“Candida Albicans Interactions With The Host: Crossing The Intestinal Epithelial Barrier”

ABSTRACT

Formerly a commensal organism of the mucosal surfaces of most healthy individuals, Candida albicans is an opportunistic pathogen that causes infections ranging from superficial to the more life-threatening disseminated infections, especially in the ever-growing population of vulnerable patients in the hospital setting. In these situations, the fungus takes advantage of its host following a disturbance in the host defense system and/or the mucosal microbiota. Overwhelming evidence suggests that the gastrointestinal tract is the main source of disseminated C. albicans infections. Major risk factors for disseminated candidiasis include damage to the mucosal intestinal barrier, immune dysfunction, and dysbiosis of the resident microbiota. A better understanding of C. albicans’ interaction with the intestinal epithelial barrier will be useful for designing future therapies to avoid systemic candidiasis. In this review, we provide an overview of the current knowledge regarding the mechanisms of pathogenicity that allow the fungus to reach and translocate the gut barrier.

Graphical abstract

Invasion of C. albicans through the intestinal epithelial barrier.

KEYWORDS:

• Candida albicans
• interaction
• adherence
• invasion
• damage
• enterocyte
• intestinal barrier
• infection models

Introduction

The importance of the intestinal mucosa in the interaction of candida sp. with its host

Formerly described as a harmless commensal of the human mucosal surfaces, C. albicans is also an opportunistic fungus responsible for candidiasis that can range from superficial to invasive and life-threatening infections in debilitated patients, mostly occurring in the hospital setting. Thus, systemic candidiasis is associated with a high crude mortality ranging from 20 to 49%,^1–4 and almost all of the organs can be secondarily infected following the hematogenous dissemination of the fungus. From a pathophysiological point of view, invasive candidiasis has various stages where C. albicans has to: (i) enter the bloodstream, (ii) survive in the blood reservoir and (iii) escape from the bloodstream in order to establish deep-seated infections.⁵ In the first stages, a clear association between the presence of C. albicans on mucosal surfaces such as the mouth or the vagina and its translocation into the blood has not been established yet. However, various observations strongly suggest that invasive candidiasis are mostly of endogenous origin with the gastrointestinal (GI) tract being the main portal of entry into the bloodstream.^5,6 Indeed, molecular typing studies have pointed to gut mycobiota as the main origin of disseminating C. albicans isolates in the blood.^7–11 On the host side, immune dysfunction such as neutropenia, damage to the mucosal barrier and dysbiosis of the bacterial microbiota have been identified as major risk factors for invasive candidiasis.^9–13 With in vivo investigations, Koh demonstrated a strong association between (i) dysbiosis of the bacterial microbiota, (ii) mechanical alteration of the gut barrier function, (iii) a decrease in the function of neutrophils (PMNs) and (iv) major risks for systemic candidiasis.¹⁴

In addition to the role of C. albicans as a commensal or a pathogen and with the advance of Next Generation Sequencing (NGS) for microbiome analysis, recent reports suggest that changes in the fungal microbiota of the gut are linked to the pathogenesis of multiple gastrointestinal (e.g. inflammatory bowel diseases (IBD))^15,16 or extra-intestinal disorders (e.g. allergic airway diseases, neurologic disorders, cancer),¹⁷ in which, an increase of colonization by fungal species, including C. albicans, was reported compared to healthy subjects.^15,18,19

All in all, C. albicans is a versatile yeast able to interact with the gut content of its human host both as a commensal, a pathogen or a major colonizer of the GI tract. This transition from commensalism to pathogenicity involves both host-related factors (e.g. environment of the digestive tract, the gut mucosa, genetics and health status), the microbiota (i.e. dysbiosis), and C. albicans itself. In this context, the purpose of this review is to discuss the interaction between C. albicans and the intestinal epithelial barrier.

Interaction of microorganisms with the intestinal mucosa

Intestinal factors influencing the interaction of micro-organisms with the gut mucosa

Human intestinal microbiota and mycobiota

The human intestinal microbiota is a complex ecosystem, largely dominated by bacteria in terms of number (i.e. around 10¹⁴ bacterial cells) and species diversity (500 to 1000 species, mainly anaerobes). However, the global gut microbiota also comprises archaea, viruses and eukariota such as parasites and fungi.²⁰ Whereas the precise contribution of gut-associated archaea, viruses and parasites in human physiology remains to be specified, research over the last few decades has given us a more comprehensive view of the contribution of bacteria and fungi to gut-associated microbiota (GM) and the potential impact on human health.^21,22 In healthy adults, the composition of the GM is unique to each individual.²³ It differs along the digestive tract²⁴ and varies according to environmental and lifestyle stimuli.²⁵

Although specialized in the digestion of nutrients, the GM contributes to the host’s gut homeostasis by (i) modulating the host’s energy metabolism,²⁶ (ii) sustaining gut integrity,²⁶ (iii) preventing gut colonization by pathogens^22,26 and (iv) promoting the development and maturation of the gut-associated immune system.^26,27 Consequently, an imbalanced GM composition can result in the dysregulation of host homeostasis contributing to the onset or progression of many diseases. Indeed, dysbiosis of the GM is associated with the pathogenesis of intestinal (e.g. IBD, irritable bowel syndrome, coeliac disease) but also extra-intestinal diseases (e.g. allergy, asthma, disruption of the immune system, metabolic diseases, cardiovascular diseases, and neurologic diseases).^28–32

Among microbiota components, fungi account for about a quarter of the species diversity but for less than 0.1% of the GM in quantity.⁷ The human mycobiota comprises 158 genera and 390 species among which only 221 are found in the digestive tract.⁷ However, despite growing interest in the gut mycobiota, several issues remain to be addressed regarding its precise composition from one individual to another. The most commonly-detected fungi in the human GM (in order of frequency) are the genera Candida, Malassezia, Aspergillus, Debaryomyces, Penicillium, Cladosporium, Trichosporon, Galactomyces, Saccharomyces and Cryptococcus (i.e. the percentage of positive samples reported in gut mycobiome studies being 80, 25, 24, 21, 20, 18, 9, 9, 6 and 4% respectively).³³

However, the mycobiota appears less stable than the bacterial microbiota and highly subject to changes brought on by environmental factors, particularly diet (e.g. animal-based versus plant-based diet).³³

Although, the role of fungi in the human GM remains poorly understood, recent studies have highlighted the role of diverse fungal communities in host physiology. Symbiotic relationships between fungi inhabiting the human GI tract and the host have indeed been reported to be essential for digestion functions.³⁴ Additionally, other findings have underscored the relationship between the gut mycobiota and human diseases. Associations between the mycobiome biodiversity in IBD or in obesity, and gut environment modifications have been reported.^15,35 For example, Mar Rodriguez et al. observed differences in the biodiversity of the mycobiome of obese compared to healthy individuals, suggesting that this fungal dysbiosis contributes to changes in the lipid and glucose metabolisms observed in diabetic patients.³⁵ Finally, antifungal or antibiotic treatments can also lead to microbial dysbiosis and so to a disruption in the balance between bacterial and fungal communities, highlighting the interdependence of fungi and bacteria in the gut.¹⁷

Intestinal Epithelial Cells (IECs)

The intestinal epithelium (IE) consists of a monolayer of cells covering a surface of ~400 m², organized into crypts and villi. Pluripotent intestinal epithelial stem cells residing in the bottom of the crypts ensure a continuous renewal of the IE, where the local environment drives their proliferation and functional differentiation. IE encompass differentiated cell types grouped under the term intestinal epithelial cells (IECs) (i.e. enterocytes, enteroendocrine cells, goblet cells, Paneth cells, Microfold (M) cells and Tuft cells) with specialized functions (i.e. absorption, hormone secretion, mucus secretion, antimicrobial peptide (AMPs) production, antigen sampling and taste-chemosensory responses respectively) in addition to Cup cells whose function remains to be specified.³⁶ Additionally, IECs exert immunoregulatory functions that are critical for the development, maturation and homeostasis of the immune system all along the gut mucosa. Finally, thanks to this complex system of cells and functions, IECs form a physical and biochemical barrier capable of segregating microorganisms from the host with the ability to discriminate commensals from pathogenic microorganisms. IECs respond differentially to commensal or pathogenic microbial signals that consequently reinforce or weaken their barrier function. In parallel, the mucosal biochemical and immune systems orchestrate appropriate immune responses to these microbial signals that can range from the tolerance of commensal microorganisms to anti-pathogenic responses.³⁷

Physical barrier function

IECs are mostly composed of enterocytes (over 80%) that form a monolayer of differentiated and polarized epithelial cells with an apical side that displays microvilli which form the brush border exposed to the digestive lumen.³⁸ Beyond their role in nutrient and fluid absorption, enterocytes also contribute to the integrity and impermeability of the IE by establishing the intestinal border. They form a physical barrier between the content of the gut lumen and the underlying tissues, limiting the translocation of microbes and their related products.^39,40 Thus, the cohesion of the IE relies particularly on junctional intercellular complexes including tight-junctions (TJs) and weak junctions (i.e. adherens-, gap-junctions and desmosomes) (Figure 1). The composition and functions of these junctional complexes vary by IEC type, especially for TJs,⁴¹ but their structural organization has been extensively studied in enterocytes (Figure 1). Between enterocytes, adherens junctions (AJs) contribute to cell-cell adhesion and intracellular signaling whereas gap-junctions and desmosomes participate in cell-cell communication and adhesion, respectively. Together with AJ, TJs form a seal (i.e. the apical junctional complex) that ensures the architectural cohesion and integrity of the IE. This complex is closely linked to the cytoskeleton that is crucial in maintaining the cohesion and structure of the IECs by extending throughout the cytosol (Figure 1).^42–44 Finally, TJs, the most apically located intercellular junctions, are crucial for the IE’s role of physical barrier since they are targeted by numerous physical and chemical factors,^44-51including cytokines, proteases, hormones, neurotransmitters, dietary components, bacterial and fungal toxins, and xenobiotics that can alter TJs and consequently modulate the permeability and integrity of the IE. Commensal bacteria can modulate impermeability and integrity in a homeostatic manner by targeting TJs,^47,52–54 whereas enteric pathogens including bacteria, viruses, parasites^55-57 and possibly fungi⁵ have developed strategies aimed at exploiting TJs to promote their invasiveness into IECs.

Figure 1. The intercellular junctions between enterocytes at the digestive barrier.

(A) Composition and organization of the enterocyte-enterocyte junctions in the intestinal epithelium. Tight junctions (or TJs) are the first intercellular junctions present at the apico-lateral region of enterocytes followed by the adherens junctions (AJs), the desmosomes and finally the GAP junctions at the baso-lateral region. (B) Composition and organization of the TJs and AJs. The TJs and the AJs form circumferential junctions composed by transmembranous proteins, including Claudins, TJ associated Marvel domain containing Occludin, Tricellulin and Marvel D3, and JAMs for TJs and E-cadherin and Nectin for AJs, all connected to the cytoskeleton through various proteins (i.e. Zonula Occludens (ZO1-3) proteins, Cinguline, Catenin and Afadin).

Secretory functions of IECs

IECs also express secretory functions reinforcing the physical barrier role of the epithelial layer and ensuring biochemical functions.

First, enterocytes secrete the glycocalix that mainly contains transmembrane mucin glycoproteins covering the apical dense microvilli surface, which locally protects the enterocyte cell membranes.⁵⁸ Moreover, the IE is lined with a viscous layer of mucus that is synthesized and extruded at the cell surface by specialized secretory IECs, i.e. the goblet cells,⁵⁹ that are dispersed throughout the gastro-intestinal tract⁶⁰ (Figure 2). In addition to other proteins and lipids, the mucus is predominantly composed of secreted mucins (gel- and non-gel-forming mucins). This class of high molecular weight glycoproteins, strongly O-glycosylated, confer the viscous, gel-like quality to the mucus.^58,61,62 MUC2 is the largest gel-forming mucin expressed in the intestinal tract.⁶³ Though its composition and amount vary along the gastrointestinal tract (Figure 2), this continuous viscous layer lubricates the IE, regulates the transport of molecules from the gut lumen to IECs and forms a physical barrier that (i) protects the IE against acidic and proteolytic environments encountered in the gut lumen and (ii) limits the access of most microorganisms and deleterious substances at the surface of the IE. However, enteric pathogens including eukariota have developed strategies to overcome the mucus layer, enabling them to reach and colonize the epithelial barrier to further cause infection⁶⁰ (reviewed in^50,64). These include (i) the production of lytic enzymes to disrupt the mucus layer^65-70 and/or (ii) the downregulation of mucus production by goblet cells or changes in its composition.⁷¹ Additionally, components of the mucus can trigger specific transcriptional responses in enteric pathogens that facilitate their interaction with the epithelial layer and/or promote their virulence.^72,73

Figure 2. Organization of the gastro-intestinal tract.

The gastro-intestinal tract forms a physical and biochemical barrier capable of segregating microorganisms from the host with the ability to discriminate commensals from pathogenic microorganisms. The physical barrier consists of a monolayer of cells, which includes various intestinal epithelial cell types (IECs) differently organized from the small intestine to the colon. The biochemical barrier consists of a mucus layer whose composition, structure and also properties differ between the small intestine and the colon. In the small intestine, the mucus is composed of a highly dynamic monolayer, not anchored to the surface of the epithelial cells. This monolayer of mucus is permeable to the bacteria. However, the distal peristaltic movements keep the microorganisms away from the surface of the epithelial cells. In the colon, the mucus is organized in two layers: the inner layer and the outer layer. The inner layer is in perpetual renewal (approximately every 1 to 2 hours) in order to remain totally germ-free. This layer is firmly anchored to the epithelial barrier through the interaction between mucins in the mucus and the mucins-binding protein located at the surface of the epithelial cells. Due to its size-exclusion filter function (i.e. exclusion of any element of more than 0.5 μm), the inner layer of mucus is impermeable to microorganisms. Finally, the outer layer is the normal habitat of intestinal commensal microbiota. In addition to the secretion of bioactive molecules such as nutrients, hormones, neuropeptides, cytokines and lipids in the gut lumen, the gut microbiota takes an active part in this permanent remodeling as highlighted in germ-free animals in which a thinner layer of mucus is observed as the result of a decreased number of goblet cells compared with conventional animals.^61–63 The outer-most layer of mucus is a reservoir of dense populations of commensal microorganisms whose composition is linked to the existing luminal populations.⁶⁴ Consequently, normal or altered GM and mycobiota will influence goblet cell function as well as the composition and volume of the mucus layer by mechanisms probably involving both the direct effect of locally released microbial factors and/or the indirect effect of bioactive or immune factors resulting from the host-response to intestinal microbes. Moreover, the mucus layer is also a biochemical barrier thanks to the presence of various secreted antimicrobial factors mostly secreted by the cytoplasmic granule-rich Paneth cells (PCs). These PCs are located at the base of the small intestinal crypt in healthy individuals.⁶⁵ Various factors, including cholinergic agonists, bacteria and bacterial products (such as lipopolysaccharides and lipoteichoic acid),^66,67 lead to the discharge of PC granules from the crypt into the mucus layer, forming a biochemical barrier that is crucial for establishing baseline homeostasis for mucosal and systemic inflammatory response. The PCs mainly secrete a panel of antimicrobial peptides (AMPs), which are released in intestinal mucus layer in the small intestine in humans. The composition of the mucus changes along the GI tract, contributing to the increase of the amount of microorganisms in the digestive microbiota between the small intestine and the colon.

In addition to the being a physical barrier, the mucus layer harbors a biochemical function. It serves as a reservoir for bioactive factors mainly secreted by IECs in order to reinforce the barrier function by maintaining GM-Host homeostasis as well as influencing the composition of the GM. Among these components, antimicrobial factors are essentially secreted by PMNs residing in the lamina propria and by specialized IECs named Paneth Cells (PCs) that reside exclusively at the base of the crypts in the small intestine of healthy individuals⁷⁴ (Figure 2). These specialized cells contain granules whose secretion is modulated, among others, by microbes and microbial products^75,76 through the stimulation of specific Pattern Recognition Receptors (PRRs) located at the surface of IECs.²⁷ These granules are mostly enriched in antimicrobial peptides (AMPs) including α- and β-defensins,^77,78 lysozyme C,⁷⁹ secretory group IIA phospholipase A2 (sPLA2),⁸⁰ Regenerating Islet-Derived 3 Gamma (REG3γ),⁸¹ α-1 antitrypsin,⁸² Angiogenin 4⁸³ and Cathelicidin.⁸⁴ PCs also release cytokines (IL-17A, TNF-α, IFN-β and IL-1), proteases (Metalloprotease 7), proteins and scaffolding molecules as well as non-granule molecules including Wnt proteins, Epidermal Growth Factor and Notch ligands.⁸² These non-granule secretions create a specific microenvironment called the “stem cell niche”, that surround the stem cells and is essential for the regeneration and differentiation of specific IECs in the small intestine.^82,85

Additional secreted antimicrobial factors are found in the intestinal mucus layers, including secretory Immunoglobulin A (sIgA). sIgA are secreted by IgA+ B cells moving from the germinal center of Peyer’s patches to the lamina propria in intestinal villi.⁸⁶ They provide an adaptive immune component which is crucial for maintaining gut homeostasis by (i) inhibiting microbial motility and improving the mucus’ ability to trap microorganisms,⁸⁷ (ii) clearing the microorganisms through peristalsis or mucocilary exclusion,⁸⁸ (iii) shaping and maintaining the GM composition⁸⁹ and (iv) neutralizing the bacterial toxins.^90,91

Furthermore, IECs constitutively produce a panel of cytokines, including TGF-α, IL-10, IL-15, and IL-18, that are crucial for the basal recruitment of immune cells, the regulation of IEC growth and so gut homeostasis.^92–95 These cytokines also up-regulate the production of other cytokines (i.e. IL-1α or β, IL-6, IL-7, IL-8, TNF-α and TGF- β) and/or chemokines by immune cells during microbial infection.^92,96–98

Finally, bacterially derived antimicrobial factors, among them the peptides called bacteriocins, are secreted by gram-positive and gram-negative bacteria in the intestinal mucus layer. These bacteriocins have a restrained antimicrobial activity spectrum that affects the composition of the microbiota (interdependency within bacterial communities).⁷⁴ Moreover, the gut microbiota produce several metabolites (from anaerobic fermentation of exogenous undigested dietary components) such as short-chain fatty acids (SCFAs), aryl-hydrocarbon receptor ligands or polyamines, which interact with host cells and thus influence immune response.⁹⁹ Altogether, the small intestine is constantly exposed to bacteria and bacterial products, which explains the constant baseline level of secreted antimicrobial factors in the mucus layer.^100,101

Gut mucosal immunity

IECs play a central role in the intestinal immune response, acting as frontline sensors between GM and the immunity of the gut mucosa.⁹⁶ A close relationship exists indeed between the digestive microbiota and the intestinal mucosa, mediated by a complex intestinal immune system. This mucosal immune system (or MIS) is divided into three physically distinct parts: (i) the intestinal epithelial barrier, (ii) the lamina propria (LP) and (iii) the gut-associated lymphoid tissue or GALT, comprising Peyers’ patches (or PPs), isolated lymphoid follicles (or ILF) and mesenteric lymph nodes (or MLNs). Functionally, the intestinal MIS comprises two immune sites, (i) the effector and (ii) the inductive site¹⁰² (Figure 3). They are challenged with discriminating entero-pathogenic microorganisms from the resident commensal flora, and adapting the immune response as needed, from tolerance to inflammation.¹⁰³ These aspects have been extensively addressed elsewhere²⁷ and will not be developed in this review. However, the specialized cells of the immune system are of particular interest in the context of this review because they contribute to the translocation of micro-organisms through the gut barrier.

Figure 3. Organization and function of gut mucosal immunity.

In this context, an important effector site of gut mucosal immunity is the lamina propria layer, which has the key functions of (i) preventing entry/spread of pathogens across gut mucosa and (ii) destroying invasive pathogens.¹⁰² Indeed, almost all of the intestinal immune cells are found in the LP layer; this includes innate immune cells (including dendritic cells (DCs), macrophages (Mφ), Mucosal-associated invariant T cells (MAIT), mast cells and innate lymphoid cells), and adaptive immune cells (Lymphoid T cells and plasma cells).¹⁰⁴ These immune cells have a crucial function in the interaction with the GM. Among them, the γδ intraepithelial T lymphocytes (that represents the major T cell population within the intestine) respond directly to microbiota signals to promote intestinal homeostasis which (i) helps preserve the integrity of damaged epithelial surfaces¹⁰⁵ and (ii) produces innate antimicrobial factors and pro-inflammatory cytokines and chemokines in response to resident bacteria that penetrate the intestinal epithelium and to intestinal injury.^106,107 Another important line of immune cells for maintaining intestinal homeostasis are the CD4 regulatory T cells (Tregs). Indeed, Tregs are important response modulators to gut microbiota, notably through the secretion of anti-inflammatory cytokines such as IL-10, which trigger an immunosuppressive response and gut homeostasis.¹⁰⁸ In parallel, the innate lymphoid cells (ILCs) contribute to host defense and immune homeostasis within GALT.¹¹³ Thus, ILCs are able to rapidly respond to gut microbiota contributing to GALT formation and function (Reviewed in 112).

Among GALT, PPs consist of dome-like structures or sub-epithelial domes (SED) containing large B cells follicles and smaller T cells areas, in addition to migratory dendritic cells (DCs) and macrophages (Mφ). These SED are separated from the gut lumen by a single epithelial layer called the Follicle-Associated Epithelium or FAE. The FAE contains epithelial cells, including microfold (M) cells. When compared to the adjacent enterocytes, these M cells display reduced irregular microvilli (short fold-like invagination or microfold) on the apical surface.²⁷ Moreover, M cells express various apical receptors (e.g. integrin β-1, GP2, C5a and poliovirus receptors) that serve as a portal of entry for luminal components.¹¹⁰ At the basolateral membrane, M cells present a large pocket-like invagination that contains B lymphocytes (LB) and T lymphocytes (LT) in equal proportions, Mφ and DCs.¹¹¹ This particular structure in the M cells act as antigen-presenting cells, sampling gut luminal antigens, macromolecules and microorganisms present in the intestinal lumen in different ways (i.e. transcytosis, phagocytosis, microvesicle shedding) without any degradation.^27,112 Consequently, the translocation of antigens through M cells is pivotal for linking the contents of the gut to the immune system, with the aime to trigger an appropriate immune response to the harmless or harmful stimuli in the gut lumen. In a healthy host, this process largely contributes to the tolerance toward antigens derived from diet or the commensal flora through signaling pathways involving an Ets (E-twenty-six-specific sequence) transcription factor, SpiB, which acts as a major regulator of the structural and functional maturation of M cells.^113,114

On the contrary, many pathogenic microorganisms including reoviruses, polioviruses,¹¹⁵ and enteric bacteria (e.g. Salmonella typhimurium, Shigella flexneri, Yersina enterocolitica)^116–119 have developed strategies to exploit (i) weaknesses in the cell-cell junctions bordering M cells and (ii) the absence of mucus covering these specific cells. Consequently, M cells form a possible “gateway” that allow pathogens to enter and invade the host, triggering deleterious immune responses.¹²⁰

Candida albicans’ interaction with the intestinal barrier

C. albicans is an opportunistic pathogen that interacts dynamically with its human host. Indeed, when colonizing diverse mucosal surfaces of the human host as a commensal, C. albicans’ extensive phenotypic plasticity makes it capable of causing life-threatening systemic infections.¹²¹ It is a polymorphic fungus that exhibits in vivo yeast or hyphal (and possibly chlamydospore) forms, and the ability to transition from one to another is considered a virulence trait of the fungus.^122,123 It also expresses metabolic flexibility that largely contributes to its adaptability and virulence to diverse host niches. This phenotypic plasticity is best exemplified by the white and opaque phenotypes that correspond to distinct morphological states of the same isolate able to undergo heritable and reversible epigenetic transitions.¹²⁴ In addition, this phenotypic and metabolic flexibility allow the fungus to adapt to specific fluctuating external stresses, including nutrient conditions, temperature and pH.¹²⁵ These stresses drive the switch of C. albicans to the white or opaque state, each displaying specific biological features affecting, among others, mating competency, immunogenicity, virulence and niche specificity.^130–135 While white cells are the most clinically relevant form of C. albicans and considered as an all-around phenotype fitting various environmental conditions, opaque cells correspond to a more metabolically specialized form of the fungus.¹²⁹

The gut environment forms a specific niche characterized by a number of environmental factors including pH, temperature, hypoxia, and the presence of serum or carbon sources. There are also a wide range of microbial and nutrient metabolites that can influence morphological changes in C. albicans and the phenotype of the fungus.¹³⁶ Interestingly, Pande K et al. observed a morphological switch to a “dark” phenotype in C. albicans cells exposed to the mammalian gut environment.¹³⁷ Indeed, when the Wor1 transcription factor is necessary and sufficient to drive the switch from the white to the opaque state,^138–140 WOR1 overexpression triggered the switch to a Gastrointestinally-IndUced Transition (GUT) morphotype optimized to the gut living in a commensal state.¹³⁷ Finally, the gut niche appears to be a complex environment capable of driving alternative phenotype switches that promote commensalism or pathogenicity in C. albicans as the result of specific combinations of environmental stresses. However, how and which variations of the gut environment favor the commensal or virulent form of C. albicans remain to be specified. Thus, the passage of C. albicans through the gut mucosal barrier is a complex process depending on host characteristics that are both extrinsic and intrinsic, associating phenotypic switches that promote C. albicans commensalism or pathogenesis.

Notably, the pathophysiology of candidiasis is intimately linked to the ability of C. albicans to adhere to and invade host cell barriers. During these processes, C. albicans has to face different components of the gut mucosal barrier, including the antimicrobial peptide armory and the mucus layer, to successfully reach and then invade the epithelial layer lining the gut mucosa, possibly resulting in blood dissemination.

Interaction of C. albicans with the gut microbiota

In the digestive tract, 221 fungal species are part of the microbiota, mostly yeast and filamentous fungi belonging to the Ascomycota, Basidiomycota and Zygomycota phyla. In healthy humans, this mycobiota encompasses almost 66 genera of fungi, mainly including Saccharomyces and Candida species.⁷

The equilibrium of the healthy microbiome is under the control of transkingdom interactions between the different inhabitants of the microbiota. Among them, bacterial-fungal interactions are particularly important for the homeostatic state, as highlighted by Iliev et al. who reported the existence of mixed fungal-bacterial biofilms in the GI tract of mice.¹⁴¹ Fungal-bacterial interactions create a complex balance between synergism (e.g. C. albicans/Streptococcus sp.) and/or antagonism (e.g. C. albicans/Lactobacillus sp.) pathways, leading to a state of harmony within the microbiota (reviewed in Wang et al., 2014).^142–144

In this context, the GI microbiota exerts anti-Candidal properties through different mechanisms. Indeed, the bacterial microbiota can modulate C. albicans colonization through the production of bacterial metabolites that limits fungal proliferation and virulence.¹⁴⁵ Such bacterial metabolites exert antifungal activities in vitro through molecular mechanisms involving the C. albicans’ TOR pathway, a central signaling pathway that controls cell growth in response to environmental nutrient signals in eukariota.¹⁴⁶ For instance, Lactobacillus sp. limits C. albicans’ growth and virulence, affecting the germ-tube formation¹⁴⁵ in the GI tract through the production of H₂O₂, organic acids¹⁴⁷ and short chain fatty acids (SCFAs). Thus, bacterial metabolites inhibit C. albicans’ hyphal growth and invasion of human enterocytes by repressing the yeast-to-hyphae transition as well as the inducibility of hyphae specific transcripts.¹⁴⁶ In addition, other metabolic products, including dietary products such as tryptophan, exert antifungal activities by enhancing the mucosal reactivity to C. albicans. Indeed, tryptophan promotes the local transcription of a bacterial gene that triggers IL-22 productive cells and the proliferation of IL-22-producing innate lymphoid cells in the GI tract.¹⁴⁸ As a result of its pivotal role in innate antifungal resistance,^149,150 the resulting increased production of IL-22 promotes resistance to colonization by C. albicans by attenuating inflammation in the IECs and damage caused by C. albicans.¹⁴⁹

As a consequence, disturbances in the bacterial community (or dysbiosis) are associated with enhanced C. albicans colonization,^151,152 as reported early in the 1970’s.^153,154 These observations were underscored in germ-free mice that experienced an increase in colonization as the result of the lack of a bacterial microbiota.¹⁵⁵ Various iatrogenic factors including antimicrobial and immunosuppressive therapies have been reported for their capacity to disrupt the GI microbiota, thus modulating C. albicans colonization.^144,156,157 Fan et al. showed that antibiotics specifically targeting anaerobic bacteria, Bacteroidetes and Firmicutes, favored colonization of C. albicans in the gut of mice.¹⁵⁸ These authors observed that anaerobic bacteria from mature adult microbiota were critical for maintaining resistance to C. albicans since Bacteroides thetaiotamicron was capable of inhibiting colonization through the activation of the Hypoxia Inducible Factor (HIF)-1α.¹⁵⁸ This transcription factor is an essential regulator of mammalian innate defense, that, once activated, induces the production of the Cathelicidin-AMPs LL37, a critical immune effector for limiting C. albicans colonization.¹⁵⁸ Conversely, the use of antifungal therapies can contribute to fungal dysbiosis in which some species are reduced while others expand. This was recently reported by Wheeler et al. who observed worsening colitis and allergic airway disease in experimental mice models. The administration of oral antifungal therapies correlated with an increase in Aspergillus sp., Wallemia sp. and Epicoccum sp. Colonization, whereas Candida sp. colonization was decreased.¹⁷

Notwithstanding iatrogenic factors, many metabolic diseases, including GI disorders, have been linked to disturbances in C. albicans colonization.^15,18 Over-colonization observed during bacterial dysbiosis may aggravate the severity of ulcers and inhibit healing associated with IBD.¹⁵⁹ Additionally, a reduced diversity in both the fungal and bacterial gut flora, with a specific increase in the Candida genus, was reported in IBD patients.¹⁹ Recently, Sokol et al., highlighted a higher proportion of C. albicans in stools of IBD patients compared with healthy subjects, which points to a disease-specific fungal mycobiota dysbiosis.¹⁵ Collectively, these data strongly suggest the involvement of C. albicans in the pathogenesis of IBD.

It is now clear that overgrowth of C. albicans on GI mucosal surfaces contributes to disseminated candidiasis, especially in weakened patients in the hospital setting who are exposed to iatrogenic factors that favor C. albicans’ colonization, such as antimicrobial therapies and/or chemotherapies.^144,156,157 However, when C. albicans colonization is associated with a metabolic disorder, including GI disorders, it is not clear whether the increased colonization results from or contributes to the pathogenesis of the disorder. Further studies are expected to specify (i) the inter-communication between C. albicans and bacteria at the species or community level in the gut, with emphasis to mixed biofilms and their role in the gut homeostasis and GI disorders; (ii) the influence of C. albicans’ over- or sub-colonization on the host metabolism and local or systemic immune response and (iii) the biological features of C. albicans associated with various digestive disorders.

Interaction of C. albicans with the gut mucosal barrier

Tackling the AMPs armory and the mucus barrier

As seen above, bacteria are essential for an efficient innate immune response against C. albicans, notably through the induction of AMP production, including LL37, histatin 5 and β-defensins.^158,160 These AMPs exert anti-Candidal activities by (i) direct induction of cell death through mechanisms promoting pore formation, membrane depolarization and/or osmotic dysregulation in yeasts, (ii) immune modulation (through the promotion of PMNs/monocyte recruitment, TNF-α expression, chemoattraction of immature DCs and T cells) restraining C. albicans proliferation, and (iii) the repression of C. albicans virulence traits (e.g. α-defensin-6 prevents adherence to and invasion into enterocytes as well as biofilm formation¹⁶¹) (Reviewed in Swidergall et al., 2014).¹⁶⁰ During commensalism, the continuous but weak Pathogen Associated Molecular Patterns (PAMPs)/PRRs interactions lead to a low level of Nuclear Factor kappa B (NF-kB) pathway activation and so to a basal level of production of AMPs.^162,163 Upon increased colonization and infection, a strong up-regulation of the ERK and JNK MAP kinase pathways occurs, boosting inflammatory and damage responses of the host cells that will secondarily increase the production of AMPs by epithelial cells.¹⁶⁴

C. albicans has developed a three-phase strategy to avoid the candidacidal activities of AMPs. First, C. albicans secretes peptide effectors which protect against a broad range of^165-167AMPs. This is exemplified by the Msb2 membrane glycoprotein whose extracellular domain, i.e. Msb2*, is released after cleavage in large amount in the extracellular environment, inactivating a large panel of AMPs.^167–172 Interestingly, some data suggest the involvement of secreted aspartyl proteases (Saps) in the cleavage and release of Msb2*, while other observations corroborate the putative involvement of Saps as direct inhibitors of AMPs. The proteolytic activities of Saps (Sap1-4, Sap8 and Sap9) were reported to reduce the antifungal activity of the AMP LL-37.¹⁷³ Second, in response to AMPs, C. albicans is able to express the drug efflux pump gene Flu1, allowing the AMPs to be outsourced from the intracellular compartments of the fungus.¹⁷² Third, the fungus is able to regulate signaling pathways, such as the HOG (High-osmolarity glycerol) pathway, which are key elements in survival to various AMPs.^174,175 Knowing that the HOG pathway is involved in both the regulation of the ROS production and the production of ATP by the mitochondria,¹⁷⁴ and that ROS production and ATP efflux is induced by various AMPs,¹⁶⁰ this HOG pathway may exert a key function in the survival of C. albicans to AMPs by downregulating ROS production and ATP efflux.

It is assumed that the AMPs release in the intestinal environment, especially in the mucus layer, is conditioned by the composition of the gut microbiota that interacts with AMP-secreting cells (mainly PCs and PMNs). Consequently, dysbiotic states modify the panel and concentration of AMPs secreted in the gut lumen.¹⁷⁶ However, less is known about the direct regulation of the secretion of AMPs by C. albicans and the gut mycobiota. Interestingly, in this context, microarray-based transcriptomic analyses of intestinal epithelial cells interacting with C. albicans did not report overexpression of AMPs-related genes following infection,¹⁷⁷ whereas a strong expression of AMPs (especially DEFB4) was observed in oral cells challenged with the fungus.^177,178 Once more, these observations corroborate the view that molecular and cellular features characterizing C. albicans interactions depend on the type of host cell. However, no in-depth study has focused on the regulation of AMP secretion by C. albicans in gut-associated PCs and PMNs.

In addition to evading AMPs, C. albicans has to cross the mucus layer to access and adhere to the layer of IECs. The mucins that mostly make up this protective layer modulate the morphology and physiology of C. albicans,¹⁷⁹ as exemplified by the mucin polymer Muc5AC expressed in the lung and the stomach that downregulate the expression of a range of genes related to adherence, filamentation and biofilm formation in C. albicans.¹⁸⁰ Moreover, the mucin Muc7 has been reported to exert a direct antimicrobial effect upon C. albicans in the oral cavity.¹⁷¹ Nevertheless, the effect of mucins on colonization and dissemination of C. albicans in the gut remains to be specified.¹⁸¹

C. albicans has however developed strategies to circumvent the mucus protective layer. C. albicans is able to adhere to mucins in a pH-independent process involving hydrophobic interactions that allow the non-specific binding of C. albicans to the 66-kDa cleavage product of the 118kDa C-terminal glycoprotein backbone of mucins.¹⁸² Whereas the involvement of specific C. albicans’ ligands remains to be specified, Böhm et al. recently identified transcription regulators in C. albicans, including ZCF8, ZFU2 and TRY4, that favor (i) the yeast form of C. albicans more prone to colonize the gut in mice and (ii) adherence to mucins and mucus-producing IECs.¹⁸³ After adhering to mucins, which is critical in initiating successful colonization to host mucosal surfaces, C. albicans degrades the GI mucus layer by secreting mucinolytic enzymes. Among them, Secreted Aspartyl Protease 2 protein (Sap2p) facilitates mucus penetration by promoting pore formation in the mucus layer¹⁸⁴ using a strategy similar to that employed by enteropathogenic bacteria.^{74,182,185–188} Finally, C. albicans produces other Saps that degrade the sIgA scattered in the mucus layer, thus escaping the host innate immune response in part.¹⁸⁹

Interaction of C. albicans with enterocytes

In a susceptible host, once it has evaded the AMPs armory and the mucus layer, C. albicans adhere to IECs, cross the epithelial barrier and finally induce cellular damages leading to the loss of gut epithelium integrity.¹⁷⁹

C. albicans adherence to enterocytes

Candida sp. adherence to IEC is a key step in the physiopathological process that contributes both to the colonization and pathogenesis of disseminated candidiasis that mainly originate from the digestive mucosa.¹⁸⁹ Heterogeneities have been reported in the adherence properties of different Candida species, C. albicans displays a greater capacity to adhere to epithelial cells in vitro, including to intestinal epithelial cells.^190,191 Thus, C. albicans’ adherence to epithelial cells is a complex, dynamic and multifactorial process that requires a close relationship between the fungus and host proteins exposed at cell surfaces.¹⁹²

The capacity of C. albicans to adhere to host tissues depends on the type of epithelial cell. Indeed, C. albicans yeast forms display decreased adherence to differentiated enterocytes in vitro as compared with oral epithelial cells,¹⁰ suggesting that (i) oral cells may express specific molecules enhancing C. albicans attachment or (ii) the brush border lining the enterocyte’s layer displays unique features limiting the fungus’ ability to adhere. Following the first contact, most of the adherent yeast cells switch to the hyphal form,¹⁰ probably as a result of the physical contact with the epithelial surface.^10,193 This is an important step since the expression of hyphal-associated genes will (i) strengthen adherence through enhanced expression of adhesins^194,195 as well as (ii) favor the expression of hyphal invasins, enabling the fungus to enter the epithelial cells (see below). Another interesting observation from the early stages of interaction is the intimate entrapment of C. albicans hyphal tips within the microvilli that cover the surface of the enterocytes, a phenomenon that probably reinforces the attachment to the epithelial surface.^10,196,197 Finally, hyphal forms of the fungus conglomerated to form aggregates that adhered to the surface of enterocytes, a phenomenon that was not observed with oral cells.¹⁰ It is possible that aggregation of Candida cells indirectly promote attachment of the fungus to the surface of enterocytes.¹⁹⁸ Collectively, these observations strongly suggest that enterocytes express specific cues that influence the attachment of C. albicans. This view is supported by observations from Sohn et al. who investigated the transcriptional response of C. albicans adhering to different substrates, including vaginal and intestinal cells, and abiotic surfaces, for different periods of time.¹⁹⁷ These authors reported subsets of genes that were specifically expressed in adherent conditions as compared to suspension cultures. Furthermore, the transcriptional response of the fungus differed significantly depending on the adhesive surface (including enterocytes). The response was globally close early on, but differences became more pronounced as the interaction progressed. These observations suggest the ability of C. albicans to sense and to adapt its transcriptional program to subtle variations in the adherent surface.¹⁹⁷

In addition to passive van der Waals forces and hydrophobic interactions that probably contribute to the initial contact of C. albicans on epithelial surfaces,¹⁹⁹ expression of surface molecules (adhesins) is a key biological feature of the fungus, allowing its attachment to host tissues (both as a commensal or a pathogen),²⁰⁰ to abiotic surfaces, and to other microorganisms or Candida cells²⁰¹ (reviewed in^190,202,203). The major C. albicans adhesins are GPI-anchored glycoproteins that are expressed by yeast or hyphal forms of the fungus or both, including the multifunctional agglutinin-like sequence (Alsp) proteins family, the hyphal wall proteins (Hwp) family, the IPF family, the F/Hyphally upregulated protein or Iff/Hyr protein family, and the Epithelial adhesin proteins (Eap) family.^190,202 Furthermore, C. albicans expresses numerous other protein adhesins that have not yet been fully characterized in terms of structure and/or capacity to mediate adherence to epithelial cells such as fimbriae, Csh1, Ywp1, Pra1, proteins of the Sap family and others (Reviewed in.^{190,202–204} Though many of these adhesins have been reported to promote adherence of C. albicans to epithelial cells, few investigations have focused on enterocyte interaction. Sohn et al. observed an upregulation of HWP1 in the early stages of the C. albicans and enterocyte interaction, suggesting its involvement in the attachment of the fungus to the Caco-2 enterocytic cell line.¹⁹⁷ Similarly, Hwp2p and Int1p have been reported to contribute to adherence of the fungus to the HT-29 human intestinal epithelial cell line in vitro.^196,205 Finally, a C. albicans double mutant for the putative β-glucanase mannoprotein Mp65, defective in hypha and biofilm formation, displayed reduced adherence to plastic²⁰⁶ and epithelial cells including enterocytes.^207,208 Given the role of Mp65 in cell wall organization and stability, it is plausible that deletion of the MP65 gene affects the expression of other adhesins at the surface of C. albicans cells, indirectly altering its adherence capabilities.²⁰⁷ However, observations that blocking antibodies specific for Mp65 strongly decrease adherence of wild-type C. albicans to plastic²⁰⁶ and epithelial cells²⁰⁸ are in line with the direct contribution of Mp65 as an adhesin in C. albicans.

In addition to the protein nature of adhesins and given the high content of the yeast cell wall in carbohydrate components, C. albicans can also adhere to enterocytes through polysaccharidic molecules expressed at the cell wall surface.^204,209 Indeed, Timpel et al. reported first that a pmt1∆/∆ deletion mutant, a gene encoding a mannosyltransferase involved in the O-glycosylation pathway, displayed reduced adherence to enterocytic Caco-2 cells, suggesting a role for the O-linked carbohydrate content of the cell wall in the adherence of C. albicans to enterocytes.²¹⁰ Additional observations reported the contribution of the N-glycosidic part of the cell wall in the adherence of C. albicans to enterocytes. Indeed, in competition experiments based on a panel of carbohydrates, Dalle et al. observed a drastic dose-dependent reduction in the adherence of C. albicans to enterocytic Caco-2 cells with synthetic β-1,2 oligomannosides and to a lesser extent with α-1,2 oligomannosides, two glycans of the N-linked moiety of the cell wall.²¹¹ Unlike α-1,2 oligomannosides, β-1, 2 oligomannosides are rare structures in living systems, and their presence has been reported in only few bacterial and yeast species.^212,213 However, they are prominently expressed in C. albicans.²¹⁴ They reside in the yeast wall associated with N-glycans of a high polymer of mannoses bound to a peptide, the phosphopeptidomannan (PPM) and mannoproteins (MPs) as well as to the glycan copula of a glycolipid of the family of mannose-inositol-phosphoceramides, i.e. phospholipomannan (PLM).^214,215 They contribute to the virulence of C. albicans thanks to their immunomodulatory and adhesin properties.²¹⁴ Interestingly, the inhibitory effect obtained with the synthetic β-1, 2 oligomannosides was higher than the other tested carbohydrates including the α-1,2 oligomannosides.²¹¹ This strongly suggests that N-glycans specific to the C. albicans cell wall contribute to their attachment to enterocytes. This confirms in vivo observations that reported less C. albicans gastrointestinal colonization in infant mice orally fed with synthetic β-1, 2 oligomannosides, then in these orally fed with α-1,2 oligomannosides.^211,216

Altogether, these data support the hypothesis that proteic and carbohydrate adhesins contribute to the attachment of C. albicans to IECs, but further studies are required to better characterize the contribution of each of these adhesins. In addition, soluble factors from the extracellular matrix as well as serum proteins have been suggested as potential mediators of the indirect attachment of C. albicans to mucosal surfaces.^204,217 Similarly, specific ligands from host-cell surfaces, including proteins of the integrin and cadherin families, have reportedly facilitated the adherence of C. albicans to epithelial cells.^196,218 However, the exact nature of the host factors that influence attachment remains to be specified.

C. albicans invasion into the enterocytes

C. albicans has developed several strategies to invade and further damage host epithelial cells. The yeast-to-hyphae transition in C. albicans is pivotal for epithelial invasion,²¹⁹ contributing to two different well-documented mechanisms of invasion: (i) epithelial-driven endocytosis of hyphal forms,^10,220 and (ii) C. albicans-driven active penetration.^10,221–223 Endocytosis is considered an early event, seeing as it occurs within the first 4 hours of the interaction,²²⁴ whereas invasion at later stages seems mainly driven by active penetration.^10,217,225 In fact, these two mechanistically distinct routes of invasion probably occur differentially depending on the epithelial cell type encountered by the fungus, as a result of the temporal and/or spatial availability of the epithelial receptors that are necessary for mediating induced endocytosis.²¹⁹ For instance, when both mechanisms are present during a C. albicans invasion of oral epithelial cells, active penetration is the only process driving invasion of the enterocytes in vitro at the early stages of the interaction.¹⁰ However, recent evidence suggest that endocytosis may occur later in time and that alternate routes may allow the translocation of C. albicans through the gut barrier⁹ (Figure 4).

Figure 4. Invasion of C. albicans through the intestinal epithelial barrier.

Schematic representation of the different mechanisms used by C. albicans to translocate through the gut mucosa: (i) the transcellular route, (ii) the paracellular route, (iii) the translocation through M cells, and (iv) the alternate route that may occur. Als3, Agglutinin-Like Sequence 3; Ssa1, Heat shock protein ssa1; Hwp1, Hyphal wall protein 1; Saps, Secreted Aspartyl Proteases; CL, Candidalysin; TJs, Tight Junctions; LDH, Lactate dehydrogenase; Ca2+, Calcium.

Endocytosis of C. albicans

Endocytosis was first described for C. albicans cells interacting with in vitro models of endothelial or oral epithelial cells. The filamentous forms of the fungus were capable of triggering the formation of pseudopod-like structures at the surface of these host cells,^194,220 leading to the partial engulfment of the hyphae and its progressive internalization into host cells.^193,194,220 This host-cell mechanism is mainly driven through a Zipper-like actin filament-dependent process as demonstrated by the drastic decrease in the internalization of C. albicans cells into oral and endothelial cells pretreated with the actin polymerization inhibitors Cytochalasin D and Lancuntrulin A.^10,194,220 However, additional molecular processes contribute to endocytosis of the fungus, including macropinocytosis¹⁰ and the clathrin-dependent endocytic machinery.²²⁶ In addition, endocytosis is independent of the viability of the fungus since heat-killed hyphal forms of C. albicans are endocytosed with the same efficiency as live hyphae.^10,220 However, endocytosis of dead hyphae was not associated with cellular damage, suggesting that other(s) factor(s) produced by viable forms of the fungus may be necessary to cause cellular damage in both endothelial and oral epithelial cells.^220,227

Two fungal invasins were reported as major stimulators of the endocytic machinery.^194,228 First, the multifunctional hypha-specific surface protein Als3, a key virulence factor for C. albicans, acts as an invasin by triggering its own endocytosis by endothelial and oral human epithelial cells. N- or E-cadherin expressed at the surface of endothelial and oral epithelial cells, respectively, were identified as putative host cell receptors for this invasin.^194,229 Finally, in silico analysis reported molecular similarities between Als3/E-cadherin and E-cadherin/E-cadherin interactions, suggesting the molecular mimicry of Als3 with cadherins.¹⁹⁴ Interestingly in the context of C. albicans interaction with intestinal epithelial cells, Goyer et al. reported that Als3 participates in the E-cadherin-independent endocytosis of C. albicans into enterocytes displaying altered tight junctions (TJs) (see below for more details about the role of TJs during C. albicans invasion into enterocytes). Indeed, the specific blockade of E-cadherin by the SHE78-7 neutralizing antibody did not modify internalization of C. albicans into enterocytes displaying altered TJs.⁹ This corroborated other observations reporting a partial reduction in C. albicans endocytosis into oral epithelial cells when inhibiting E-cadherin function,²³⁰ suggesting that other(s) host-cell receptor(s) may contribute to the Als3-driven endocytosis of the fungus. A second invasin, Ssa1, expressed at the surface of C. albicans hyphae and belonging to the Hsp70 family of heat shock proteins, was found to contribute to cadherins-dependent endocytosis of the fungus into endothelial and oral epithelial cells.²²⁸ Indeed, C. albicans ssa1∆/∆ exhibited reduced endocytosis into endothelial and oral epithelial cells, that was associated with less cellular damage, a decrease in adherence to endothelial and oral epithelial cells and attenuated virulence in a mouse model of oropharyngeal candidiasis.²²⁸ Surprisingly, the C. albicans double mutant ssa1∆/∆als3∆/∆ displayed reduced endocytosis, similar to the single mutant als3∆/∆, suggesting (i) the possible binding of Als3 and Ssa1 to the same host receptor and/or (ii) a cooperative function of both Als3 and Ssa1 acting as a multiprotein complex.²²⁸ Finally, other invasins probably contribute directly or indirectly to the endocytosis of C. albicans,^217,230 as highlighted by Wächtler et al., who observed reduced endocytosis of killed hyphae of the C. albicans sap1-3∆/∆ and sap4-6∆/∆ deletion mutants into oral cells.²¹⁷ However, the exact nature of the Saps as well as their direct or indirect involvement in the endocytic mechanism is not yet clear.

Collectively, these observations strongly suggest that endocytosis of C. albicans into epithelial cells is a complex mechanism that involves C. albicans hyphal-invasins including Als3, Ssa1 but possibly other invasins that remain to be specified. The interaction of such invasins with host-receptors, including cadherins but also probably other receptors, is necessary for triggering this process.²¹⁷ Zhu et al. highlighted that the involvement of Epithelial Growth Factor Receptor (EGFR) and HER2 as cadherin-independent host receptors during the endocytosis of C. albicans into oral epithelial cells.²³⁰ However, the exact nature and function of such invasins and host receptors in the endocytosis of C. albicans into enterocytes displaying altered TJs remains to be specified.

Active penetration

Active penetration is a fungal-driven process and requires viable forms of the fungus. Similar to endocytosis, active penetration exploits the yeast-to-hyphae transition of C. albicans, which germinate and produce hyphae that progressively elongate and penetrate the epithelial cells.^10,219 This invasion mechanism is of prime interest since it is the only one that allows C. albicans to gain entry to enterocytes with intact intercellular junctions in vitro.^9–11 Scanning Electron Microscopy (SEM) observations of Candida-infected enterocytes did not reveal pseudopod-like structures surrounding C. albicans hyphae, which is a typical feature of endocytosis. By contrast, a depression in the cell surface was observed around the C. albicans entry site, probably as the result of an active process of the extending hyphae forcing entry either directly into the apical surface or at intercellular junctions between adjacent enterocytes. These observations were supported by the fact that pretreatment with endocytosis inhibitors does not inhibit the uptake of the fungus into enterocytic Caco-2 cells.¹⁰

The contribution of fungal hydrolases to this process has been investigated but remains unclear.²³¹ Notably in the gut barrier, Saps contribute to the active penetration of C. albicans. The C. albicans deletion mutants sap1-3∆/∆ or sap4-6∆/∆²¹⁷ displayed moderate but significant reduction in their ability to invade enterocytes. In addition, treating the fungus with the aspartic protease inhibitor pepstatin A reduced enterocyte invasion in a dose-dependent manner.¹⁰ The role of other hydrolases, including lipases and phospholipases B1, did not obviously contributing to active penetration of C. albicans into enterocytes.¹⁰ Altogether these results suggest that active penetration of C. albicans into intestinal cells is likely to occur through molecular mechanisms combining the mechanical pressure exerted by elongating hyphae and the lytic activities of Saps.

Interestingly, Als3, a major invasin that contributes to induced endocytosis, reportedly facilitates the active penetration of C. albicans into oral epithelial cells,¹⁹⁴ but probably through an indirect mechanism that favors the binding of the two structures.²¹⁹ Als3 also seems to play a role in invasion since a C. albicans deletion mutant als3Δ/Δ, which is not defective in hyphal formation, was less able to invade intact enterocytes,⁹ which are not able to internalize the fungus through endocytosis.¹⁰

Interestingly, in an attempt to decipher the molecular basis of the interaction of C. albicans to epithelial cells, Wächtler et al. investigated the abilities to adhere to, invade and cause damage to oral and intestinal epithelial cells, of a set of 26 mutants deleted for genes predicted to be and/or previously reported as major contributors to host-C. albicans interaction. Whereas EFG1 and its upstream regulators appeared essential for all stages of the infection process, other genes displayed discrete but significant functions in specific stages of the interaction. Nine genes were identified as “damage-associated” since the corresponding deletion mutants were still able to invade oral epithelial cells but without damaging them. Surprisingly, six of these mutants (i.e. gpd2∆, gpp1∆, cka2∆, bcr1∆/∆, hwp1∆, bud2∆ and rsr1∆) displayed a strong and significant decrease in invasion into enterocytes.²¹⁹ This suggests that the genes involved in glycerol accumulation (GPP1 and GPD2) and directed hyphal growth (BUD2 and RSR1) are necessary for the active penetration of C. albicans into enterocytes but not into oral cells.²¹⁹

Together, these observations strongly support the view that C. albicans requires some specific genes to invade enterocytes but not oral epithelial cells,²¹⁹ reflecting the ability to adapt to specific cell environments by expressing biological factors that can contribute differentially to specific stages of the infection process.

The paracellular route

Several studies have suggested a paracellular route involving the proteolytic cleavage of the intercellular adherens junction (AJ) protein E-cadherin in both oral epithelial cells^225,232 and IECs.²³³ AJs are located immediately below the TJs, and are mainly composed of proteins of the cadherin family including E-cadherin. AJs together with TJs form a junction belt that ensures the architectural cohesion of the intestinal epithelium.^234,235

While active penetration has been described as the predominant mechanism of Candida invasion into IECs, in 2016 Goyer et al. observed that endocytosis of C. albicans can also occur during interaction with IECs in vitro. They demonstrated that TJs play a protective role in the early steps of the interaction by limiting endocytosis-mediated C. albicans invasion into enterocytes. While intact IECs are mainly invaded through active penetration, pharmacologically altered TJs are invaded by active penetration and induced endocytosis, through molecular mechanisms involving, at least in part, the hyphal-associated invasin Als3. So, in addition to demonstrating the role of TJs in limiting the invasion of IECs by C. albicans, this study highlights the putative harmful effects of extrinsic factors that alter the integrity of TJs, thereby allowing C. albicans to invade through the gut barrier.⁹ This corroborates reports of clinical situations where therapies or medical practices may alter the permeability of the digestive tract, consequently promoting the translocation and hematogenous dissemination of common residents of the gut microbiota, including C. albicans.^236,237 Interestingly, recent findings indicate that C. albicans itself can alter the integrity of TJs. Using an in vitro model of the intestinal barrier, Böhringer et al. observed a barrier breakdown after 8 h of infection, as illustrated by the drastic decrease in the Trans-Epithelial Electrical Resistance (TEER) of C2BBe1 (a clone of Caco-2 cells) intestinal epithelial cells interacting with C. albicans. This breakdown was not related to cell death since C2BBe1 cells were not damaged at this point in time. In addition, the drop in TEER required the fungal yeast-to-hyphae transition since the efg1∆/cph1∆ deletion mutant, unable to produce hyphae, did not induce barrier breakdown or cell death. Along with barrier breakdown, a strong decrease in the level of TJs proteins, including occludin, JAM-A, Claudin −1, −3 and −4, was observed by Western-blotting on whole infected-cell extracts, suggesting that the barrier breakdown results from the loss of these proteins. This was confirmed by immunofluorescence testing which found a progressive extinction of Occludin, Claudin-1 and 4 and JAM-A staining after 10 h of infection. All in all, this study showed that a barrier breakdown occurs during infection by C. albicans and this is related to a decrease in TJs protein levels and their concomitant disappearance from cell-cell borders.¹⁷⁷ Recently, Allert et al. observed the ability of the ece1∆/∆ mutant to compromise epithelium integrity and to translocate across the intestinal epithelial barrier through damage-independent factors and without damaging the cells.¹⁷⁹ The authors hypothesized that these factors probably alter the cell-cell connections (i.e. TJs or adherens junctions), facilitating paracellular translocation. Focusing on the effect of Saps on the promotion of the degradation of the E-cadherin protein during C. albicans infection, they studied mutants lacking various SAP genes (i.e. sap1-3∆/∆, sap5∆/∆, sap4-6∆/∆ and sap9/10∆/∆). Unfortunately, no translocation defect was observed with the mutants. However, the reduced ability of the sap1-3∆/∆ mutant to translocate through the blank transwell insert without cells cannot rule out the possible minor role of SAPs in the paracellular route of translocation.¹⁷⁹

Collectively, these are important observations that point to the paracellular route as a possible mechanism used by C. albicans to translocate through the intestinal barrier. This mechanism may favor the passage of C. albicans cells between enterocytes without invading or damaging intestinal cells.^10,193 However, it is clear now that easier access to the basolateral site of intestinal cells promotes fungal invasion through both active penetration and induced endocytosis.⁹ Finally, C. albicans itself is capable of altering the integrity of TJs, secondarily promoting its own translocation.¹⁷⁷ However, the extent to which the paracellular route promotes intercellular transmigration of C. albicans (associated or not to epithelial invasion and damage) remains to be specified.

C. albicans-induced epithelial cell death and mucosal damage

Late stages of the infection of intestinal cells are characterized by a loss of barrier integrity associated with the destruction of IECs invaded by C. albicans (Figure 4).²⁰³ It is possible that the barrier breakdown observed in the late stages, as a consequence of alterations in intercellular junctions,^177,233 triggers host-cell signaling pathways that promote inflammation, the activation of inflammasomes and apoptotic events, ultimately leading to necrosis of the enterocytes.⁴³ Invasion by C. albicans also actively contributes to enterocyte damage and cell death.²⁰³ Until recently, it was generally believed that Candida invasion resulted in the death of infected epithelial cells possibly through mechanisms involving necrosis and/or apoptosis.²³¹ However, the invasion of epithelial cells per se is not always clearly associated with cell damage.¹⁷⁹ In fact, initial invasion of oral, vaginal or intestinal epithelial layers causes little measurable cell death.^10,193 Conversely, the later phases of C. albicans invasion into IECs are associated with epithelium damage mainly due to the mechanical elongation of the hyphae.²¹⁹ However, the identification of a specific virulence factor inducing cell damage has remained elusive. A cytolytic toxin produced by C. albicans, named Candidalysin (CL), was recently identified; it is thought to be the missing link between hyphae formation and damage to host cells.^179,238 CL is a 31-amino-acid peptide generated from the pre-protein Ece1 (Extent of cell elongation 1) (271 amino acids), which is encoded by the ECE1 gene belonging to the eight core filamentation genes of C. albicans and highly expressed during hyphae formation.²³⁹ Previous studies had shown that the seven lysine-arginine dibasic amino acid motifs present in the Ece1 pre-protein can be recognized and cleaved by a subtilisin-like serine protease, Kex2, in the yeast Golgi apparatus.²⁴⁰ CL is one of the cleaved peptides generated from Ece1, and its α-helical structure allows it to intercalate and permeabilize the host cell membranes and subsequent cell lysis when secreted in sufficient amount. In oral epithelial cells, CL triggers a danger-response signalling pathway and activates epithelial immunity.²³² Moreover, the permeability of the cell membrane is enhanced by a positive charge at the carboxy terminus of the peptide which triggers an inward current concomitant with calcium influx.²³² Recent observations have confirmed the importance of CL during the translocation of C. albicans through the intestinal barrier in vitro. Allert et al. screened more than 2000 C. albicans gene deletion mutants to identify fungal factors involved in C. albicans gut translocation, leading to the identification of several genes associated with cellular damage including PRN4, orf19.2797, NRP2, AAF1, HMA1, TEA1, orf19.3335, PEP12 and ECE1.¹⁷⁹ Focusing on these nine corresponding mutants, eight of them, including ece1∆/∆, showed a significant decrease in their ability to induce damage at 24 h post-infection. Given the critical role of CL for oral and vaginal epithelial cell damage, the authors investigated the role of CL during C. albicans gut translocation using a C. albicans mutant that lacked only the CL-encoding region within ECE1 (ece1 ∆/∆ + ECE1_∆184–279 mutant). They observed that both ece1 ∆/∆ and ece1 ∆/∆ + ECE1_∆184–279 mutants were unable to induce damage and that their ability to induce damage was partially restored with the administration of synthetic CL. This demonstrated that a combination of both (i) hyphae formation and (ii) CL secretion were required for optimal damage induction.¹⁷⁹ Necrotic cell death seems to be the major mechanism involved in damage observed during C. albicans translocation through enterocytes. Indeed, 24 h post-infection, no induced apoptosis (i.e. Annexin V staining) was observed while 40% of IECs were necrotic.¹⁷⁹ In addition, damage was associated with a loss of epithelial integrity during the translocation of C. albicans through the intestinal epithelial layer.¹⁷⁹ These studies highlight the key role of CL in Candida pathogenesis and provide a molecular explanation for the destructive activity of hyphae on the mucosal surface.²³⁹

Interaction of C. albicans with M cells

As mentioned above, several bacterial pathogens exploit M cells to invade mucosal tissues and cross the digestive epithelial barrier to reach the bloodstream.²⁴¹ This suggests that entry through M cells is an essential step in the pathophysiology of many infectious diseases. In a recent in vitro study, Albac et al. demonstrated that C. albicans co-localizes with and preferentially invades M cells in a model of the enterocytic Caco-2 cells co-cultured with M-like cells, providing novel evidence that the fungus can use M cells as a portal of entry through the intestinal barrier. In addition to active penetration, F-actin-dependent endocytosis contributed to internalization of the fungus into M cells through a mechanism involving hypha-associated invasins such as Ssa1 and Als3.¹¹ Moreover, fungal β-glucans can be endocytosed by M cells which, by virtue of their APC function, expose β-glucans to macrophages and dendritic cells associated with Peyer’s patches, thereby modulating the systemic immune response to C. albicans.²³⁶ Altogether these recent findings strongly support the view that understanding the contribution of M cells in sensing C. albicans as a commensal or a pathogen should provide new insights to the commensal role of C. albicans as well as to the pathophysiology and initiation of life-threatening systemic candidiasis. The exact molecular mechanisms by which C. albicans targets and translocates through M cells as well as the relevance of these interactions in vivo remain to be elucidated.

Models to study interaction of C. albicans with the intestinal mucosa

For now, several models have been already used to study the interaction between C. albicans and the epithelial intestinal barrier. Surprisingly, when the gut barrier forms a complex environment involving several cell types, each with specialized functions, the in vitro models already described are mostly based on the use of only one cell type: enterocytes. Only one model investigated the interaction of C. albicans with two co-cultured cell types: enterocytes and M-like cells. However, new models are aimed at specifying the interaction of the fungus with AMPs, mucus and immune cells in the gut environment, as well as with other specialized cells including PCs, and Goblets cells. All of the existing models (i.e. in vivo and in vitro) are summarized in Table 1 and Table 2.

Table 1. In vitro models developed to study the interactions between Candida albicans and the epithelial intestinal barrier.

Cell lines	Experimental procedures	Features	References
In vitro models to study the interaction between Candida albicans and IECs
Colon adenocarcinoma derived cell line Caco‐2 (ATCC® HTB-37™)	-Routinely cultured in DMEM medium supplemented with 10% FBS, pyruvic acid, l‐glutamine and nonessential amino acids without antibiotics or antifungal agents in a humidified incubator at 37°C in 5% CO2 2- – Seeded onto 12mm diameter glass coverslips previously placed in 24 well plates or in polycarbonate filter inserts. Differentiation obtained 15 days post seeding	Monolayers displayed several morphological and functional characteristics of mature enterocytes	^Citation10,Citation233
C2BBe1	-Routinely cultured in DMEM/Ham’s F12 medium supplemented with 0.1 Vol. FBS, glucose, stable glutamine and proxyferrin in a humidified incubator at 37°C in 10% CO2 -Seeded in Transwell polycarbonate inserts coated with Matrigel. Differentiation obtained 12 days post seeding	Polarized monolayers (Caco-2 subclone)	^Citation177
In vitro models to study the interaction between Candida albicans and M cells
Human enterocytes Caco-2/Isolated murine B Lymphocytes from Peyer‘s patches	Caco-2 culture in insert for 14 days then co-culture with isolated murine B Lymphocyte from Peyer‘s patches. M cell differentiation obtained 6 days post LB initiation.	Heterologous co-culture	^Citation241
Human enterocytes Caco-2/B Lymphocytes Raji (ATCC® CCL-86™) (LB Raji)	Caco-2 culture in insert for 14 days then co-culture with LB Raji. M cell differentiation obtained 6 days? days post LB initiation	Homologous co-culture	^Citation242
Human enterocytes Caco-2/LB Raji	Caco-2 culture in insert for 5 days, then insert culture returned for 10 days then co-culture with LB Raji. M cell differentiation obtained 6 days post LB initiation	Homologous co-culture	^Citation11,Citation243

Table 2. In vivo models described to study the interaction between Candida albicans and the epithelial intestinal barrier Murine models of intravenous disseminated candidiasis.

Mice strains (sex)	Experimental procedures	Features	References
DBA/2N (M or F)	- Immunocompetent or immunocompromised mice - Infected intravenously with C. albicans via the tail vein	Well characterized Reproducible Rapid spread	^Citation244–Citation247
CFW (M)
CD-1 (F)
CBA/H (us)
BALB/c (M or F)
Murine models of disseminated candidiasis originating from the gastrointestinal tract
Crl:CD1 (ICR)BR (M)	Immunocompetent mice (± antibiotic therapy) - Diet supplemented with yeasts administrated (i) in food or (ii) in drinking water or (iii) by gavage - Gut colonization argued by stool culture - Gut mucosa breakdown induced by (i) hypo-protein diet or (ii) administration of DSS or (iii) administration of immunosuppressive treatments (Cyclophosphamid, succinate methylprednisolone sodium, cyclosporine, 5-Fluoroouracil)	Mimic the yeast physiopathology in Human Blood dissemination delayed 14 to 21 days after the induction of the barrier rupture	^Citation14,Citation245,Citation248–Citation253
C3H/HeJ (M)
BALB/c (M or F)
CD-1 (F)
C3H/HeN (F)
CBA/H (us)

Conclusions

Formerly described as a commensal that exists on the mucosal surfaces of most healthy individuals, Candida albicans is an opportunistic pathogen that causes infections ranging from superficial to life-threatening disseminated infections, especially in the ever-growing population of vulnerable patients in hospital settings. In these situations, the fungus takes advantage of its host thanks to disturbances in the host-defense system and the mucosal microbiota. Overwhelming evidence suggests that the gastrointestinal tract is the main source of disseminated C. albicans infections. Some of the major risk factors for disseminated candidiasis include damage to the mucosal intestinal barrier, immune dysfunction, and antibiotic-induced dysbiosis of the resident microbiota.

Clearly, translocation of C. albicans through the intestinal barrier into the bloodstream is a critical step in the initiation of invasive candidiasis, which makes it important to understand the mechanisms by which C. albicans interacts with the gut barrier. Given the specific cues shared by intestinal cells and the specific environment of the gut lumen, the bulk of knowledge acquired when studying interaction of C. albicans with other mucosa, including the oral mucosa, is not fully transferable to the behavior of the fungus in the gut context.

A more precise view has emerged recently on how C. albicans translocates through the gut epithelial layer. However, the nature of the fungal and host factors contributing to the translocation remain to be specified. In addition, many questions are still unsolved and remain to be explored. For instance, a lack of knowledge exists with regard to the specific molecular mechanisms allowing IECs to prevent fungal invasion and how C. albicans can hijack them. The mechanisms by which C. albicans interacts with specialized IECs such as PCs and goblet cells and their effects on gut homeostasis and/or the pathogenesis of candidiasis is still unknown. Similarly, the extent to which the maintenance of inter-cellular junctions is crucial to restraining C. albicans invasion and how the fungus alters these junctions is still ongoing. In addition, little is known about how the biochemical components of the gut environment modulate the fungal factors and how C. albicans resists these biochemical factors. Finally, from an immune point of view, the specific mechanisms and pathways involved during host response associated to commensalism or pathogenicity of C. albicans need to be precise in the gut reservoir. For instance, though the contribution of the M cells in the translocation of C. albicans is now obvious, the consequence of this mechanism on the priming of tolerant or protective immune response is still unknown. A better overall understanding of the interaction between C. albicans and the intestinal epithelial barrier will help with the development of future therapies to avoid systemic candidiasis.

Importantly, with the advance of molecular tools to study microbiota components, new insights have emerged with regard to the association of GM dysbiosis with gastrointestinal or extra-gastrointestinal disorders where C. albicans colonization is disturbed. When highlighting the intricacy of bacterial-fungal relationships in the GM,²⁵⁴ the consequences of these complex interactions upon the host and C. albicans virulence needs to be clarified.¹⁵⁸ Such information will help with the creation of new strategies to combat candidiasis, taking advantage of the existing interactions between bacteria and fungi in the gut microbiota. Co-infection models that closely mimic the interactions of the digestive microbiota and IECs are needed if we are to fully understand the pathophysiological mechanisms of candidiasis.

Abbreviation

Adenosin Triphosphate	=	ATP
Adherens Junctions	=	AJs
Agglutinin-Like Sequence	=	ALS
Anti-Microbial Proteins	=	AMPs
B Lymphocytes	=	LB
Candidalysin	=	CL
c-Jun N-terminal Kinase	=	JNK
Dendritic Cells	=	DCs
Epithelial Adhesin Protein	=	EAP
Epithelial Growth Factor Receptor	=	EGFR
E-twenty-six-specific sequence	=	Ets
Extent of cell elongation 1	=	Ece1
Extracellular-signal-Regulated Kinase	=	ERK
Follicle-Associated Epithelium	=	FAE
GastroIntestinal	=	GI
Gastrointestinally-IndUced Transition	=	GUT
Glyco-Protein 2	=	GP2
Gut-Associated Lymphoid Tissue	=	GALT
Gut-associated Microbiota	=	GM
Heat Shock Protein	=	HSP
High-osmolarity glycerol	=	HOG
Hyphal wall protein	=	Hwp
Hypoxia Inducible Factor	=	HIF
Inflammatory Bowel Disease	=	IBD
Intestinal Epithelial Cells	=	IECs
Intestinal Epithelium	=	IE
Isolated Lymphoid Follicles	=	ILF
Macrophages	=	Mφ
Mannoproteins	=	MPs
Mesenteric Lymph Nodes	=	MLNs
Microfold cells	=	M cells
Mitogen-Activated Protein Kinases	=	MAPK
Mucosal Immune System	=	MIS
Neutrophil cells	=	PMNs
Next Generation Sequencing	=	NGS
Nuclear Factor kappa B	=	NFkB
Paneth Cells	=	PCs
Pathogen Associated Molecular Patterns	=	PAMPs
Pattern Recognition Receptors	=	PRRs
Peyers’ Patches	=	PPs
PhosphoLipoMannan	=	PLM
PhosphoPeptidoMannan	=	PPM
Reactive Oxygen Species	=	ROS
Regenerating Islet-Derived 3 Gamma	=	REG3γ
Scanning Electron Microscopy	=	SEM
Secreted Aspartyl Proteases	=	SAPs
Secretory group IIA PhosphoLipase A2	=	sPLA2
Secretory Immunoglobulin A	=	sIgA
Short-Chain Fatty Acids	=	SCFAs
Sub-Epithelial Domes	=	SED
T Lymphocytes	=	LT
Tight Junctions	=	TJs
Trans Epithelial Electrical Resistance	=	TEER
Zonula Occludens	=	ZO

Disclosure of interest

The authors report no conflict of interest

Additional information

Funding

This work was supported by the Regional Council of Bourgogne—Franche Comté under Grant and the Fonds Européen de DEveloppement Régional (FEDER) under Grant; Conseil régional de Bourgogne-Franche-Comté; FEDER.