Epithelial hypoxia maintains colonization resistance against Candida albicans

Abstract

Antibiotic treatment promotes the outgrowth of intestinal Candida albicans, but the mechanisms driving this fungal bloom remain incompletely understood. We identify oxygen as a resource required for post-antibiotic C. albicans expansion. C. albicans depleted simple sugars in the ceca of gnotobiotic mice but required oxygen to grow on these resources in vitro, pointing to anaerobiosis as a potential factor limiting growth in the gut. Clostridia species limit oxygen availability in the large intestine by producing butyrate, which activates peroxisome proliferator-activated receptor gamma (PPAR-γ) signaling to maintain epithelial hypoxia. Streptomycin treatment depleted Clostridia-derived butyrate to increase epithelial oxygenation, but the PPAR-γ agonist 5-aminosalicylic acid (5-ASA) functionally replaced Clostridia species to restore epithelial hypoxia and colonization resistance against C. albicans. Additionally, probiotic Escherichia coli required oxygen respiration to prevent a post-antibiotic bloom of C. albicans, further supporting the role of oxygen in colonization resistance. We conclude that limited access to oxygen maintains colonization resistance against C. albicans.

Keywords: Candida albicans; Clostridia; Escherichia coli; colonization resistance; epithelial hypoxia.

Figure 1:. Metabolic footprinting identifies resources that support C. albincans growth in the gastrointestinal tract of mice.

(A) Schematic of steps involved in metabolic footprinting. (B) Workflow of metabolomics data analysis with the reactive web application omuShiny. (C-E) Comparative analysis of the cecal metabolome of germ-free mice (n = 6), and gnotobiotic mice three days after engraftment with C. albicans (strain ATCC28367) (n = 4). (C) Principal component analysis of the cecal metabolome in the indicated groups of mice. Ovals indicate the 95% confidence interval. (D) Volcano plot of metabolites colored by class. The Y-axis shows the decadic logarithm of the false discovery rate (FDR)-corrected P value. The dashed line is set at an FDR corrected P value of 0.05. Metabolites with a negative fold-change value decreased in mice engrafted with C. albicans compared to germ-free mice, while metabolites with a positive fold-change value increased. (E) The graphs show the abundance of the indicated sugars in mice engrafted with C. albicans compared to germ-free mice. (C and E) Each dot represents data from one animal. (F) In vitro growth of C. albicans (strain ATCC28367) in yeast nitrogen broth supplemented with cysteine and the indicated sugars under aerobic (left panels) or anaerobic (right panels) growth conditions. Data are represented as mean ± SEM. GC-MS, gas chromatography time-of-flight mass spectrometry; KEGG, Kyoto Encyclopedia of Genes and Genomes; **, P < 0.005. See also Figure S1 and Table S1.

Figure 2:. Treatment with streptomycin depletes Clostridia and short-chain fatty acids.

Groups (n = 8) of special pathogen-free C57BL/6J mice were started on chow supplemented with 5-aminosalicylic acid (5-ASA: +) or control chow (5-ASA: −) and four days later mice received a single dose of streptomycin by oral gavage (Strep: +) or were mock-treated (Strep: −). Feces were collected for analysis one day after streptomycin treatment. (A) The absolute abundance of Clostridia in fecal samples was determined by real-time PCR using class specific primers. (B-D) Concentrations of acetate (B), propionate (C) or butyrate (D) in fecal contents were determined by gas chromatography-mass spectrometry. Data are represented as mean ± SEM. ***, P < 0.0005. See also Table S2

Figure 3:. E. coli requires aerobic respiration to restore colonization resistance against C. albincans after streptomycin treatment.

(A) Groups (n = 4) of special pathogen-free C57BL/6J mice were either given a single dose of streptomycin (20 mg/animal) by oral gavage (Strep: +) or were mock treated (Strep: −) and were challenged 48 hours later with the indicated doses of C. albicans (strain ATCC28367). Shown are geometric means (bars) +/− geometric standard deviation (error bars) of C. albicans colony-forming units (CFU) recovered per gram (g) of cecal contents. (B) Special pathogen-free C57BL/6J mice were infected with the indicated dose of C. albicans (strain ATCC28367) and were either mock-treated (n = 4) or given a single dose of streptomycin (20 mg/animal) by oral gavage five days after C. albicans infection (n = 8). Shown are geometric means (symbols) +/− geometric standard deviation (error bars) of C. albicans CFU recovered per gram of feces at the indicated time points. (C-E) Special pathogen-free C57BL/6J mice were either mock treated (−) or given a single dose of streptomycin (20 mg/animal) by oral gavage (+). The next day, mice were either mock inoculated, inoculated with 10⁹ CFU of E. coli Nissle 1917 (WT) or 10⁹ CFU of an E. coli Nissle 1917 cydA mutant (cydA). One day later, mice were challenged with 10⁵ CFU of C. albicans (strain ATCC28367) and organs were collected two days later. (C and D) Shown are geometric means (bars) +/− geometric standard deviation (error bars) of C. albicans CFU recovered one day after challenge from feces (C) or two days after challenge from cecal contents (D). (E) Shown are geometric means (bars) +/− geometric standard deviation (error bars) of E. coli CFU recovered from feces or cecal contents at the indicated time points relative to C. albicans challenge. (A and C-E) Each symbol represents data from one animal. NS, P > 0.05; *, P < 0.05; **, P < 0.005; ***, P < 0.0005. See also Figure S1.

Figure 4:. 5-ASA restores colonization resistance against C. albicans after streptomycin treatment.

(A) Mice maintained on control chow or chow supplemented with 5-aminosalicylic acid (5-ASA) were challenged with 10⁶ colony-forming units (CFU) of C. albicans (strain ATCC28367). Three days later, mice were mock-treated or received a single dose of streptomycin (Strep) by oral gavage. Shown are geometric mean +/− geometric standard deviation of C. albicans CFU recovered from feces in each group (n = 6 for strep + 5-ASA, n = 3 for remaining groups) at the indicated time points. (B-E) Mice were started on chow supplemented with 5-aminosalicylic acid (5-ASA: +) or control chow (5-ASA: −) and two days later mice received a single dose of streptomycin by oral gavage (Strep: +) or were mock-treated (Strep: −). Each symbol represents data from one animal (i.e., n for each group is provided by the number of symbols). (B, D, and E) Mice were challenged with 10⁵ C. albicans CFU (strain ATCC28367) one day after streptomycin treatment and samples were collected the next day. (C) Mice were challenged with 10⁶ C. albicans CFU (strain ATCC28367) two days after streptomycin treatment and C. albicans CFU in cecal contents were determined seven days later. (B and C) Shown are geometric means (bars) +/− geometric standard deviation (error bars) of C. albicans CFU recovered from cecal contents. (D) Sections from the cecum were blinded and scored by a veterinary pathologist. (E) RNA was extracted from preparations of colonic epithelia cells. Transcript levels of Lcn2, Tnfa, Il17a, and Ifng were determined by quantitative real-time PCR. Shown are geometric means (bars) +/− geometric standard deviation (error bars) of fold-changes in transcript levels of Lcn2, Tnfa, Il17a, or Ifng compared to transcript levels in mock-treated mice on control chow. NS, P > 0.05; *, P < 0.05; **, P < 0.005. See also Figure S2 and Table S4

Figure 5:. 5-ASA restores epithelial hypoxia in the large intestine after streptomycin treatment.

Mice were started on chow supplemented with 5-aminosalicylic acid (5-ASA: +) or control chow (5-ASA: −) and two days later mice received a single dose of streptomycin by oral gavage (Strep: +) or were mock-treated (Strep: −). Mice were challenged with 10⁵ C. albicans CFU (strain ATCC28367) one day after streptomycin treatment. One day after challenge, mice were injected with pimonidazole HCl (PMDZ) and the cecum was collected one hour later. (A-C) Binding of pimonidazole was detected using hypoxyprobe-1 primary antibody and a Cy-3 conjugated goat anti-mouse secondary antibody (red fluorescence) in histological sections from the cecum that were counterstained with DAPI nuclear stain (blue fluorescence). (A) Representative images are shown. Scale bar is 50 μm. (B) Pimonidazole staining was quantified by measuring mean PMDZ intensities from the lumen (distance of 0.0 arbitrary units) to the border of the colonocytes (distance of 0.1), and into the tissue. (C) The graph shows the peak PMDZ intensity for each mouse (symbols) and the mean peak intensity for each group (lines) (n = 7–8). (D) Butyrate concentrations in cecal contents were determined by gas chromatography-mass spectrometry (n = 4). Data are represented as mean ± SEM. NS, P > 0.05; **, P < 0.005. See also Table S2

Figure 6:. 5-ASA can functionally replace Clostridia species to restore colonization resistance against C. albicans after streptomycin treatment.

Germ-free Swiss Webster mice received a cecal microbiota transplant from streptomycin-treated C57BL/6J mice (Strep) or from mock-treated C57BL/6J mice (Mock). Seven days later, some mice were inoculated with a community of human Clostridia isolates (C17), received a second cecal microbiota transplant from mock-treated C57BL/6J mice (Mock), or were switched to chow supplemented with 5-aminosalicylic acid (5-ASA: +). Seven days later, all mice were challenged with 10⁵ C. albicans CFU (strain ATCC28367) and organs were collected one week later. (A) Schematic of the experimental timeline. (B) Shown are C. albicans CFU recovered 7 days after challenge from cecal contents of each animal (symbols; n = 3 for Strep + Mock, n = 6 for Strep + Mock and Strep + C17, n = 8–9 for remaining groups) and geometric means (bars) +/− geometric standard deviation (error bars). (C and D) Blinded sections from the cecum of each animal were scored by a veterinary pathologist. (C) Representative images of colonic sections for each group are shown. Scale bar is 100 μm. (D) Average histopathology score for each group. (E-G) Mice were injected with pimonidazole HCl (PMDZ) and the cecum was collected one hour later. Binding of pimonidazole was detected using hypoxyprobe-1 primary antibody and a Cy-3 conjugated goat anti-mouse secondary antibody (red fluorescence) in histological sections from the cecum that were counterstained with DAPI nuclear stain (blue fluorescence). (E) The graph shows the peak PMDZ intensity for each mouse (symbols) and the mean peak intensity for each group (lines). (F) Pimonidazole staining was quantified by measuring mean PMDZ intensities from the lumen (distance of 0.0 arbitrary units) to the border of the colonocytes (distance of 0.1), and into the tissue. (G) Representative images are shown. Scale bar is 50 μm. NS, P > 0.05; ***, P < 0.0005; ****, P < 0.00005. See also Table S3