Candida albicans Morphogenesis Programs Control the Balance between Gut Commensalism and Invasive Infection

Abstract

Candida albicans is a gut commensal and opportunistic pathogen. The transition between yeast and invasive hyphae is central to virulence but has unknown functions during commensal growth. In a mouse model of colonization, yeast and hyphae co-occur throughout the gastrointestinal tract. However, competitive infections of C. albicans homozygous gene disruption mutants revealed an unanticipated, inhibitory role for the yeast-to-hypha morphogenesis program on commensalism. We show that the transcription factor Ume6, a master regulator of filamentation, inhibits gut colonization, not by effects on cell shape, but by activating the expression of a hypha-specific pro-inflammatory secreted protease, Sap6, and a hyphal cell surface adhesin, Hyr1. Like a ume6 mutant, strains lacking SAP6 exhibit enhanced colonization fitness, whereas SAP6-overexpression strains are attenuated in the gut. These results reveal a tradeoff between fungal programs supporting commensalism and virulence in which selection against hypha-specific markers limits the disease-causing potential of this ubiquitous commensal-pathogen.

Figure 1.. Activators of C. albicans filamentation inhibit gut commensalism.

A) Schematic of commensalism screens. ~650 C. albicans mutants and wild type (WT, SN250) were gavaged into BALB/c mice (n= 2-3 animals/inoculum). Feces was sampled after 3-5 and 10 days, and strains recovered by plating. Sequencing libraries were prepared from genomic DNA abutting the disrupted ORFs (mutants) or C.a.LEU2 gene (WT). Competitive indexes (CI) were calculated as the log2 function of (R/I), based on samples recovered from the host (R) and the inoculum (I). See also Tables S1, S2, and S3. B) Results from one animal after 10 days. efg1 (red) outcompeted WT (black) and all other mutants (gray). C) Hypercompetitive mutants were recovered from three commensalism screens. Screen 1 tested all mutant and WT; Screen 2 excluded efg1; Screen 3 excluded efg1, brg1, and rob1. Each data points indicate the CI for the indicated strain in one animal. Bars indicate the mean. D) Hyphal gene regulatory circuit (note that not all regulators are shown). Efg1, Brg1, Rob1, Tec1, and Ume6 activate expression of genes required for morphogenesis and genes that are upregulated in hyphae. Efg1 has been shown to activate UME6 by epistasis analysis and ChIP-Seq (Banerjee et al., 2013; Zeidler et al., 2009). Brg1 and Tec1 have been shown to bind to the UME6 promoter by ChIP-Seq (Childers and Kadosh, 2015; Nobile et al., 2012).

Figure 2.. efg1, brg1, rob1, tec1, and ume6 exhibit enhanced commensal fitness, while UME6^OE has reduced fitness.

A) Schematic of competition experiments. Mice were gavaged with 1:1 mixtures of WT and each commensalism mutant, strain abundance in feces was monitored for ≥25-days using qPCR. B-G) Results for: (B) WT (ySN226) vs. efg1 (ySN119), (C) WT (ySN425) vs. brg1 (ySN1180), (D) WT (ySN250) vs. rob1 (ySN1440), (E) WT (ySN250) vs. tec1 (ySN1442), (F) WT (ySN250) vs. ume6 (ySN1479), (G) WT (ySN1556) vs. UME6^OE (ySN1558). Bars represent the mean. Significance was determined using the paired student’s t-test: n.s. not significant, *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001. Additional results are presented in Figures S1, S2, and S3.

Figure 3.. C. albicans colonizes the gut as a mixed population of yeast and hyphae.

A)Examples of FISH-stained wild-type C. albicans (SN425 or SN250) yeast and hyphae in the mouse cecum (left) and in vitro (Spider medium, 37°C, right). A Cy3-coupled fungal -specific oligonucleotide (red) was hybridized to C. albicans 28S rRNA. Yellow arrowheads indicate hyphae. Scale bar denotes 15 μm. B) WT (SN250) and ume6 (SN1479, SN1478) in different GI compartments after 10 days of colonization. “Epithelium” indicates cross sectional region near the mucosa, “Lumen” indicates region near the center of the GI compartment. Mucus (green) was stained with a FITC-coupled lectin (UEA-1 +/− WGA-1), host epithelial cell nuclei (blue) were stained with DAPI. Scale bar indicates 20 μm. C) Quantification of C. albicans yeast and hyphae. 30 fields of view were scored per GI compartment per animal colonized with WT (n=6 animals) and ume6 (n=9 animals). Data represent the mean ±SEM. Significance was determined using the unpaired student’s t-test. See Figure S4 for ume6 morphology in vitro and Figure S5 for FISH staining of other transcription factor mutants in the gut.

Figure 4.. NanoString reveals induction of hypha-associated and pH-responsive genes in commensally propagated C. albicans.

A) Schematic of NanoString experiment. Animals were colonized with wild type (ySN425) for 1, 4 and 10 days. RNA recovered from stomachs, ceca, and large intestines was analyzed with 182 NanoString primer sets. B) Heatmap of expressed hypha-associated genes. Upregulation relative to the inoculum is indicated in yellow, downregulation in blue. Final two columns indicate the −log₁₀(adjusted p-value) of expression differences between the gut vs. laboratory conditions (Gut/YPD) or the stomach vs. distal compartments (Stomach/Distal). Significance was determined using a linear fit model. C) Heatmap of expressed pH-associated genes. The full NanoString dataset appears in Table S2.

Figure 5.. mRNA-Seq reveals Ume6 regulation of a subset of hypha-associated genes in the gut.

A) Schematic of mRNA-Seq experiment. PolyA RNA was recovered from WT C. albicans (ySN250) and ume6 (ySN1479) propagated under standard in vitro conditions (YPD, 30°C; n=3 cultures), filament-inducing conditions (YPD+10% serum, 37°C; n=3 cultures), or for 10 days in the GI commensalism model (n=5 animals). B) Analysis of sequencing depth and coverage. Bars represent % reads that align to C. albicans vs. mouse transcriptomes. Green numbers denote the read count (in millions) of indicated strains under each condition. C) Heat map of hypha-associated gene expression under laboratory conditions and in the host digestive tract. Values represent the log₂ function of the tpm under the indicated condition divided by the tpm under standard in vitro conditions (YPD, 30°C). Final two columns represent the significance [−log₁₀(adjusted p-value)] of expression differences between wild type propagated in the gut vs. in vitro (Gut/YPD) and ume6 vs. WT when both are propagated in the gut (ume6/WT). D) Volcano plot depicting the log2 transformed ratio of hypha-associated gene expression in commensally propagated ume6 vs. wild type (x-axis) versus significance (y-axis). Full mRNA-Seq datasets are presented in Table S4.

Figure 6.. Hypha-associated secreted and cell surface proteins inhibit GI commensalism.

The commensal fitness of mutants affecting SAP6 and HYR1 was determined in 1:1 competition with wild-type C. albicans, as in Figure 2. A) WT (SN250) vs. sap6 (SN1664); B) WT (SN250) vs. hyr1 (SN511); C) WT (SN250) vs. sap6+SAP6 gene addback strain (SN1796, note that a single copy of SAP6 was restored to a sap6Δ/sap6Δ strain lacking both natural alleles); D) WT (SN235) v. SAP6^OE (SN1798). Significance was determined using the paired student’s t-test; ns not significant, *p<0.05, **p<0.01, ***p<0.001. E) sap6 has normal morphology in the gut. sap6 (SN1664 and m886) was visualized using FISH (left panels), as in Figure 3. The plot on the right indicates the mean percentages (±SEM) of hyphae in stomachs, small intestines, ceca, and large intestines of six animals after ten days. SAP6 expression levels in transcription factor mutants is shown in Figure S6.