extracellular signal-regulated protein kinase) [85••, 86]. This results in the activation of the p65/p50 transcription factor via NF-κB and the c-Jun transcription factor via JNK and ERK1/2. However, this is insufficient to fully activate the epithelial cells as proinflammatory cytokines are not induced. This recognition event may also occur during mucosal C. albicans colonization (commensalism). This lack of activation by C. albicans cell wall polysaccharides has also recently been demonstrated for skin keratinocytes [87], suggesting that these polysaccharides play a limited role in inducing epithelial/keratinocyte immune responses to fungal cells.

The second recognition event only occurs when C. albicans hyphae (but not yeast) are present at higher burdens. In these circumstances, there is strong activation of NF-κB and all three MAPK pathways which results in the activation of the c-Fos transcription factor (via p38) rather than c-Jun (via ERK1/2 and JNK). The combination of c-Fos and NF-κB activation appears to be sufficient for full activation of epithelial cells resulting in the induction of proinflammatory cytokines [85••]. This recognition event may occur during C. albicans invasion (pathogenicity). Thus, it has been proposed that epithelial MAPK-p38/c-Fos activation may constitute a ‘danger response’ mechanism that is kept in check to permit immune quiescence in the presence of low C. albicans burdens but immune activation when the C. albicans burden increases and becomes hyphal [27, 83, 84]. Interestingly, recent work has also demonstrated that tissue macrophages also respond selectively to C. albicans hyphae, although recognition in macrophages appears to occur through hyphal wall-specific β-glucan moieties [88]. Importantly, the MAPK-p38/c-Fos pathway is activated only by hypha-forming Candida species (C. albicans and C. dubliniensis) but not by C. tropicalis, C. glabrata, C. parapsilosis or C. krusei [60], and is also present in human vaginal epithelial cells [89].

Therefore, this MAPK-p38/c-Fos mechanism may enable different mucosal tissues to recognize invading Candida hyphae, thereby potentially identifying when this commensal fungus has become pathogenic. Together with similar findings in Caenorhabditis elegans (nematode worm) with C. albicans [90] and in murine intestinal epithelial cells with bacterial pathogens (Citrobacter rodentium) [91•], the data suggest that MAPK-p38 signalling may be required for epithelial recognition of ‘pathogenic’ microbes and to initiate inflammatory responses. The epithelial receptors mediating hyphal recognition via p38 are currently unknown but do not seem to involve Toll-like receptors or C-type lectins [85••], even though they are expressed in epithelial cells [92•, 93]. Therefore, epithelial cells appear to utilize sensing mechanisms for immune activation that are different from those in myeloid cells [94], possibly as they target alternative fungal moieties. This field of epithelial pattern recognition of Candida will no doubt be an active area of future research.

Conclusion

Over the past few years our understanding of the mechanistic events involved in host mucosal responses to Candida pathogens has grown significantly. Originally, the role of epithelial cells in these events was considered to be relatively unimportant, but it has become increasingly apparent that they are critical in fungal recognition and host defence. For C. albicans, hypha formation appears key to the host–pathogen interaction event at mucosal surfaces, since it is required by the fungus to promote adhesion, colonization and invasion, and is targeted by the epithelial cell to potentially discriminate between commensal and pathogenic states of the fungus. These advances provide important insights into the complex mechanisms by which appropriate host immune responses may be initiated and how this affects fungal colonization and infection.