To generate GUT-like C. albicans cells we overexpressed the gene WOR1 in a/α MTL background, as reported elsewhere (Pande et al., 2013). For this purpose, we chose in this study the doxycycline regulated promoter (Park and Morschhauser, 2005) to express WOR1, as it represents a strong promoter widely used in C. albicans research and, most importantly, because of its ability to regulate gene expression in vivo. Therefore, a genetic construction carrying a myc-epitope tagged version of WOR1 regulated under the TET-OFF system was integrated at the ADH1 locus in C. albicans. The strain obtained, CAI4-WOR1^OE, was shown to produce Wor1-myc as a protein of ≈90 kDa whose expression was doxycycline-dependent and similar to the 84.44 kDa predicted molecular weight of the construct (Supplementary Figure 1A). In addition, CAI4-WOR1^OE retained MTL heterozygosity (Table 1 and Supplementary Figure 1B) excluding the possibility that the transformation altered the mating type (Selmecki et al., 2010). We observed that CAI4-WOR1^OE cells and colonies in vitro appeared as opaque. The colonies stained with phloxine B, as opaque cells, (Table 1 and Supplementary Figure 1C) and pimples were absent at the cell surface. This constrasts with opaque cells where these structures are typically found (Table 1 and Supplementary Figure 1D). Phloxine B positive phenotype was stable over time and independent of the temperature (37 or 21°C). The addition of doxycycline reversed en-masse the population to phloxine B negative cells, even at 21°C. This effect was reversed upon removal of doxycycline (Supplementary Figure 1E).

In order to determine whether overexpression of WOR1 generates cells with increased colonization fitness, we performed competitive colonization experiments between CAI4-WOR1^OE and an isogenic strain expressing RFP instead of WOR1 (Prieto et al., 2014). Analysis of CFUs from stools of mice revealed that colonization of both strains was in the range of 10⁷ CFUs/g; however, strain CAI4-WOR1^OE showed an increased colonization ≈2 weeks after gavage (Figure 1A) that was lost over time. This was more apparent in the relative ratios of both strains determined by the log₂(R/I) index (Pande et al., 2013) which reflects the ratio of the abundance of each strain compared to the starting (inoculum) amounts (Figure 1B). However, these changes were not maintained at later time points when differences in colonization were statically non-significant after 24 days (p = 0.93) and higher times (not shown).

We conclude from these set of experiments that sustained ectopic expression of WOR1 from the tetracycline promoter results in GUT-like cells in vitro, which behave similar to the previously described GUT phenotype. Importantly, we note that increased fitness is not maintained in prolonged colonization studies of greater than 3 weeks.

While performing competitive colonization experiments we noticed that although precise equal proportions of viable cells (as determined by CFU counting on solid media) of wt and WOR1^OE were adjusted in the inoculum, the proportions of CAI4-WOR1^OE and wt in the first days of colonization were significantly different. This was especially surprising, as fecal CFUs counts during the first days after inoculation closely resemble the initial inoculum dose (Prieto and Pla, 2015). As shown in Figure 2A, colonization initiated with a dose of 10⁷ total cells (1:1 proportion) rendered ≈10⁶ CFUs/g of WOR1^OE at day 1 and ≈10⁵ at day 3, with the ability to reach wt colonization levels observable 2 weeks later. In order to define more precisely this phenomenon, we performed time course experiments for competitive colonization. We analyzed the CFUs of stools at very early times points of 15, 24, and 72 h after gavage (Figure 2B). Although, the inoculum proportions of CAI4-WOR1^OE in the particular experiment shown in the figure was about 60%, it diminished to ≈40% after 15 h, ≈30% after 24 h and to only ≈2% after 3 days. One caution in the interpretation of this data is that CFU actually determines distal (cecum/large intestine) colonization, but the location of the cells (and their relative proportions) could vary between different regions of the intestine regions. We, therefore, performed a post mortem analysis of mice colonized simultaneously with both CAI4-WOR1^OE and CAI4-RFP after 28 days, when colonization levels were approximately similar (≈10⁷ CFUs/g). Different regions of the intestine, stomach, proximal small intestine, distal small intestine, cecum, and large intestine, were processed and CFUs were counted. Interestingly, the percentage of WOR1^OE cells varied between 19% in the stomach, 23% in the proximal intestine and 16% in the distal intestine. By contrast we obtained levels as high as 47% in the cecum and 42% in the large intestine (Figure 2C). Therefore, we conclude that a prevalence of wt over CAI4-WOR1^OE occurs in the upper intestinal tract. We finally performed co-housing experiments, where new non-colonized mice are added to the same cage of mice already colonized so that they are inoculated “naturally” via coprophagy. Mice inoculated with a mixture of wt and WOR1^OE cells were co-caged with new mice at day 28. As shown in Figure 2D, CAI4-WOR1^OE and wt cells colonized these new mice with similar efficiency and there was no apparent decrease in CFUs in stools of CAI4-WOR1^OE (see days 29 and 31). Collectively, these experiments indicate that overexpression of WOR1 results in a “barrier” effect upon entry of in vitro cultured cells in the mouse gastrointestinal tract, which is reflected in its preferential colonization of specific gastrointestinal regions.

One of the reasons for the altered proportions of WOR1^OE cells in early stools and gut locations could be the susceptibility to bile salts, which are mainly present in the small intestine. We tested this assumption by plating dilutions of overnight growing (YPD) wild type and WOR1^OE cells in solid YPD medium supplemented with bile salts. As observed in Figure 3A, overexpression of both RFP and WOR1 caused a slight reduction in growth (compare lanes ± DOX), slightly more pronounced in WOR1^OE cells. CAI4-WOR1^OE cells were found to be significantly sensitive to bile salts. This susceptibility is indeed determined by WOR1 overexpression, as it is dependent on the presence of doxycycline in the medium, which mediates the repression of the ectopic WOR1 expression and restores wild type bile salt susceptibility (Figure 3A). In order to better mimic the gut atmosphere, we confirmed the phenotype also under oxygen-limiting microaerophilic conditions at 37°C, and noted that the sensitivity seems to be slightly higher than during normoxic growth (Figure 3B). As this phenotype is a new trait assigned to GUT-like cells and given the similarity between GUT and opaque cells (Table 1), we checked whether WO-1-derived opaque cells also shared it. We tested this at 21°C as 37°C forces a conversion of WO-1 opaque cells to white cells. As shown in Figure 3B, WO-1 opaque cells, but not white cells, present a clear sensitivity to bile salts and again, this effect was somewhat increased at low oxygen levels. Therefore, overexpression of WOR1 leads to susceptibility of cells to bile salts, especially under specific oxygen limiting environments.

A feature of the gut is the limited oxygen availability, especially in the distal portions (He et al., 1999). We reasoned that one of the mechanisms by which GUT-like cells could be adapted to this niche would be a different respiratory metabolism. We, therefore, tested oxygen consumption in wt and WOR1^OE cells in the presence of different inhibitors of respiration. The oxygen consumption of both strains was found to be similar (0.56 ± 0.05 min⁻¹/10⁶ cells for wt and 0.59 ± 0.02 min⁻¹/10⁶ cells for WOR1^OE) under standard laboratory conditions for cells grown in YPG medium (Supplementary Figure 2B). We used Antimycin A and sodium azide to inhibit Complex III and IV of the electron transport chain, respectively (Supplementary Figure 2). When high doses were used, both compounds blocked oxygen consumption of both wt and WOR1^OE strains. Nevertheless, under limiting concentrations of these compounds, the CAI4-WOR1^OE strain displayed a higher sensitivity to both Antimycin A and sodium azide. In the presence of Antimycin A at 10 μg/mL the relative respiration rate values were 0.86 ± 0.16 for CAI4-RFP and 0.13 ± 0.074 % for CAI4-WOR1^OE cells. Cells exposed to 0.01‰ sodium azide displayed a relative respiration rate of 1.2 ± 0.02 in wt cells and 0.74 ± 0.05 in WOR1^OE cells (Figure 4A). In fact, when we tested sodium azide on solid YPD medium, we observed that CAI4-WOR1^OE cells were hypersensitive to this compound, a phenotype that did not occur in the presence of doxycycline (Figure 4B). Interestingly, this feature was not shared by the WO-1 strain (Figure 4C) that seems to be intrinsically more resistant to this drug. Therefore, the respiratory metabolism of CAI4-WOR1^OE is altered compared to wild type cells.

Susceptibility to bile salts does not explain why within a certain range of time (2-3 weeks in our model) colonization of CAI4-WOR1^OE is favored over wt cells. To address this question, we tested the adhesion capacity using a competition assay. In this type of experiments, RFP-labeled wild type cells were used as internal control and mixed in equal proportions with a tester strain. After interaction with the adhesion surface, cells are recovered and the relative proportions of both cells types relative to the inoculum is determined to calculate an Adhesion Relative Index (ARI). This approach results in less experimental variability than that observed in absolute adhesion tests (Prieto et al., 2014). We tested CAI4-WOR1^OE using a biotic (the large intestine mucosa) and an abiotic (polystyrene) surface. WOR1^OE cells showed increased adhesion to the mouse gut mucosa relative to wt cells in the samples that employed large intestine samples with an ARI = 1.40 ± 0.13 (mean ± SEM) (Figure 5). This effect was also observed in the small intestine tissue samples (data not shown). Interestingly, CAI4-WOR1^OE cells showed the opposite effect when tested on 24-well plates (polystyrene abiotic surface), ARI = 0.43 ± 0.08 (mean ± SEM). This indicates that WOR1 overexpression results in cells with enhanced competitive adhesion to the mouse intestine although this effect is not general to all surfaces.

The strains used are described in Table 2. Cells were grown at 37°C in YPD medium (1% yeast extract, 2% peptone, and 2% dextrose) unless otherwise stated. Two independent WOR1^OE clones (c1 and c2) were generated which gave similar expression levels and in vitro phenotypes (not shown) and only one (c1) was used for in vivo studies. The susceptibility/resistance to different compounds was performed through drop test as follows. Cultures grown at 37°C from either stationary or exponential phase (O.D. = 1) were adjusted to 2 × 10⁷ cells/mL, serially 10-fold diluted and deposited (5 μL) onto solid YPD plates supplemented (or not) with the indicated compounds. Plates were incubated at 37°C for 24 and 48 h before scanned. Microaerophilia was achieved using an anaerobic chamber and a commercial system that ensures the adequate percentages of O₂ and CO₂ (GENBox Microaer, BioMérieux, ≈15% CO₂ and ≈6% O₂)_. For the observation of white-opaque switching, C. albicans strains were grown on YPD plates supplemented with phloxine B (10 μg/mL) at 37 and 21°C. When necessary, doxycycline was added to either liquid or solid media at 10 or 20 μg/mL respectively.

To achieve ORF ectopic expression, RFP labeling and myc fusion, plasmids pNRUX-RFP, and pNRUX-WOR1 were constructed. These plasmids contain a tetracycline repressible promoter. Strains were obtained by integrating ectopically a Kpn I-Sac II fragment of plasmids pNRUX-RFP or pNRUX-WOR1 in the ADH1 locus of CAI4. pNRUX-RFP plasmid was generated by replacing the 5′ ADH1 Xba I-Sac II fragment (880 bp) from pNRU-RFP (Correia et al., 2016) with a 1,630 bp Xba I-Sac II fragment from the pNIMX vector (Chauvel et al., 2012) containing the 5′ADH1 and the TDH3 promoter (TDH3^PR). The WOR1 ORF was amplified by PCR from the clinical isolate SC5314 strain using the primers up-WOR1-myc (GAGATGTCGACAATGTCTAATTCAAGTATAGTCCCTACATATAATG) and rev-WOR1-myc (TCGCGGCCGCGAGTACCGGTGTAATACGACCCAG). The 2364 bp PCR product was cloned in the intermediate pGEMT plasmid (Promega), digested with Sal I and Not I and accommodated in the pNRUX-RFP, previously digested with Sal I and Not I to generate the pNRUX-WOR1 vector. Homologous recombination occurs at the ADH1 locus following a Kpn I-Sac II digestion and transformation of C. albicans using described procedures (Kohler et al., 1997). The correct integration was checked by Southern Blot. MTL status was determined by PCR of genomic DNA using the following primers: MTLa-up (TTGAAGCGTGAGAGGCTAGGAG), MTLa-lo (ATCAATTCCCTTTCTCTTCGATTAGG), MTLα-up (TTCGAGTACATTCTGGTCGCG) and MTLα-lo (TTCGAGTACATTCTGGTCGCG).

All procedures involving cell lysis, protein extraction, gel electrophoresis, and transfer to nitrocellulose membranes were made as previously described (Martín et al., 1993; Martin et al., 2000). Protein extracts were measured at A_280nm to equalize the amount of protein loaded for western blot analysis and Blots were probed with anti-myc, clone 4A6 (Millipore). Western blots were developed according to the manufacturer's conditions using the Hybond ECL kit (Amersham Pharmacia Biotech).

C. albicans strains were grown in YPG at 30°C until they reached an O.D. ≈1–1.5. Cells from 20–30 mL cultures were recovered by low speed centrifugation (5,000 rpm, 3 min), washed twice with PBS and suspended in 1 mL PBS. To quantify oxygen consumption, an oxygen electrode (Hansatech Instruments) was used. The oxygen electrode chamber was maintained at 30°C, filled with 1 mL PBS and 10 μL of the cell suspension was introduced in the electron chamber. Different compounds that inhibit specific complexes of the electron transport chain were added to the electrode chamber to quantify the contribution of different complexes to the respiration rate. Antimycin A (Sigma-Aldrich) and sodium azide (Panreac) were added at the concentration indicated in the figures to the electron chamber. The relative respiration rate was determined by dividing the respiration rate in the presence of the inhibitor by the respiration rate in its absence (basal).

The gut colonization assay was performed following the protocol described previously (Prieto et al., 2014). Briefly, after 4 days of antibiotic pre-treatment (2 mg/mL streptomycin, 1 mg/mL bacitracin, and 0.1 mg/mL gentamycin), 10⁷ C. albicans cells were intragastrically inoculated by gavage. Stool samples were obtained at different days and homogenized in PBS prior to plating on SD plates with chloramphenicol (20 μg/mL) and YPD plates with phloxine B (10 μg/mL) to quantify CFUs. To analyze C. albicans loads in the gastrointestinal tract, mice were sacrificed and samples from the stomach, cecum, small and large intestine were aseptically obtained, homogenized in sterile PBS and cultured in SD plates. Female mice C57BL/6 were purchased from Harlan Laboratories, Inc. (Italy) and used within an age of 7–10 weeks-old. Mice housing and other non-invasive procedures took place in the animal facility at the Medical School of the Universidad Complutense de Madrid. All experiments involving animals performed in this work were carried out in strict accordance with the regulations in the “Real Decreto 1201/2005, BOE 252” for the Care and Use of Laboratory Animals of the “Ministerio de la Presidencia,” Spain. The protocol used in the commensalism model was approved by the Animal Experimentation Committee of the University Complutense of Madrid (CEA 33-2015) and Comunidad de Madrid according to Artículo 34 del RD 53/2013 (PROEX 226/15). The treatments here did not result in disease in the animals; nevertheless, all procedures were conducted minimizing any suffering. The number of animals per experiment was adjusted to a minimum for ethical reasons. Experiments were done at least twice (normally three) and only one representative experiment is shown in the figures.

CAI4-WOR1^OE cells were incubated in YPD medium for 24 h at 37°C. After washing, cells were fixed overnight at 4°C with 2.5% glutaraldehyde in 0.1 M Na-cacodylate (pH 7.4). Then cells were washed and post-fixed for 2 h in 1% osmium tetroxide in 0.1 M sodium cacodylate buffer (pH 7.4), gradually dehydrated in ethanol and dried. A JEOL JSM-6400 microscope was used for visualization.

Adhesion to polystyrene was performed in 24-well flat bottom plates for culture cells. C. albicans strains from overnight YPD cultures were mixed (1:1) and adjusted to 2 × 10⁴ cells/mL concentration. 10⁵ cells were added to each well in YPD medium and allowed to adhere for 90 min at 37°C. Once non-adhered cells were washed out with sterile PBS (3 times), then adhered cells were mechanically removed and spread on SD plates with chloramphenicol (20 μg/mL) for CFUs count. Adhesion to intestinal mucosa we assessed as previously described (Prieto et al., 2014). Briefly, a 1 cm-piece of the large intestine, was opened, washed and placed in a 4 mm-diameter methacrylate chamber, which was filled with RPMI medium pre-warmed at 37°C. Then, C. albicans strains from overnight YPD cultures were mixed (1:1) and adjusted to 2.5 × 10⁷ cells/mL concentration in serum-free RPMI medium. 10⁶ yeast cells from this suspension were placed in the lumen side from the colonic tissue and incubated for 150 min at 37°C. Then, the piece of intestine was carefully washed with sterile PBS twice and mechanically disaggregated. This fraction was spread on SD plates with chloramphenicol (20 μg/mL) and YPD plates with phloxine B (10 μg/mL) for CFUs determination. An internal control (CAI4-RFP) was introduced in all adhesion assays; therefore, adherence is quantified by the ARI. This index is calculated by dividing the relative amounts of the strain under analysis in the adhered cells' fraction relative to the value of that same strain in the inoculum.

Statistical differences between two groups were calculated using Student's two-tailed unpaired t-tests. Statistical differences between more than two groups were calculated using One-way ANOVA correcting for multiple comparisons using Tukey method. Only p < 0.05 were considered significant.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.