ERG3 and ERG11 genes are critical for the pathogenesis of Candida albicans during the oral mucosal infection

The hyphal development of Candida albicans (C. albicans) has been considered as an essential virulent factor for host cell damage. However, the missing link between hyphae and virulence of C. albicans is also been discovered. Here, we identified that the null mutants of ERG3 and ERG11, two key genes in ergosterol biosynthesis pathway, can form typical hyphae but failed to cause the oral mucosal infection in vitro and in vivo for the first time. In particular, the erg3Δ/Δ and erg11Δ/Δ strains co-cultured with epithelial cells significantly reduced the adhesion, damage, and cytokine (interleukin-1α (IL-1α)) production, whereas the invasion was not affected in vitro. Importantly, they were incapable of extensive hyphal invasion, formation of micro-abscesses, and tongue epithelium damage compared to wild type due to the decrease of the colonization and epithelial infection area in a murine oropharyngeal candidiasis model. The fluconazole (FLC), an antifungal targeted at ergosterol biosynthesis, relieved the epithelial infection of C. albicansin vitro and in vivo even under non-growth inhibitory dosage confirming the virulent contribution of ergosterol biosynthesis pathway. The erg3Δ/Δ and erg11Δ/Δ strains were cleared by macrophages similar to wild type, whereas their virulence factors including agglutinin-like sequence 1 (Als1), secreted aspartyl proteinase 6 (Sap6), and hyphal wall protein-1 (Hwp1) were significantly reduced indicated that the non-toxicity might not result from the change on immune tolerance but the defective virulence. The incapacity of erg3Δ/Δ and erg11Δ/Δ in epithelial infection highlights the contribution of ergosterol biosynthesis pathway to C. albicans pathogenesis and fluconazole can not only eliminate the fungal pathogens but also reduced their virulence even at low dosage. The damage from oral infection with the fungus Candida albicans can be contained by targeting two cell membrane-building genes. C. albicans cells transition from a rounded shape into long filamentous structures called hyphae prior to invading and damaging host epithelial cells. Researchers led by Lei Cheng at Sichuan University have now identified a key intermediate step between hyphae formation and virulence. They determined that fungal cells lacking either of two genes that manufacture ergosterol, a component of the C. albicans membrane, still form hyphae and attach to epithelial cells. However, these mutant fungi inflict no cellular damage, and did not cause disease in mice. Furthermore, treatment with low-dose fluconazole, a drug that inhibits ergosterol synthesis, rendered the fungus non-virulent without killing it, indicating that this pathway represents an important ‘missing link’ for fungal pathogenesis.


INTRODUCTION
Oral candidiasis, a worldwide medical challenge for fungal superficial infection, is responsible for the high morbidity especially in children, denture wearers and the immunocompromised population, such as human immunodeficiency virus (HIV) infected patients and head/neck cancer patients received radiation or chemo therapy. [1][2][3][4] Candida albicans (C. albicans) is the most pronounced conditional fungal pathogen colonized in oral cavity. 5 The filamentous growth of C. albicans is considered as the most essential virulence factor for the adhesion and invasion. 6,7 C. albicans can also produce many virulent molecules companied with the hyphal development, such as the cell-surface adhesin and secreted aspartyl proteases (Sap). 8,9 Agglutinin-like sequence (ALS) genes encoded cell-surface glycoproteins have important roles in the adherence to host surfaces, 10 such as agglutinin-like sequence 1 (Als1), which is capable of the inducing adherence to endothelial and epithelial cells. 11,12 Although the major protein of the hyphal cell wall hyphal wall protein-1 (Hwp1) also functions as a cell-surface adhesion with the ability to mimic mammalian transglutaminase substrates for the formation of covalent crosslinking between C. albicans and epithelial cells. 13 The family of Sap of C. albicans is responsible for the adhesion, cell-surface integrity, and tissue damage. 7,14,15 SAP6 is the predominant protease gene expressed in the patients with oral candidiasis and the expression occurs concomitantly at the place of tissue damage. 16 The epithelium is thought to be the first mechanical barrier against tissue invading by C. albicans. When the epithelial cells are infected by the C. albicans hyphae, they activate the activating protein-1 (AP-1), c-Fos, and mitogen-activated protein kinase 1 (MKP1) to sense the C. albicans hyphal damage and produce the epithelial cytokine (such as interleukin(IL)-1α, IL-1β, IL-6, and IL-17), and then recruit immune cells (such as macrophages). 17,18 However, it remains unclear that which cell components of C. albicans hyphae are important for mediating the damage of epithelial cells. Recently, the first fungal cytolytic peptide toxin "Candidalysin" (encoded by ECE1) was identified in C. albicans. 19 The ECE1 deleted mutant can form normal hyphae similar to the wild type strain but not cause the epithelial cell damage, suggesting that candidalysin is a critical factor for the potential of C. albicans hyphae to cause invasive mucosal infections and tissue damage without the impact upon filamentous growth. The morphological identity between ECE1 deletion and wild type strains combined the opposite capabilities on epithelial cell damage highlight the idea that there are "missing links" between hyphal growth and host cell damage. This type of "missing link" genes will provide further insight into the transformation process from commensal to pathogenic state of C. albicans, and perhaps additional therapeutic targets.
To combat with C. albicans infections, several types of antifungal drugs are developed, such as azoles targeted at ergosterol (key element in cell membrane) biosynthesis, 20 polyenes binding to ergosterol to form poles in cell membrane, 21 and echinocandins targeted at cell wall biosynthesis. [22][23][24] Fluconazole (FLC), a clinical first-line fungistatic antifungal azole, can bind to Erg11 to inhibit the ergosterol biosynthesis and cause the accumulation of toxic sterols, indicating the importance of ergosterol in C. albicans. [25][26][27] ERG3 and ERG11 are the most important genes in ergosterol biosynthesis pathway and they have key roles in azole drug resistance. [28][29][30] However, their contributions to oral epithelial infections are not under investigated. Here we identified that the ERG3 and ERG11 genes were also belonged to the "missing link" type of genes for the first time since their deletions were incapable of causing oral mucosal infection similar to ECE1 gene, but they can also form hyphae. Meanwhile, fluconazole can relieve the epithelial infection even at non-growth inhibitory dosage both in vitro and in vivo, indicating its dual-functional abilities to not only eliminate the C. albicans but also inhibit the interaction between fungal pathogens and host cells by reducing the infective virulence.

RESULT
ERG3 and ERG11 genes are critical for epithelial cell damage in vitro The expression of both ERG3 and ERG11 genes were significantly upregulated when C. albicans strains co-cultured with epithelial cell, indicating the positive relationship between ERG3 and ERG11 and the epithelial pathogenesis ( Figure S1a, b). Then we subjected wild type, erg3Δ/Δ and erg11Δ/Δ to epithelial cell culture to probe the functions of ERG3 and ERG11 genes during epithelium infection in vitro. The erg3Δ/Δ and erg11Δ/Δ strains both can form typical hyphae identical with wild type (Fig. 1a), but they were incapable of inducing epithelial cell damage (Fig. 1b) after co-cultured with epithelial cell for 24 h compared to wild type, indicating that erg3Δ/Δ and erg11Δ/Δ only formed non-virulent hyphae. Meanwhile, the erg3Δ/Δ and erg11Δ/Δ strains significantly reduced the adhesion to the epithelial cells compared to wild type (Fig. 1c). Interestingly, both erg3Δ/Δ and erg11Δ/Δ strains were capable of extensive epithelial invasion and penetrating through multiple epithelial cells same as the wild type after 24 h cocultured with epithelial cells, in line with the morphological similarity of hyphae between the mutants and wild type strain (Fig. 1d, e). Although the invasion was not affected, both erg3Δ/Δ and erg11Δ/Δ strains significantly reduced cell damage and inflammatory through the decrease of the reactive oxygen species (ROS) (Fig. 1f) and cytokine (IL-1α) production ( Fig. 1g) in epithelial cells compared to the wild type strain. To identify the reason for the non-virulent hyphae of erg3Δ/Δ and erg11Δ/Δ, we measured the expressions of some important virulence factors. The expressions of ALS1 and SAP6 tested in this study were significantly downregulated in erg3Δ/Δ strain as well as the obvious downregulation of SAP6 and HWP1 in erg11Δ/Δ strain (Fig. 1h), respectively, consistent with their incapability of causing the cell damage and induction of ROS and cytokine, indicating the critical roles of ERG3 and ERG11 for epithelial infection in vitro.
FLC protect the epithelial cells from the damage of C. albicans in vitro As azole drugs can target at the ergosterol biosynthesis pathway, 22 we employed FLC as an inhibitor to confirm the critical roles of ERG3 and ERG11 of C. albicans during the epithelial cell infection in vitro. The growth of C. albicans was inhibited by FLC at the concentration of 1 µg·mL −1 ( Figure S2a), so we tried low doses of FLC, at which the growth was not inhibited ( Figure S2b), to evaluate its inhibitory ability on the interactions between C. albicans and epithelial cell. Interestingly, C. albicans wild type stains was significantly reduced the epithelial cell damage at both 0.25 and 0.125 µg·mL −1 (Fig. 2a). FLC also significantly suppressed the inductions of ROS and IL-1α productions (Fig. 2b, c). Meanwhile, the non-growth inhibitory concentration could remarkably reduce the adherence rate to the epithelial cells (Fig. 2d). In combination, these results indicate that FLC can not only work as fungal eliminative drug, but also be served as virulence inhibitor when at non-growth dosage due to the critical functions of ergosterol biosynthesis pathway in epithelial cell infection.
ERG3 and ERG11 genes are essential for mucosal pathogenesis in vivo We next assessed the role of ERG3 and ERG11 in murine oropharyngeal candidiasis model. The ERG3 and ERG11 null mutants were considered as non-virulent hyphae, whereas the wild type served as normal-virulent hyphae control. As expected, the mice infected with wild type strains exhibited typical hyperplastic white plaques on the lingual surface, whereas the erg3Δ/Δ and erg11Δ/Δ strains failed to form the lesions (Fig. 3a). Quantification of histology sections including micro-abscesses, the extensive hyphal invasion of the tongue epithelium and tissue damage indicated the typical tongue epithelial infectious disease symptoms when the mice infected with the wild type. In contrast, mice infected with erg3Δ/Δ and erg11Δ/Δ strains showed no obvious invasive fungal hyphae and no inflammatory infiltrates or tissue damage ( Fig. 3b-d). The incapability of erg3Δ/Δ and erg11Δ/Δ strains to cause mucosal infection in mice was not due to the intolerance to phagocyte challenge as both the null mutants and wild type strains were cleared by macrophage as the same (Fig. 3e), but may mainly related the decrease of virulence factors in erg3Δ/Δ and erg11Δ/Δ strains (Fig. 1h). Therefore, ERG3 and ERG11 genes in C. albicans were critical for mucosal infection in vivo.
FLC can cure the epithelial infection caused by C. albicans at low dosage in vivo The powerful efficacy of FLC in epithelial cell infection in vitro highlighted the expectation of its potential effects in vivo at low dosage. As expected, after the mice infected by C. albicans, the FLC-treated group demonstrated absent white patches and low fungal burdens on the tongues at both 0.25 and 0.125 µg·mL −1 compared to the no drug treatment group (Fig. 4a, b). These dosages of FLC also significantly reduced the epithelium infection area, inflammatory infiltrates, and local epithelial damage (Fig. 4c, d), indicating the curative efficacy of FLC even at non-growth inhibitory dosage and suggesting the ability of FLC to block the interactions between fungal pathogens and host. The macrophage clearance rate of C. albicans was not effected by FLC ( Fig. 4e) in line with the wild type strain, indicated the inhibition of FLC on C. albicans virulence.

DISCUSSION
Ergosterol, the most important component in fungal cell membrane, functions many of the same as cholesterol in animal cells to regulate the fluidity and biogenesis of plasma membrane. 31,32 Here we identified the functions of ergosterol biosynthesis pathway contributed to the oral epithelial infection caused by C. albicans for the first time. The ergosterol biosynthesis dysfunction mutants erg3Δ/Δ and erg11Δ/Δ failed to damage the oral epithelial cells in vitro and importantly they lost the function to cause the mucosal infection in vivo. Miyazaki et al. 33 observed that the ERG3 null mutants showed defective hyphal formation when induced by human serum in vitro and in systemic infected mice. However, when the erg3Δ/Δ strain co-cultured with the epithelial cells in vitro, it can form the typical hyphae and show the same invasion rate similar to the wild type strain in our study, indicated that the attachment between C. albicans and epithelial cells may regulate the morphological development of C. albicans. As the typical hyphae of erg3Δ/Δ mutant were observed in vitro, there were no extensive fungal invasion in murine oropharyngeal candidiasis model in line with the observation in systemic C. albicans infection mice, indicated the defective virulence of ERG3 deletion. Becker et al. 34 found that ERG3 and ERG11 were required for virulence in murine model of systemic infection. Our results corroborated the finding that ERG3 and ERG11 were critical for the virulence in murine oropharyngeal candidiasis model. Furthermore, we dominated that the reason for that may result from the decrease of key virulence factors such as Als1, Hwp1 and Sap6 in ERG3 and ERG11 null mutants instead of their immune intolerance to macrophage. The production of ROS from immune cells, such as macrophage, is an important immune weapon to wipe out C. albicans during the early infectious stage. We found that the ERG3 and ERG11 gene from C. albicans contributed to the ROS production in oral epithelial cells, which may contribute to the C. albicans-infected epithelial cell damage. The dysfunction of ERG11 gene was reported as more sensitive to ROS generated from neutrophils, 35,36 which may also one of the reasons for the failure of erg11Δ/Δ mucosal infection in vivo in our study. In combination of the previous findings in C. albicans systemic infection murine model 37,38 and our results in murine oropharyngeal candidiasis model, ergosterol biosynthesis pathway is proved to be essential for C. albicans pathogenesis both in invasive and superficial fungal infection.
The epithelial cells are physically the first defensive surface barrier against C. albicans caused invasion and tissue damage. The f Decreased ROS released by TR146 epithelial cells after infected with erg3Δ/Δ and erg11Δ/Δ strain compared with WT strain (fold change to 0 min). g IL-1α production at 24 h post infection, 1 × 10 4 cells per mL. h Relative expression of virulence factors in C. albicans erg3Δ/Δ and erg11Δ/Δ strain compared with WT strain measured by RT-qPCR. LDH lactate dehydrogenase transition between C. albicans yeast and hyphal forms has been proved as the most essential virulence factor by numerous investigations both in vitro and in vivo. 39,40 However, the recent identified fungal peptide toxin Ece1 functions as the key element to damage the epithelial cell instead of the hyphae of C. albicans as the deletion of ECE1 also formed the morphological identical hyphae compared to the wild type with the same invasive ability. 19 This type of genes whose deletion will lost epithelial cell damage without the effect on hyphal formation is important to understand the pathogenesis of C. albicans during the superficial infection and can be served as new therapeutic potential targets for the treatment of mucosal candidiasis. Our results from the contribution of ERG3 and ERG11 in epithelial cell infection model and murine oropharyngeal candidiasis model confirmed that these two genes were another "missing-link" genes between epithelial damage and hyphal development. The erg3Δ/Δ and erg11Δ/Δ in this study combined previous ece1Δ/Δ suggest that this type of "missing-link" genes is likely related to the virulence factors as ERG3 and ERG11 deletion decreased the virulence factors in our study while Ece1 acted as toxin to epithelium itself. In view of C. albicans infection in vivo, some of its immune evasion-related genes whose deletion will not affect the hyphal development may also lost the infectious ability due to the clearance of immune system in vivo (Y. Zhou et al. unpublished data.2017). Therefore, the "missing-link" genes may typically include virulence and immune evasion correlative genes or pathways, which will be new type of targets for antifungal drug discovery beyond the killing targets. These targets may reduce the fungal drug resistance as the inhibition to these genes will not kill the fungi but cause their incapability in infection.
Owing to the important role of ergosterol in fungal membranes, azole drugs that inhibit ergosterol biosynthesis are widely used for the treatment of fungal infections. 41 Usually, azole drugs, such as FLC, are clinically used at killing dosage to inhibit the growth of C. albicans. Here we identified the multi-functions of FLC, which can not only eliminate the fungi but also inhibit the infective virulence   Relative quantification of differentially expressed genes by real time PCR C. albicans cultures were harvested by centrifugation at 6 000 r·min -1 at 4°C for 5 min. The pellets were flash frozen in liquid nitrogen and stored at −80°C until RNA preparation. RNA isolation was carried out according to the E.Z.N.A. ® Yeast RNA Kit (OMEGA Bio-tek.) instructions. Then 1 µg RNA was subjected to the One Step RNA PCR kit (Takara Inc.) to prepare the cDNA according to the manufacturer's instructions. The RT-PCR were then proceeded by using the SYBR® PremixEx TaqTM kit (Takara Inc.) with following two-step strategy: (1) 94°C for 30 s; (2) 40 PCR cycles (94°C for 30 s, a gene-specific annealing temperature for 30 s). All primer sequences are listed in Table S2. Real time PCRs of triplicate samples were performed using CFX 96 Touch TM (Bio-Rad, Hercules, CA, USA). The gene expression level relative to the calibrator was expressed as 2 −ΔΔCT .

MATERIALS AND METHODS
Cytokine production qualification assay Cytokine levels produced from cell culture supernatants were determined using Quantikine ® ELISA kit (R&D Systems, USA) and a Varioskan Flash machine (Thermo Scientific) according to the manufacture's introduction. All of the experiments were performed in triplicates.
Cell damage assay TR146 cells were grown to confluence on 24-well plates for 48 h in DMEM medium. Lactate dehydrogenase (LDH) array were conducted to determine the TR146 cell damage after the cells co-cultured with C. albicans or FLC. Briefly, the culture supernatants were collected after incubation for 24 h and subjected to the LDH activity test by using a Roche cytotoxicity detection kit plus according to the manufacturer's instructions. All of the experiments were performed at least in triplicates.
Epithelial cell adhesion assay TR146 cells were grown to confluence on 24-well plates for 48 h in DMEM medium. After washed by PBS for three times, 1 mL serumfree DMEM of C. albicans yeast cells (2 × 10 5 cells per mL) were added and then followed the incubation for 60 min (37°C, 5% CO 2 ). The non-adherent C. albicans cells were aspirated and washed with PBS for three times.  19,42 Briefly, female BALB/c mice were injected subcutaneously with 3 mg per mouse (in 200 μL PBS with 0.5 % Tween 80) of cortisone acetate on days before and post infection. The second day after injected, mice were in a coma for at least 75 min with an intra-peritoneal injection of 5 % chloral hydrate 10 mL·kg −1 . Then a swab soaked in a 1 × 10 7 CFU·mL −1 of C. albicans yeast in sterile saline was placed under the tongue. After 2 days, mice were executed. The tongue was cut out and divided longitudinally in two. After weighed, one half was homogenized and cultured to quantify candida counts on CHROMagar TM Candida plate, whereas the other one was processed for histopathology analysis. To monitor the efficacy of fluconazole, different dosages of fluconazole were added into the drinking water after the mice were infected.
Immunohistochemistry of murine tissue C. albicans-infected murine tongues were fixed in 10% (v/v) formaldehyde before being embedded and processed in paraffin wax using standard protocols. For each tongue, 5-μm sections were prepared using a Leica microtome and silane-coated slides. Sections were dewaxed using xylene. Then C. albicans and infiltrating inflammatory cells were visualized by staining using Periodic Acid-Schiff (PAS) stain and hematein eosin (HE) stain. Sections were then examined by light microscopy. Histological quantification of infection was undertaken by measuring the area of infected epithelium and expressed as a percentage relative to the entire epithelial area.
ROS assays TR146 cells were grown to confluence on 96-well plates for 24 h in DMEM medium. The ROS production was determined using a Reactive Oxygen Species Assay Kit (Beyotime, China) according to the manufacturer's instructions. Briefly, cells were loaded with 10 μmol·L -1 H2DCF-DA in serum-free DMEM for 20 min in a humidified incubator (37°C, 5% CO 2 ) in the dark. After washing with serum-free DMEM for three times, 100 mL DMEM medium containing Rosup (100 mg·mL −1 ) was served as positive controls and equal volumes of C. albicans strains with or without fluconazole at the indicated concentrations were added.
Chemiluminescence was measured at 15 min intervals at 37°C with a Varioskan Flash machine (Thermo Scientific). Data are expressed as relative luciferase per well TR146 cells over time.

Statistics
Statistical significance was decided by Student's t-test with Welch's correction, one-way ANOVA with Dunnett's or Tukey's multiple comparison test, or two-way ANOVA with Tukey's multiple comparison test using GraphPad Prism software. For data plotted on a logarithmic scale the geometric mean is indicated, and data were log-transformed before statistical analysis.