Phenotypic differences among the environmental Candida albicans isolates

We used the three oak trees isolates (S1 Table: Strains used in this study) to characterize their phenotypes and investigate the possibility of an environmental reservoir for pathogenic C. albicans strains. To this end, we investigated their growth under different conditions and survival upon amoebae predation, and then tested clinically relevant parameters, including virulence attributes and antifungal susceptibility.

First, we evaluated the metabolic plasticity of the oak tree isolates by assessing their growth on a synthetic defined medium supplemented with different carbon (Fig 1A) and nitrogen (Fig 1B) sources. Importantly, these included compounds that are present in the human gut, such as the host dietary sugars galactose, fructose, and sucrose [32,33], and alternative carbon and nitrogen sources like short-chain fatty acids, N-acetylglucosamine (GlcNAc), and amino acids [34–36]. Significant differences in the patterns of nutrient use were found across the environmental strains. One of the C. albicans oak tree isolates, Oak 2, grew comparable to the reference strain SC5314 in most tested carbon sources, with the exception of citric acid, acetate, and citrate (Fig 1A). Oak 3 grew similarly well in most tested carbon sources, except for a few sugars such as galactose and maltose, and showed significantly better growth in citric acid. Oak 1 showed reduced growth in host dietary sugars like glucose, fructose, and galactose, as well as in multiple alternative sugars such as maltose, fucose, rhamnose, raffinose, and ribose. Similar to Oak 3, Oak 1 grew better in citric acid than SC5314. Moreover, Oak 1 showed deficits in growth on all tested alternative nitrogen sources, except for arginine (Fig 1A, B). Oak 3 grew slower on certain nitrogen sources like phenylalanine, tryptophan, and with low levels of ammonium sulfate. Finally, Oak 2 grew similarly to SC5314 on the tested nitrogen sources (Fig 1B), with the exception of phenylalanine, tryptophan, and very low concentrations of ammonium sulfate. In general, the oak tree isolates were not unusual in their nutrient utilization spectrum, as we observed similarly reduced growth (compared to SC5315) in galactose and maltose for at least two of the four tested clinical C. albicans isolates (S1A Fig). Moreover, all four clinical isolates had a somewhat lower growth in alternative nitrogen sources, including amino acids such as phenylalanine, tryptophan, histidine, and lysine (S1B Fig), similar to Oak 1 and Oak 3. Overall, the nutrient utilization spectra of clinical and environmental isolates therefore overlapped significantly for our tested strains.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. The oak tree strains display significant differences in nutrient use, and Oak 2 shows resistance to amphotericin B.

Growth of all three oak tree isolates was assessed as growth curves at 30 °C in different media comprising specific carbon (A) or nitrogen (B) sources. The growth is shown as area under the curve relative to the reference strain SC5314 (% growth). Asterisks indicate significance compared to SC5314 calculated using a one-way ANOVA with Tukey’s multiple comparisons test (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001) (n = 3). (C, D) Resistance to fluconazole and amphotericin B was determined by performing a drop test of serially diluted C. albicans solutions on YPD agar plates supplemented with the indicated antifungal concentration. Pictures were taken (C) after 2 days at 30 °C and (D) after 6 days at 37 °C. For all pictures, contrast and brightness were increased by 20%.

https://doi.org/10.1371/journal.ppat.1013542.g001

We hypothesized that the oak tree isolates were potentially exposed to amoeboid predators in the environment, similar to other environmental fungal pathogens [37,38], which may have selected for the ability to withstand such attacks. To test this, we evaluated the survival of the oak tree strains upon amoeba predation after 3 h. For all three environmental isolates, almost 100% of the inoculum survived the amoebae contact, which is significantly more than SC5314 (54% inoculum survival) (S1C Fig). Moreover, when analyzing the phagocytosis rate of the oak tree isolates after 2 h of amoeba predation, Oak 1, Oak 2 and Oak 3 were significantly less phagocytosed by the amoebae than SC5314 (S1D Fig), indicating a potential adaptation to amoeba predation.

We assumed that the oak tree isolates had not been in contact with clinically prescribed antimycotics. We therefore tested their antifungal drug susceptibilities by performing drop tests of serial dilutions on agar plates supplemented with commonly prescribed antifungals, namely azoles and polyenes. All oak tree isolates, especially Oak 1, showed better growth in the presence of fluconazole than SC5314 (Fig 1C). Most surprisingly, when the strains were tested for growth in the presence of amphotericin B (AmB), Oak 2 showed a high degree of resistance (Fig 1D).

One oak tree C. albicans isolate with reference strain-like host cell damage potential

We expected that the environmental strains are not adapted to mammals, and therefore tested the damaging potential of all three oak tree isolates on epithelial cells. Host cell damage was determined via lactate dehydrogenase (LDH) release 24 h and 48 h after infection of oral or intestinal epithelial cells, respectively, to account for the different infection kinetics with the different host cells (Fig 2A, B). We found that Oak 1, Oak 3, and the tested clinical isolates did not significantly damage the oral and intestinal epithelial cells, when compared to SC5314 (Fig 2A, B). Interestingly, the AmB-resistant strain, Oak 2, induced damage to intestinal epithelial cells at a level similar to the high-damaging SC5314 strain [39,40] (55% of complete lysis), and reached 64% of the full lysis control on oral epithelial cells (Fig 2A, B). On both epithelial cell types, Oak 2 was among the most damaging strains tested.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. One oak tree C. albicans isolate shows an increased virulence potential on epithelial cells and augmented survival in an ex vivo blood model.

(A) Damage to oral epithelial cells by the environmental oak tree strains and clinical isolates was determined after 24 h by LDH release (n = 4). The lower dashed line indicates the uninfected control. The upper dashed line indicates the full lysis control. One-way ANOVA with Tukey’s multiple comparisons test (** p < 0.01, **** p < 0.0001). (B) Damage to intestinal cells was monitored 48 h post infection by quantifying LDH release (n = 3). One-way ANOVA with Tukey’s multiple comparisons test (* p < 0.05). (C) Adhesion of environmental strains was determined relative to SC5314 at 1 h post-infection of oral cells by staining and microscopy (n = 3). (D) Hypha formation on oral cells was determined by microscopic evaluation 3 h post infection. One-way ANOVA with Tukey’s multiple comparisons test (**** p < 0.0001) (for oak strains and SC5314: n = 3-6, clinical isolates: n = 2-3). (E) Invasion was measured by differential staining of oral cells 3 h post infection. Invasion was calculated as percentage of invading hyphae compared to all counted hyphae (n = 3). (F) Survival of C. albicans in whole human blood was assessed over the course of 4 h by plating surviving colony-forming units (4 donors). Survival is shown relative to the inoculum. Heat map displays fold change to SC5314 and p-values calculated with a two-way ANOVA with repeated measures and Dunnett’s multiple comparison test. Significant p-values are indicated in bold.

https://doi.org/10.1371/journal.ppat.1013542.g002

To investigate the basis for this damaging potential further, its underlying features – hypha formation, adhesion, and invasion – were quantified on oral epithelial cells. Adhesion of all oak tree isolates to host cells was comparable to SC5314 after 1 h, except for Oak 3, which displayed a 2-fold higher, albeit not statistically significant, level of adhesion (Fig 2C). Oak 1 and Oak 3 showed a significantly lower rate of filamentation (26% and 42%, respectively) and invasion compared to SC5314 on oral cells (Fig 2D, E), and grew mainly as yeast cells. Interestingly, Oak 2 showed SC5314-like filamentation (99%) and invasion into oral epithelial cells.

To further characterize the virulence potential of Oak 2, the survival rate of this strain was quantified as compared to SC5314 in an ex vivo human whole blood model [41]. Oak 1 was included as a non-virulent, environmental control strain. Remarkably, both strains, the high-damaging Oak 2 and the low-damaging Oak 1, survived significantly better in whole human blood than SC5314 over 4 h, with a 3-fold higher number of surviving cells at the final time point (Fig 2F). Collectively, Oak 2 could therefore represent a new C. albicans clade with high damaging potential and a high level of antimicrobial resistance, which can also survive and thrive outside of the human host.

Differences in candidalysin sequence and expression kinetics among the oak-derived isolates

We went on to investigate the basis for the high damage potential of Oak 2. The most important contributor to host cell damage by C. albicans is candidalysin (CaL), a hypha-associated pore-forming peptide toxin encoded (together with other peptides) by the ECE1 gene [42–44]. We therefore determined the transcript levels of ECE1 during infection of epithelial cells by quantitative real-time PCR (Fig 3A, B). On oral epithelial cells, Oak 2 expressed lower levels of ECE1 transcripts compared to SC5314 (fold change of 0.5) at both tested time points (24 h and 48 h post infection), but the levels were still significantly higher than in Oak 1 and Oak 3 (Fig 3A). In fact, these strains did not produce measurable transcript levels of ECE1 during in vitro oral infection, although the gene is present in their genomes (Fig 3C). On intestinal cells, the ECE1 transcript level of Oak 2 reached 0.7 times that of SC5314 after 24 h, and even exceeded its level after 48 h (at 1.8-fold) (Fig 3B). Similar to oral cells, Oak 3 and Oak 1 had lower ECE1 levels than SC5314 on intestinal cells. Only Oak 1 reached a measurable transcript level, at 0.36 times that of SC5314 (Fig 3B).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. ECE1 (Extent of cell elongation 1) characteristics of Oak 2 resemble the SC5314 reference strain.

ECE1 expression on oral (A) and intestinal epithelial cells (B) is shown as fold change relative to SC5314 at the indicated time points. The expression is normalized to ACT1 mRNA levels. Statistical significance was calculated using a one-way ANOVA with Tukey’s multiple comparisons test (* p < 0.05, **** p < 0.0001). (C) Predicted Ece1 sequences were aligned using Clustal Omega. Identical amino acids are indicated as asterisks (*) and similar ones with dots (. or:). Blue indicates the signal sequence and pink amino acids (KR, KK, R) mark the Kex protease cleavage sites of the polypeptide. (D) Secretion of Ece1 peptides and peptide fragments from the indicated strains into culture medium, evaluated by LC-MS/MS analysis. The mean peptide-spectrum matches (PSM) scores for each identified peptide are presented (n = 2); if no number is given, the peptide was not detected. Oak 3 was not included in this measurement.

https://doi.org/10.1371/journal.ppat.1013542.g003

Although the amino acid sequence of Ece1 is largely conserved among C. albicans strains, some regions of candidalysin and the non-candidalysin-Ece1 peptide (NCEP) sequences are known to differ between clinical isolates, which can have an impact on candidalysin-mediated host cell damage [44–46]. We, therefore, performed whole genome sequencing of the environmental strains and aligned the predicted amino acid sequences of CaL and NCEP peptides of Oak 1, Oak 2, and Oak 3 to SC5314 (Fig 3C). There is little variation in Ece1 sequences overall (Ece1 sequence similarity to SC5314: 0.971 for Oak 1, 0.989 for Oak 2, 0.982 for Oak 3), and the majority of amino acid substitutions were found in peptide 2, peptide 3 (CaL), and peptide 7. In Oak 1 and Oak 2, we identified new amino acid sequence variants of CaL (S2E Fig) that differed from the 10 known variants from 182 clinical C. albicans isolates [46] – somewhat supporting their non-human origin. Only the amino acid sequence of Oak 3 is identical to one of the most common candidalysin variants identified in clinical isolates (S2E Fig) [46]. CaL sequence differences might contribute to the different levels of epithelial damage (Fig 2), as shown previously [46]. The other peptide sequences as well as the seven lysine-arginine (KR) motifs, which serve as cleavage sites for Golgi-located Kex proteases [42], are highly conserved among the environmental strains (Fig 3C). Alignments of the ECE1 promoter sequences (S2A, B, D Fig) revealed that both, Oak 1 and Oak 3 lack one out of two TATA boxes present in the SC5314 ECE1 promoter. While the first TATA box 109 bp upstream of the open reading frame of ECE1 is the primary regulator of gene expression, the second box at -284 bp is considered mostly dispensable for the transcription of ECE1 [47]. These differences may account for the lower ECE1 transcript levels in the two oak isolates. The promoter sequence of Oak 2 contained a single TATA box at a position much further upstream (-165 bp instead of -109 bp) (S2B Fig), potentially explaining the different expression dynamics on intestinal cells, with high transcript levels at a very late time point.

Like SC5314, all oak tree isolates also code for a secretion signal sequence at the 5’ end of the open reading frame of the ECE1 gene. We investigated the secretion efficiency of the Ece1 peptides, including candidalysin, NCEPs, and the corresponding fragments [48] as another potential explanation for the differences in host cell damage. To this end, the cell-damaging Oak 2, SC5314 as a positive control, and Oak 1 as a non-damaging isolate were cultured in strong hypha-inducing conditions and the protein content of the supernatant was analyzed by LC-MS/MS [49]. As expected, we found most Ece1 peptides and peptide fragments, including candidalysin, in the SC5314 supernatant, but no peptides other than peptide 7b in the non-virulent Oak 1 (Fig 3D). Oak 2 showed a moderate secretion of candidalysin and measurable levels of peptides 6a, 7a, and 7b (Fig 3D).

Overall, our results indicate that Oak 2, but not Oak 1 or Oak 3, is capable of inducing SC5314-like damage to oral and intestinal epithelial cells due to strong filamentation and subsequent invasion into epithelial cells, associated with a strong expression of ECE1 and secretion of candidalysin into the invasion pocket [49].

Oak 2, but not Oak 1 or Oak 3, inflict damage to macrophages

Since Oak 2 showed a high virulence potential toward human epithelial cells and a high survival rate in whole blood, we asked whether Oak 2 would also damage immune cells and thus investigated the interaction of the three oak tree isolates with primary human macrophages. In macrophages, Oak 2 again showed SC5314-like damage after 24 h, whereas Oak 1 and Oak 3 induced significantly less damage compared SC5314, similar to the other tested clinical isolates (Fig 4A). Similarly, Oak 2 induced macrophage membrane rupture to an extent similar to SC5314 (S8B Fig). However, in this assay Oak 3 reached a level comparable to Oak 2. Oak 1 induced significantly less host cell lysis than SC5314 (S8B Fig). Interestingly, all oak tree isolates survived inside the phagocytes significantly better than SC5314 (Figs 4B and S8D), with Oak 1 having the highest survival rate (388% of SC5314) (Fig 4B). To determine the contribution of filamentous vs. yeast growth to these differences, we determined the fungal morphology microscopically at 3 h post infection. For Oak 1 we observed stretched macrophages filled with yeasts, reflecting its high intracellular survival ability (Figs 4C and 8E). In line with the high damage induction, Oak 2 formed filaments similar to SC5314, and Oak 3 replicated as yeasts, but less than Oak 1. In conclusion, Oak 2 exerts strong pathogenicity toward macrophages, while Oak 1 showed an increased adaptation and survival inside macrophages. To investigate whether these different phenotypes are driven by the macrophages or due to the fungal-intrinsic strain variations, we investigated the immune activation of macrophages by the oak tree isolates (S9B Fig). Interestingly, all three strains did not differ significantly in triggering IL-8, TNFα, IL-6, and IL-1β release, except for Oak 3, which induced a significantly higher level of IL-8.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. The oak tree strains show differences in their interaction with primary macrophages compared to the reference strain.

(A) Damage to monocyte-derived macrophages by the environmental oak tree strains was determined after 24 h by measuring LDH release. The lower dashed line indicates the uninfected control. The upper dashed line indicates the full lysis control. Each circlerepresents one donor. Significance was calculated by a one-way ANOVA with Tukey’s multiple comparisons test (* p < 0.05, ** p < 0.01) (n = 4-6 donors). (B) Intracellular survival of the C. albicans strains in human monocyte-derived macrophages was assessed 3 h post infection by lysis and plating of the intracellular fungal cells. Survival is depicted relative to the reference strain SC5314. Each circlerepresents one donor. Significance was calculated by a one-sample t-test against 100% (* p < 0.05) (n = 6 donors). (C) Microscopic pictures were taken either 3 h after infecting human monocyte-derived macrophages (37 °C, CO₂) with a 63 × magnification with immersion oil (scale bar, 20 µm). Samples were stained with DAPI (nuclei staining, purple) and Concanavalin A-AlexaFluor647 (fungal staining, blue). Figure shows representative pictures from at least 4 donors.

https://doi.org/10.1371/journal.ppat.1013542.g004

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. Experimental evolution with a host dietary sugar leads to antifungal resistance and metabolic flexibility.

(A) Schematic overview of the evolution experiment in YNB medium supplemented with 2% galactose. Oak 1 and SC5314 overnight cultures were diluted 1:50 into the galactose-containing medium and were incubated for 6 weeks at 30 °C, with passaging them into fresh medium every 2-3 days when the early stationary phase was reached. Growth was measured as OD_600nm over the whole time course. Samples were taken and frozen after every passage. (B) After the evolution experiment, growth of the sugar-adapted strain (Evo) was characterized in multiple carbon and nitrogen sources. The growth is shown as area under the curve relative to the reference strain SC5314 (% growth). Black asterisks indicate significance compared to SC5314 and blue asterisks indicate whether the sugar-adapted strain grew significantly different compared to its parental Oak 1 strain. Significance was calculated by using a one-way ANOVA with Tukey’s multiple comparisons test (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001) (n = 3). (C, D) Antifungal resistance of the galactose-evolved strain was assessed by performing growth curves in YPD supplemented with either 32 µg/ml fluconazole or 0.5 µg/ml amphotericin B at 30 °C for 72 h, and is depicted as area under the curve. Significance was calculated by using one-way ANOVA with Tukey’s multiple comparisons test (* p < 0.05, ** p < 0.01, *** p < 0.001) (n = 3).

https://doi.org/10.1371/journal.ppat.1013542.g005

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. The sugar adaptation leads to transcriptomic remodeling.

(A) Principal component analysis of the transcriptomes for the SC5314, two oak tree isolates, and Evo on intestinal cells. Samples clustered together depending on their origin, and the sugar-adapted strain underwent a positive shift along the PC1 axis. (B) KEGG pathways enriched for genes that were significantly up- or down-regulated on intestinal epithelial cells 24 or 48 h post infection. % genes in pathway is indicated by the dot size and represents the number of genes within each pathway that were differentially regulated compared to the indicated reference strains (below the graph). The color scale indicates the adjusted p value.

https://doi.org/10.1371/journal.ppat.1013542.g006

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. The sugar-evolved Evo strain shows increased adhesion, filamentation, invasion, and host cell damage during interaction with epithelial cells.

(A) Damage to oral cells and (B) intestinal cells was evaluated by quantifying epithelial LDH release 48 h post infection (oral n = 3, intestinal n = 4). The lower dashed line indicates the uninfected control. The difference between Oak 1 and Evo was not statistically significant in (A) and (B). (C) Adhesion of C. albicans strains was assessed by fixing and staining the fungal cells 1 h post infection of oral cells. The average number of attached C. albicans cells per microscopic image is plotted relative to the reference strain SC5314 (n = 3). (D) Hypha formation was determined by microscopic evaluation counting the hypha-forming cells relative to all C. albicans cells on oral cells 3 h post infection (n = 3). (E) Hyphal length was measured 3 h post infection of oral epithelial cells (n = 3). The difference between Oak 1 and Evo was not statistically significant. (F) Invasion was measured by infecting oral cells, fixing, and differentially staining 3 h post infection. Invasion was calculated as percentage of invading hyphae compared to all counted fungal cells (n = 3). (G) ECE1 expression was analyzed after 3 h of growth in hypha-inducing medium. The expression is normalized to ACT1 (ΔCt). The control condition is the yeast morphology of the reference strain SC5314 (ΔΔCt) (n = 3). The difference between Oak 1 and Evo was not statistically significant. (A-G) Significance by one-way ANOVA with Tukey’s multiple comparisons test (* p < 0.05, ** p < 0.01, *** p < 0.001). (H) Predicted Ece1 sequences were aligned using Clustal Omega. Identical amino acids are indicated as asterisks (*). Blue indicates the signal sequence, and pink-colored amino acids show the Kex protease cleavage sites of the polypeptide.

https://doi.org/10.1371/journal.ppat.1013542.g007

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 8. Virulence of the parental Oak 1 isolate and the sugar-adapted Evo strain in a transwell assay and a complex gut-on-chip model.

(A) Translocation of the strains was measured in an intestinal transwell model by plating the translocated colonies 24 h post infection. Significance was calculated using a one-way ANOVA with Tukey’s multiple comparison test (** p < 0.01) (n = 3). (B) Barrier integrity was determined by measuring TEER (transepithelial electrical resistance) in an intestinal transwell model 24 h after infection. Significance was calculated using a one-way ANOVA with Tukey’s multiple comparison test (*** p < 0.001, **** p < 0.0001) (n = 3). (C) Percentage of adhesion to the gut epithelium in the gut-on-chip model was determined by plating the amount of non-attached C. albicans in the flow-through relative to the injected inoculum (n = 3). (D) Damage was measured as LDH release from both, the gut side and the vascular compartment of the gut-on-chip 24 h post infection (n = 3). (E) Fungal burden was determined by plating the intestinal dispersion (gut flow-through), intestinal fungal burden (intestinal lysate), vascular invasion (vascular lysate), and dissemination (vascular flow-through) 24 h post infection (n = 3). To show the variation between the different gut-on-chip experiments, the individual values of the biological replicates are shown in (D) and (E). Significance was calculated using a two-way ANOVA with Tukey’s multiple comparison test (** p < 0.01). The difference between Oak 1 and Evo was not statistically significant in panel (A), (B), (C), (D), and (E).

https://doi.org/10.1371/journal.ppat.1013542.g008

Mimicking C. albicans’ evolution in sugar-rich medium

Sugar-rich diets consumed in westernized countries typically contain low-digestible oligosaccharides [33,50]. These are comprised of sugars such as galactose and fructose, and can reach the large intestine where C. albicans mainly resides as a commensal [32,33]. Oak 1 has shown diminished growth with all tested host dietary sugars and a low virulence potential, which feasibly could be increased if the strain were to adapt better to the human host. We hypothesized that adaptations to components of westernized diets can influence C. albicans-host interactions, and therefore performed an in vitro evolution experiment that mimicked aspects of this diet change. To this end, we continuously passaged Oak 1 in triplicate with a host dietary sugar, galactose, at 2% as its sole carbon source (Fig 5A). For 6 weeks, the strain was transferred into fresh medium every two to three days, whenever it reached the late exponential growth phase. In preliminary screens for hypha formation, antifungal resistance, and interaction with host cells, two of the three evolved strains showed no significant differences compared to the parent strain (S3 Fig), and were hence not used for follow-up analyses. In contrast, one of the evolved replicates, from here on indicated as ‘Evo’, adapted well to growth on galactose and was therefore chosen for further investigation.

The sugar-adapted C. albicans strain shows an increased metabolic flexibility and augmented resistance to antifungals

To investigate the sugar-evolved Evo strain, we re-characterized its growth on dietary sugars and alternative carbon sources which can be present in the human body [32–36]. Additionally, several different nitrogen sources were investigated. Interestingly, the sugar adaptation significantly improved growth of the evolved strain on the tested dietary sugars, including fructose, galactose, and sucrose, as well as within most of the alternative nitrogen sources, except for two higher concentrations of ammonium sulfate, compared to the parent strain (Fig 5B, indicated with blue asterisks). Thus, the long-term adaptation to galactose increased the nutritional flexibility of the evolved strain from its parental stage. However, compared to SC5314, the sugar-evolved strain still grew at a significantly lower rate in glucose, galactose, and with low levels of ammonium sulfate (Fig 5B, indicated with black asterisks). Nevertheless, the overall nutrient utilization pattern became more similar to SC5314 – for instance with citric acid, where the Evo strain grew worse than its parent, but at a rate similar to SC5314 (Fig 5B).

We then tested the sugar-adapted Evo strain for its susceptibility to the antifungals fluconazole and amphotericin B. Strikingly, Evo displayed significantly better growth and thus an increased antifungal resistance compared to its parental strain or SC5314 in the presence of either antimycotic (Fig 5C, D). It seems that long-term exposure to a host dietary sugar induced resistance to multiple classes of antifungals in this strain.

Sugar adaptation led to a transcriptome shift in C. albicans during interaction with epithelial cells

In order to identify the underlying mechanism for the increased metabolic flexibility and antifungal resistances, we investigated the changes in the sugar-adapted C. albicans strain on a genomic and transcriptomic level. To this end, we performed whole genome sequencing of the sugar-adapted strain and aligned it to the haploid genome of the parental Oak 1 [17]. We reliably identified nine mutations, however, we did not find a likely candidate in the form of, e.g., a change in the coding sequence of a relevant gene. The majority of mutations were single nucleotide polymorphism or frame shifts in intergenic regions, except one missense mutation in a gene coding for a transmembrane sorting receptor for vacuolar hydrolases (S3 Table). These mutations may have contributed to the observed phenotypes, but their individual effects are difficult to ascertain especially in a non-reference strain of C. albicans. Similarly, we did not find any likely causative mutations in the other evolved strains (S3 Table).

We therefore performed RNA sequencing of Oak 1 and 2 and the sugar-adapted Evo after 24 h and 48 h on intestinal epithelial cells, which simulate a relevant host environment. Principal component analysis (PCA) analysis revealed that the transcript profiles of both oak tree isolates cluster together, independent of the time point of infection (Fig 6A). After the sugar adaptation, a shift along the PC1 axis was observable, suggesting that the evolution on the host dietary sugar led to the transcriptional shift.

We performed functional GO term enrichment analysis to find pathways differently regulated in the environmental and sugar-adapted strains compared to SC5314 during intestinal cell infection (Fig 6B). For Oak 1, we observed a higher expression of glyoxylate cycle and amino acid metabolism genes in comparison to SC5314. Furthermore, Oak 1 showed reduced transcript levels of sugar metabolism pathways and other metabolic processes like glycolysis, autophagy, fatty acid metabolism, general carbon metabolism, and protein degradation compared to SC5314. This likely led to the diminished growth of Oak 1 on most of the tested carbon and nitrogen sources (Fig 1A, B). Most of the differentially regulated GO terms enriched in Oak 2 were also enriched in Oak 1. The sugar-adapted Evo strain showed an enrichment of highly expressed galactose, sucrose, and fructose metabolism pathway genes (Figs 6B and S4), as well as of glycolysis and autophagy, compared to its parental strain.

Based on these findings, we specifically looked at the expression of genes involved in sensing of extracellular glucose and galactose levels (HGT4), transport of glucose (e.g., HGT17, HGT2, etc.) and galactose (HGT7), as well as genes involved in signaling downstream of the sugar sensor (S4 Fig). For all these genes we observed increased transcript levels in the sugar-adapted strain compared to its parental oak tree isolate. Other host dietary sugar transporter genes like MAL31, and downstream signaling genes such as MIG1, GAL7, GAL10, and GAL1 were similarly more transcribed in the Evo strain. The notable exception was the galactose transporter gene HGT7, which exhibited low transcript levels – very likely due to selection on high levels of galactose. Overall, the transcriptional changes indicate that the long-term adaptation to one host dietary sugar led to the emergence of a C. albicans strain adapted for better growth on multiple sugars.

Interestingly, ABC transporter gene transcript levels were higher in the sugar-evolved Evo strain than in its parental strain or SC5314. The ABC transporter class includes multidrug transporters that confer resistances to azoles [51], and their upregulation may explain the fluconazole resistance observed in the sugar-adapted strain.

The sugar-adapted C. albicans strain has increased virulence potential in vitro and in an intestinal organ-on-chip model of infection

Next, we characterized damage induction by the sugar-evolved Evo strain on oral and intestinal epithelial cells via release of LDH at 24 h and 48 h post infection. On oral cells, damage by the Evo strain, like its parent, remained only slightly above the non-infected control (Fig 7A). Interestingly, on intestinal cells, we observed a drastic increase in damage that reached SC5314-like levels (Fig 7B). Thus, our laboratory evolution on dietary sugars led to an increased damage toward epithelial cells of C. albicans’ intestinal niche.

To determine the causes for this surprising phenotype, we investigated hypha formation and invasion at 6 h after infecting intestinal epithelial cells. All three oak tree isolates as well as the sugar-adapted Evo showed shorter hyphae and less invasion compared to SC5314, while their filamentation rate was similar (S5 Fig). To investigate the phenotype of the sugar-evolved strain in the oral niche, we tested adhesion, hypha formation, and invasion with oral cells at 1 h and 3 h post infection. Adhesion of Evo was higher compared to both, SC5314 and the parental oak tree isolate (about 270% of SC5314 and 245% of the parent adhesion level, respectively) (Fig 7C). Similarly, hypha formation and invasion of Evo significantly increased, up to SC5314-like levels (Fig 7D, F). The hyphal length, however, remained significantly shorter than SC5314 (Fig 7E).

We hypothesized that the increased filamentation might have led to enhanced expression of hypha-associated genes [52] in the sugar-evolved strain, contributing to the increased damage. We indeed found increased expression of two aspartic protease genes, SAP4 and SAP6, at both, 24 h and 48 h after infection of intestinal cells, as well as for the adhesin gene ALS1 at the later time point (S6 Fig). All other hypha-associated genes, however, such as ALS3, ECE1, or HWP1 did not differ in their expression. This suggested that the increased damage by the sugar-adapted strain at these later infection stages is mainly driven by its enhanced filamentation and not by an increased transcription of hypha-associated genes. One possible mechanism linking filamentation and invasion to epithelial cell damage is the delivery of the toxin candidalysin into the invasion pocket [49]. We therefore first verified, by qRT-PCR, that there were no differences in ECE1 transcript levels between the sugar-adapted strain and its parent on oral and intestinal cells at 24 h and 48 h (S6 Fig). We additionally examined the expression of ECE1 at 3 h, when the ECE1 transcript levels are generally highest in C. albicans, which is known to be necessary for damage to the host [49]. Interestingly, the sugar-adapted strain showed increased ECE1 transcript levels compared to its parental strain at 3 h in hypha-inducing conditions (Fig 7G). We also investigated the Ece1 peptide sequences and searched for mutations that may have evolved during our sugar evolution experiment, but found no genetic changes (Fig 7H). For the promoter region of ECE1, only minor mutations were found (S2C Fig), which did not affect the TATA box at -109 bp, the main cis-factor for ECE1 transcription in C. albicans [47].

In conclusion, adaptation to a dietary sugar led to the emergence of a C. albicans strain with increased filamentation on host cells and stronger invasion into the epithelial cell layer.

Higher damage capacity and increased filamentous growth are often associated with an increased ability for translocation through the intestinal epithelium [53,54]. Therefore, we tested the translocation abilities of the different strains using a transwell model with intestinal cells [53]. SC5314 showed the highest translocation 24 h post infection, while Oak 1 showed the lowest and Oak 2 had an intermediate level (Figs 8A and S7A), largely matching our intestinal cell damage findings (Fig 2B). For the sugar-adapted Evo strain, the translocation rate was increased compared to its parental strain, even though the level was still significantly lower than the SC5314 reference strain (Fig 8A). Measurements of the transepithelial electrical resistance (TEER), an indicator for barrier integrity, revealed that SC5314 decreased the overall integrity of the barrier significantly more than either oak tree strain or Evo (Figs 8B and S7B).

To further investigate the increased virulence of Evo in a more complex infection scenario, we used a previously established gut-on-chip model [55]. The C. albicans strains were individually used to infect the gut compartment by perfusing them within that compartment in the organ-on-chip. Unadhered CFUs obtained from the flow-through showed that all strains adhered to the gut-side intestinal epithelial cells to a similar degree (Figs 8C and S7C). Damage to the host cells in the gut and vascular compartments was measured by LDH release after 24 h. Surprisingly, the otherwise non-virulent Oak 1 isolate damaged both, intestinal and endothelial cells to the same extent as SC5314 in this setting (Fig 8D). The lowest damage to both the epithelial and endothelial compartment was observed for Oak 2, in stark contrast to its high pathogenicity in the earlier in vitro assays (S7D Fig). The sugar-adapted Evo, however, induced damage to the intestinal cells to the same extent as SC5314 and even showed the highest damage of all strains to the vascular cells. This again supports the previously observed increase in its virulence (Fig 8D) and suggests a galactose-driven adaptation to the gut niche, which among our models is best simulated by this organ-on-chip system.

To determine the colonization, invasion, and translocation abilities for each strain in this setting, CFUs recovered from the different compartments of the chip were plated after 24 h. Chips infected with Oak 1 or with Evo showed more CFUs in the gut flow-through. A likely reason for this is that they form more less-adherent yeasts than the more filamentous SC5314 and Oak 2 strain (Figs 8E and S7E). When the total intestinal epithelium was scraped and plated to determine the intestinal fungal burden, the majority of strains showed similar degrees of growth and invasion. Only Evo had a significantly higher fungal load in the intestinal compartment (Fig 8E).

Based on the CFUs levels on the vascular side, all strains had a similar ability to invade the endothelium and translocate. The fungal burdens of the oak tree isolates and Evo strain were even higher than that of SC5314 (Figs 8E and S7E, Vascular invasion). Although the translocation into the endothelium was similar for all strains, they differed in the ability to disseminate into the vascular medium (Figs 8E and S7E). The sugar-adapted Evo disseminated better than the Oak 1 strain, matching its increased gut- and vascular-side damage, and its increased translocation in the transwell model (Fig 8A, D, E). Overall, these data show that adaptation of an environmental isolate of C. albicans to host dietary sugars can increase its virulence potential in the intestinal niche.

Lastly, we checked whether the sugar adaptation changed C. albicans interaction with macrophages. Similar to the parental oak tree isolate, the sugar-adapted Evo strain showed an increased intracellular survival in macrophages compared to SC5314 (S8C, D Fig). Moreover, when investigating the damage induction to human macrophages, the end-point LDH assay showed that Evo still inflicted significantly less damage than SC5314, similar to its parent Oak 1 (S8A Fig). However, when measuring the host cell membrane lysis over time, the sugar-adapted Evo strain induced macrophage lysis to a similar extent than SC5314, significantly more than Oak 1 (S8B Fig).

We investigated the immune activation by the evolved strain in whole blood, with macrophages, or in the complex gut-on-chip model. For the latter, no difference in cytokine release was observed among the strains (S9C Fig). Interestingly, in whole blood, the sugar-adapted Evo induced less pro-inflammatory cytokines such as IL-8, TNFα, IL-6, and IL-1β than SC5314, and less IL-8 than its parental strain (S9A Fig). This trend, however, was inverted in infected macrophages, as Evo induced the highest release of IL-8 and TNFα (S9B Fig).

Ethics statement

Human peripheral blood was collected from healthy volunteers with ethics approval and after obtaining written informed consent. This study was conducted according to the principles of the Declaration of Helsinki. The blood donation protocol and use of blood for this study were approved by the institutional ethics committee of the University Hospital Jena (permission number 2207–01/08).

Fungal strains and culture conditions

C. albicans SC5314 [98], C. albicans oak tree isolates 1, 2, and 3 [17,20], and a sugar-adapted C. albicans strain were used for this publication (S1 Table: Strains used in this study). For all experiments single colonies were picked from Yeast Extract Peptone Dextrose (YPD) agar plates and grown overnight in liquid YPD medium in an orbital shaker at 180 rpm at 30 °C. Yeast cells were then harvested by centrifugation (20,000 g, 1 min), washed twice with phosphate-buffered saline (PBS), and the yeast cell number was adjusted.

Culture of oral and intestinal cells

The human oral epithelial cells (TR146, European Collection of Authenticated Cell Cultures ECACC #10032305) were cultivated at 37 °C and 5% CO₂ in Dulbecco’s modified Eagle medium/Nutrient Mixture F-12 (DMEM/F12, Life Technologies) supplemented with 10% fetal bovine serum (FBS) (Bio&Sell) for no longer than 15 passages. TR146 cells were seeded in 6-well plates at a concentration of 8 × 10⁵ cells/well for RNA isolation. For damage assays TR146 cells were seeded at a total concentration of 2 × 10⁴ cells/well in a 96-well plate. For invasion/adhesion/ filamentation assays 1 × 10⁵ TR146 cells/well were seeded on a coverslip in a 24-well plate. Confluent TR146 cells were washed once and subsequent experiments were carried out in serum-free medium.

The intestinal epithelial Caco-2 brush border expressing 1 cell line (C2BBe1; ATCC, CRL2102) and the human intestinal goblet cell line (HT29-MTX; ATCC, HTB-38; CLS, Lot No. 13B021) were routinely cultivated in Dulbecco’s Modified Eagle’s Medium (DMEM) (Gibco, Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (FBS) (Bio&Sell), 10 µg/ml Holotransferrin (Calbiochem, Merck), and 1% non-essential amino acids (Gibco, Thermo Fisher Scientific) at 37 °C with 5% CO₂ for no longer than 15 passages. C2BBe1 cells were seeded in 6-well plates at a concentration of 5 × 10⁵ cells/well for RNA isolation. C2BBe1 and HT29-MTX cells were seeded in 96-well plates and transwell inserts (polycarbonate membrane with 5 μm pores; Corning) at a 70:30 ratio (C2BBe1:HT29-MTX) and a total concentration of 2 × 10⁴ cells/well or insert for damage and translocation assays, respectively. For quantification of filamentation and invasion, intestinal cells were seeded in a 70:30 ratio (C2BBe1:HT29-MTX) and a total concentration of 1 × 10⁵ cells/well on coverslips. All well plates and transwell inserts were coated with collagen I (10 µg/ml for 2 h at room temperature; Thermo Fisher Scientific) and maintained for 12 d at 100% confluency for differentiation with regular medium exchange before infection. Just prior to infection with C. albicans, the medium was removed and fresh DMEM without FBS, Holotransferrin, or 1% non-essential amino acids was added to the cells.

Growth curves

Washed Candida cells were diluted to an OD_{600 nm} of 0.1. As medium, 0.67% yeast nitrogen base (YNB, BD Bioscience) was supplemented with 2% of a carbon and 0.5% of a specific nitrogen source (for amino acids 10 mM was used). For all tested carbon sources, 0.5% ammonium sulfate was used as nitrogen source and for all tested nitrogen sources, 2% glucose as carbon source. Growth was monitored in 96-well-plates by measuring the absorbance at 600 nm every 15 min for up to 72 h at 30 °C in a microplate reader (Tecan M-Plex). Prior to each measurement plates underwent ten seconds orbital shaking followed by ten seconds waiting time. The first measured OD_{600 nm} was subtracted, the remaining measured values were plotted, and the area under the curve was calculated. The experiment was repeated three times (n = 3).

Antifungal drop tests and growth curves

For the drop test, 5 µl of serial diluted C. albicans solution were dropped on YPD plates supplemented with the indicated concentrations of amphotericin B (Sigma) and fluconazole (Sigma), and were incubated at 37 °C for 6 days. For the growth curves, YPD was supplemented with either 0.5 µg/ml and 1 µg/ml of amphotericin B, or 32 µg/ml and 64 µg/ml of fluconazole and growth was measured and analyzed as described above.

Amoebae infections

Protostelium aurantium var. fungivorum [99] was cultured as described previously [100]. The amoeba trophozoites were regularly grown in standard-size petri dishes (Greiner Bio-One) in 2 mM phosphate buffer (PB) (0.8 mM KH₂PO₄, 1.2 mM K₂HPO₄, pH 6.6) with Rhodotorula mucilaginosa as a food source at 22 °C. The survival of the C. albicans oak tree strains in presence of the amoebae was determined as described previously [100]. Briefly, amoebae were grown to confluency in 20 × wMY (40 mg/ml yeast extract, 40 mg/ml malt extract, 0.75 g/l K₂HPO₄), harvested by scraping, and counted. C. albicans cultures were prepared as indicated above, washed twice in 20 × wMY, and added to the amoebae at an MOI of 10 in a 96-well plate. After an incubation step at 22 °C for 3 h, the samples were appropriately diluted in PB and plated on YPD plates. The number of surviving yeasts was calculated as a percentage of CFUs relative to the inoculum. To assess the amount of phagocytosed fungi, C. albicans and amoebae were prepared as described above and were added in a ratio of 10:1 (fungus:amoeba) in a 24-well plate. After an incubation step for 2 h at 22 °C, samples were fixed with Roti-Histofix 4%. After rinsing with PBS, amoebae were permeabilized with 0.5% Triton X-100 for 10 min. Fungal cells were stained with calcofluor white (CFW, Sigma) and after subsequent washing steps with ddH₂O, amoeba actin was stained with Phalloidin-AlexaFluor 633 (Thermo Fisher Scientific). Micrographswere taken using a Zeiss Celldiscoverer 7 (Zeiss, with LSM 900) and the number of amoebae with ingested yeasts (%) was determined.

Quantification of adhesion, invasion, hypha formation, and hyphal length

Oral cells were infected in a 24-well plate with Candida cells at a multiplicity of infection (MOI) of 1 as described above, and incubated for 1 (adhesion) or 3 h (invasion, hypha formation, hyphal length). Non-adherent Candida cells were removed by rinsing with PBS and samples were fixed with Roti-Histofix 4% (Roth). To measure adhesion, adherent Candida cells were stained with AlexaFluor647 conjugate of succinylated concanavalin A (ConA; Invitrogen) and visualized using a fluorescence microscope (Leica DM5500B, Leica DFC360 FX). Six representative images were counted, the average amount of cells per image was calculated, and the numbers were plotted relative to the reference strain (SC5314). For invasion, extracellular, non-invasive fungal components were stained with ConA. After rinsing with PBS, oral cells were permeabilized with 0.5% Triton X-100 for 10 min. The entire fungal cells (invasive and noninvasive) were stained with Calcofluor white (CFW; Sigma-Aldrich) and visualized by fluorescence microscopy. The total hyphal length was measured, and the percentage of filaments and invasive hyphae (only CFW-stained), counted from at least 100 hyphae per strain for each biological replicate.

Intestinal cells were infected at a MOI of 1 and incubated for 6 h for quantification of filamentation and invasion. Non-adherent Candida cells were removed by rinsing with PBS and samples were fixed. Extracellular hyphae were stained with CFW, and the host cells were permeabilized as stated above. Intracellular C. albicans hyphae were incubated with 25 µg/ml rabbit anti-C. albicans antibody (Acris) for 3 h at 4 °C. After rinsing with PBS, 4 µg/ml goat anti-rabbit antibody labeled with AlexaFluor488 were added, and incubated for 1 h at 37 °C.

Epithelial damage assay with oral and intestinal cells

Oral and differentiated intestinal cells were seeded in a 96-well plate as described above, infected with C. albicans (MOI 1), and incubated for 24 or 48 h. Release of the cytoplasmic enzyme lactate dehydrogenase (LDH) was measured as a marker for necrotic epithelial damage [101] using a Cytotoxicity Detection Kit (Roche) according to the manufacturer’s instructions. The LDH release was expressed as % of full lysis control (maximum LDH release induced by the addition of 0.5% Triton X-100 to uninfected epithelial cells) unless otherwise stated.

Translocation assay

Intestinal cells were seeded in inserts in a 24-well plate as described above. The infection and read-outs were performed as described previously [53]. Briefly, the inserts were infected with C. albicans (MOI 1) and incubated at 37 °C and 5% CO₂. After 24 h, the translocated hyphae were detached by treating in the lower compartment with 20 U/ml zymolyase (Amsbio) for 2 h at 37 °C and 5% CO₂. Detached hyphae were then plated on YPD agar, and the number of colony-forming units (CFUs) was determined.

The trans-epithelial electrical resistance (TEER) was determined using a volt-ohm meter (EVOM2, World Precision Instruments) after 24 h post infection.

RNA isolation

For fungal RNA extraction on epithelial cells, oral and intestinal cells were seeded in 6-wells as described above. On the day of infection, media in each well was replaced with 2 ml DMEM (intestinal) or DMEM F12 (oral cells) without FBS, infected with C. albicans cells (2 ml of 4 × 10⁵ yeast/ml in medium without FCS), and incubated at 37 °C, 5% CO₂. Samples for RNA isolation were collected at different time points: 24 and 48 h post infection. More specifically, the well content was removed and replaced with 500 µl of RNeasy Lysis (RLT) buffer (Qiagen), containing 1% β-mercaptoethanol (Roth). Cells were detached using a cell scraper (< 3 min), immediately shock-frozen in liquid nitrogen, and stored at -80 °C until further use. Collected samples were defrosted and centrifuged for 10 min (20,000 g, 4 °C). Fungal RNA was isolated from the pellet, using a freezing-thawing method, as described previously [102]. RNA concentrations were quantified using a NanoDrop 1000 Spectrophotometer (Thermo Fisher Scientific) and RNA quality was assessed with an Agilent 2100 Bioanalyzer (Agilent Technologies).

To compare ECE1 expression levels during filamentation of the oak tree isolates, C. albicans strains were adjusted to 10⁷ cells/ml in 25 ml RPMI 1640 (hypha-inducing) or 5 ml YPD (yeast-maintaining). For hyphae samples, fungal suspensions were distributed in 150 cm² petri dishes and incubated at 37 °C and 5% CO₂ for 3 h. Afterwards, medium and nonadherent Candida cells were discarded. Adherent Candida cells were washed once with ice-cold PBS, detached using a cell scraper, and collected. For yeast samples, fungal suspensions were incubated in Erlenmeyer flasks at 30 °C and 180 rpm for 3 h. The cells were collected at 4000 g for 2 min at 4 °C and resuspended in 10 ml ice-cold PBS. Both, yeast and hyphae samples, were centrifuged again at 3000 g for 2 min at 4 °C, the supernatant was discarded, and the C. albicans pellets were shock-frozen in liquid nitrogen. Thawed pellets were resuspended in 600 µl RNeasy Lysis (RLT) buffer (Qiagen) and RNA was isolated as described previously [49].

Reverse transcription-quantitative PCR (RT-qPCR)

Isolated RNA (500 ng) was treated with DNase I (Fermentas) following the manufacturer’s recommendations and subsequently transcribed into cDNA using 0.5 µg Oligo(dT)12–18 Primer, 200 U Superscript III Reverse Transcriptase and 40 U RNaseOUT Recombinant RNase Inhibitor (Thermo Fisher Scientific). Obtained cDNA was diluted 1:5 and used for qPCR with GoTaq qPCR Master Mix (Promega) in a CFX96 thermocycler (Bio-Rad). The expression levels were normalized against beta-actin. All primers used are listed in S2 Table.

RNA sequencing and transcriptional analysis

RNA sampling and isolation was performed from intestinal cell infections as described above. Library preparation and RNA sequencing were carried out at Genewiz/Azenta (Leipzig, Germany). After poly(A) filtering, mRNA was fragmented, and cDNA libraries were generated for each sample. Libraries were sequenced with 2x150 read lengths using an Illumina NovaSeq platform. Reads were aligned to the SC5314 standard genome (version A22-s07-m01-r146 [103]) using bowtie2 v2.5.2 [104] with standard settings. Transcripts were counted by Subread featureCounts v2.0.5 [105] followed by median ratio normalization.

Next, we performed gene set enrichment analysis and visualized clustering patterns of our data across different strains and conditions using R (v4.3.2). To visualize the effect of different strains and time on gene expression, first principal component analysis was performed using the prcomp function in the stats package (v4.3.2). The first two components were selected and k-means clustering with two centers was performed using factoextra package (v1.0.7). For KEGG pathway enrichment analysis, up- and down-regulated genes were defined with a log₂ fold-change threshold of ±1.5. Significantly enriched pathways (p < 0.05) were identified using the compareCluster function in the ClusterProfiler package (v4.10.1) [106]. A subset of pathways of interest was manually curated and visualized.

Ece1 detection in hyphal supernatants by LC-MS/MS analysis

Analysis of hypha-secreted Ece1 peptides was optimized for the detection of candidalysin and has been published recently [48]. Briefly, C. albicans strains were incubated for 18 h under hypha-inducing conditions. Secreted peptides were enriched from culture supernatants by solid-phase extraction, dried in a vacuum concentrator, dissolved in 0.2% formic acid in 71:27:2 (vol/vol/vol) acetonitrile (ACN)-H₂O-dimethyl sulfoxide (DMSO), and passed through a 10-kDa molecular mass cut-off filter in order to remove intact proteins. LC-MS/MS analysis was carried out on an Ultimate 3,000 nano RSLC system (Thermo Fisher Scientific) coupled to a QExactive HF mass spectrometer (Thermo Fisher Scientific). Peptide separation was performed on an Accucore C4 column (75 μm I.D. × 150 mm, 2.6 μm). Proteome Discoverer 3.0 (Thermo Fisher Scientific) and the Sequest HT algorithm were used to search against the protein database of C. albicans SC5314 (https://www.uniprot.org/proteomes/UP000000559; 2023/05/03). Mass spectra were searched for both unspecific cleavages (no enzyme) and tryptic peptides up to four missed cleavages. Precursor and fragment mass tolerances were 10 ppm and 0.02 Da, respectively. A strict false discovery rate (FDR) < 1% (Target Decoy PSM validator node) and a Sequest score (cross correlation, Xcorr) >3 were required for positive protein/peptide hits. Only rank 1 proteins and peptides of the top scored proteins were counted. Label-free protein quantification was based on the Minora algorithm of PD3.0 using the precursor abundance based on intensity and a signal-to-noise ratio >5.

Monocyte isolation from buffy coats and macrophage differentiation

Preparation of human monocyte-derived macrophages (hMDMs) was performed as described before [107] based on selection of monocytes by magnetic automated cell sorting of CD14 positive monocytes and a differentiation period of seven days. Adherent hMDMs were detached with 50 mM EDTA in PBS and seeded in 96-well plates (4 × 10⁴ hMDMs/well) in RPMI 1640 (Gibco, Thermo Fisher Scientific) supplemented with 10% FBS and 50 ng/mL M-CSF (ImmunoTools) and incubated overnight. Before infection with C. albicans, the hMDMs were washed once with PBS, and RPMI without FBS was added.

Whole blood fungal survival

Human whole blood was freshly drawn from healthy volunteers in anticoagulation tubes with recombinant Hirudin (Sarstedt). C. albicans overnight cultures were prepared as described above, 10⁶ yeasts were added to 490 µl whole blood, and incubated at 37 °C under gentle rolling of the tubes. Samples were taken after 15 min, 30 min, 60 min, 120 min, and 240 min and plated on YPD in appropriate dilutions in PBS. The plates were incubated at 30 °C and colony-forming units (CFUs) were determined after 2 d to assess fungal survival.

Macrophage infections

C. albicans strains were prepared as described above and hMDMs were infected with a MOI of 5. To assess damage, LDH release from the macrophages was measured with a Cytotoxicity Detection Kit (Roche) after 24 h of infection. To determine intracellular fungal survival 3 h post infection, the supernatant was removed and macrophages were lysed by adding 200 µl ddH₂O, scraping the well, and pipetting up and down to break the cells. This was repeated with each well five times. The supernatant and lysate were each diluted with PBS and plated on YPD plates to determine the CFUs. To evaluate the morphology of the C. albicans oak tree isolates inside the macrophages, micrographs were taken using a Zeiss Celldiscoverer 7 (Zeiss, with LSM 900) after 3 h of infection.

Micro-evolution experiment

C. albicans oak tree isolate 1 (NCYC 4144) and the reference strain SC5314 were prepared as described above. The washed overnight cultures were diluted 1:500 in the YNB + 0.5% ammonium sulfate + 2% galactose, and incubated at 180 rpm and 30 °C. The strains were passaged every two to three days into fresh medium for 6 weeks. Growth of the strains was measured as OD_{600 nm} using Aquila biolabs sensor plates (Aquila biolabs GmbH). After each passage, 750 µl of the cultures were mixed with 1 volume of 60% glycerol, shock-frozen in liquid nitrogen, and stored at -80 °C for further analysis. The sugar-adapted Evo strain used in this study was collected with the last passage.

Genomic DNA isolation and whole genome sequencing

For DNA extraction, C. albicans cultures were grown in YPD for 16 h at 30 °C and 180 rpm. The cultures were centrifuged for 5 min at 4000 rpm, resuspended in 1 ml ddH₂O, and centrifuged for 2 min at 13000 rpm. The pellet was resuspended in 1 ml lysis buffer (1 M sorbitol, 100 mM sodium citrate, 50 mM EDTA, 0.6 mg/ml lyticase, 2.5% β-mercaptoethanol), incubated at 37 °C for 45 min, and centrifuged for 5 min at 13000 rpm. The pellet was resuspended in 800 µl of proteinase buffer (10 mM Tris-Cl, 50 mM EDTA, 0.5% SDS, 1 mg/ml proteinase K) and incubated at 60 °C for 30 min. 800 µl of phenol:chloroform:isoamylalcohol (25:24:1) were added, the samples were vortexed for 4 min, and centrifuged for 3 min at 13000 rpm. The aqueous phase was transferred to a fresh tube, 700 µl ice-cold isopropanol were added, and the precipitated DNA was pelleted by centrifugation for 5 min at 13000 rpm. The pellets were washed with 70% ethanol and centrifuged for 2 min at 13000 rpm. After air-drying, the pellet was resuspended in water supplemented with RNase. The gDNA was stored at -20 °C until sequencing. The sequencing was carried out by Genewiz/Azenta (Leipzig, Germany) with 2 × 150 read lengths and 10 M raw paired-end reads using an Illumina NovaSeq platform. Whole genome sequences were used for variant detection followng the best practices workflows as outlined by the GATK Team (gatk.broadinstitute.org, [108]). Briefly, the reads were pre-processed using Picard v3.1.1 [109] to remove any adapter sequences. Reads were then aligned to the NCYC4144 haploid genome [110] using the bwa v0.7.17 BWA-MEM algorithm (with parameters -M -t 11 -B 3 -p). Duplicate reads were removed using Picard in MarkDuplicates mode and the resulting SAM files sorted and indexed. Variants were detected by comparison to the re-sequenced NCYC4144 strain using GATK v4.5.0.0 with mutect2, followed by FilterMutectCalls with standard settings and SelectVariants to exclude the filtered variants.

Ece1 amino acid sequence analysis

To compare the sequences of the Extent of cell elongation 1 polyprotein in the oak tree C. albicans strains and in the sugar-adapted Evo strain, DNA sequences were extracted from whole genome data or for SC5314 from candidagenome.org and the amino acid sequence was predicted using Expasy (Swiss Institute of Bioinformatics). The amino acid sequences were aligned with Clustal Omega (European Bioinformatics Institute). Asterisks indicate identical amino acids and dots structurally similar amino acids. Similarly, promoter regions 400 bp upstream of the ATG from ECE1 were retrieved either from candidagenome.org or the WGS and aligned with Clustal Omega.

Gut-on-chip infection model

The gut-on-chip model was performed as described previously [55,111]. Briefly, biochips were prepared, assembled [111], and infected with 1250 fungal cells per chip under linear perfusion. After 10 min, the intestinal compartment was switched to linear perfusion with fresh medium. The unadhered fungal cells were collected for a total of 50 min and plated as a measurement of adhesion to the gut compartment. After 24 h post infection, damage induction, membrane permeability, and fungal burden were measured as described before [111]. For cytokine levels, supernatants were collected from vascular and intestinal compartments 24 h post infection, centrifuged, and stored at -20 °C.

Cytokine release quantification

Supernatants of infected primary macrophages (MOI 5) were collected 24 h post infection. Whole blood supernatants were collected 4 h post infection after centrifuging samples for 10 min at 500 g. Gut-on-chip supernatants were collected as described above. The release of IL-8, IL-6, Il-1β, and TNFα was measured using commercially available human enzyme-linked immunosorbent assay kits (R&D Systems) according to the manufacturer’s protocols.

Statistics and reproducibility

Experiments were performed in at least biological triplicates (n = 3). For experiments using primary macrophages or whole blood, at least 4 different donors were used. Data were analyzed using GraphPad Prism 9.5.1 (GraphPad Software, La Jolla California USA). Values are presented as mean ± standard deviation (SD). Statistical tests are indicated in each figure legend. When applicable, statistical comparison was always performed between the parent strain Oak 1 and the sugar-evolved Evo strain. Statistical significance is indicated in the figures as follows: *, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001; ****, p ≤ 0.0001.