Malassezia Is Associated with Crohn's Disease and Exacerbates Colitis in Mouse Models

Abstract

Inflammatory bowel disease (IBD) is characterized by alterations in the intestinal microbiota and altered immune responses to gut microbiota. Evidence is accumulating that IBD is influenced by not only commensal bacteria but also commensal fungi. We characterized fungi directly associated with the intestinal mucosa in healthy people and Crohn's disease patients and identified fungi specifically abundant in patients. One of these, the common skin resident fungus Malassezia restricta, is also linked to the presence of an IBD-associated polymorphism in the gene for CARD9, a signaling adaptor important for anti-fungal defense. M. restricta elicits innate inflammatory responses largely through CARD9 and is recognized by Crohn's disease patient anti-fungal antibodies. This yeast elicits strong inflammatory cytokine production from innate cells harboring the IBD-linked polymorphism in CARD9 and exacerbates colitis via CARD9 in mouse models of disease. Collectively, these results suggest that targeting specific commensal fungi may be a therapeutic strategy for IBD.

Keywords: CARD9; Crohn disease; Malassezia; c-type lectin; mycobiome.

Figure 1.. Crohn’s disease mucosae-associated mycobiome characterization

(A) Sequences from healthy control and Crohn’s disease sigmoid colon and cecum were identified to the genus level, and the relative representation of each genus in the datasets is illustrated. “Others” include 67 relatively rare genera, and “Unmapped” indicates the fraction of the sequences that could not be convincingly identified as belonging to specific genera (defined as ≥97% identity to a known reference sequence). (B) The data are shown broken down by individual patient (x-axis), showing the relative abundance of sequences detected in each sample. (C) MaAsLin analysis reveals that Crohn’s disease (CD) is associated with a decrease in ascomycetes and an increase in basidiomycetes compared to healthy controls (HC). Notched box plots show individual arcsine square-root transformed relative abundance’s median and confidence intervals as well as the first and third quartiles. The strength of the difference in the CD samples is noted by the brightness (lower in CD – green, higher in CD – red). (D) Specific genera and their associations with CD (all patients) are illustrated. (E) Analysis as in (D) but illustrating associations specific to the indicated types of Crohn’s disease (ileocolonic or colonic). See also Figure S1 and Tables S1–S4.

Figure 2.. The presence of Malassezia is linked to the Crohn’s disease CARD9 risk allele

(A) Presence or absence of the CD-associated variant of CARD9 was assessed for association with fungal genera. Fungal genera approaching or exceeding statistical significance (Bonferroni-corrected 6.7×10⁻⁴) are listed. (B) Box-and-whisker plots showing relative abundance of Malassezia and Pichia in sigmoid colon CD samples (all patients, left, and patients with ileocolonic CD, right) according to the CARD9^S12N SNP genotype. (*, p<0.05; **, p<0.01; one-way ANOVA with Tukey’s multiple comparison test). (C) Serum samples were selected from a biobank of CD patient sera previously characterized as “High” or “Low” for ASCA IgG or IgA (n=18–22 per group) and screened by flow cytometry for IgA and IgG reactivity against Malassezia restricta. (Mann-Whitney U Test) Notches indicate 95% confidence interval and whiskers extend no further than 1.5 times IQR from the hinge.

Figure 3.. M. restricta exacerbates colitis in mice

(A, B) Colon length upon termination of experiment in which mice were gavaged with the indicated live yeast and exposed to DSS in their drinking water for 7 days and then without DSS for 5 days (n=5/group). (C) Disease activity over duration of experiment. (D) Fecal lipocalin-2 levels were measured by ELISA on day 6. (E, F) Representative H&E-stained colon sections of an experiment terminated on day 7 (E) and histological assessment of disease severity (F). (G, H) Percentage of IL-17A- and IFN-γ-producing colonic lamina propria CD4⁺ T cells was determined upon sacrifice (day 12). (*, p<0.05; **, p<0.01; ***, p<0.001; one-way ANOVA with Tukey’s multiple comparison test). Each dot represents an individual mouse. Data are representative of three independent experiments. See also Figures S2 and S3.

Figure 4.. M. restricta is sufficient to exacerbate colitis in gnotobiotic mice

(A) Detection of bacterial and fungal rDNA by quantitative PCR in feces of specific pathogen-free (SPF), germ-free (GF), altered Schaedler flora-colonized (ASF) mice, and ASF mice gavaged with M. restricta (n=5/group). (B) Levels of the 8 ASF bacteria were assessed by quantitative PCR of 16s rDNA before and after exposure to M. restricta. Each column is a different mouse as labeled (C, D) Colon length upon termination of experiment in which fungal-free altered Schaedler flora (ASF) mice were gavaged with live yeast and exposed to DSS in their drinking water. (E) Disease activity over duration of experiment. (F) Fecal lipocalin-2 levels were measured by ELISA on day 5. (G, H) Percentage of IL-17A- and IFN-γ-producing colonic lamina propria CD4⁺ T cells was determined upon sacrifice. (*, p<0.05; **, p<0.01; ***, p<0.001; one-way ANOVA with Tukey’s multiple comparison test). Each dot represents an individual mouse. See also Figure S4.

Figure 5.. M. restricta elicits a strong inflammatory response from myeloid phagocytes

(A) C. albicans, S. cerevisiae, and M. restricta were grown in liquid culture and imaged by differential interference contrast microscopy. Diameters (n>150 each) were measured using ImageJ. (B) M. restricta, C. albicans or S. cerevisiae were fixed (killed) in paraformaldehyde, and human dendritic cells were exposed to killed yeasts at the indicated multiplicities of infection (MOI), or to E. coli lipopolysaccharide (LPS, 100 ng/ml) for 24 hours. TNF-α in culture supernatants was measured by ELISA. (C) Mouse bone marrow-derived dendritic cells were stimulated as in (B) and TNF-α and IL-6 levels were measured in culture supernatants. (D) Mouse bone marrow-derived macrophages were stimulated as in (B) and TNF-α and IL-6 levels were measured in culture supernatants. (E) Mouse bone marrow-derived dendritic cells were stimulated with fungi as in (B) and expression of CD86 was assessed by flow (MOI 10). (F, G) Mouse bone marrow-derived dendritic cells were stimulated with the indicated fungi in the presence of anti-CD3ε anti-bodies and co-cultured with naïve CD4⁺ T cells. Production of IL-17A and IFN-γ by CD4⁺ cells was assessed by flow cytometry. (*, p<0.05; **, p<0.01; ***, p<0.001; one-way ANOVA with Tukey’s multiple comparison test). Data are representative of three independent experiments.

Figure 6.. Innate inflammatory responses to M. restricta are CARD9/C-type lectin dependent and are enhanced by the Crohn’s disease-associated CARD9 polymorphism

(A) Mouse bone marrow-derived dendritic cells from wild-type (WT) or Card9^−/− mice were stimulated with fixed M. restricta at the indicated multiplicities of infection (MOI), or to E. coli lipopolysaccharide (LPS, 100 ng/ml) for 24 hours. TNF-α and IL-6 levels were measured in culture supernatants. (B) Neutrophils from wild type (WT) or Card9^−/− (KO) mice were purified from bone marrow and stimulated with C. albicans or M. restricta yeast (MOI 5, fixed in paraformaldehyde) or E. coli lipopolysaccharide (LPS, 100 ng/ml) for 24 hours. Cytokines in culture supernatants were measured by ELISA. (C) Mouse bone marrow-derived dendritic cells from wild-type (WT) or the indicated knockout (KO) mice were stimulated with M. restricta yeast (MOI 5, fixed in paraformaldehyde), E. coli lipopolysaccharide (LPS, 100 ng/ml), or zymosan (30 μg/ml) as indicated for 24 hours. TNF-α and IL-6 levels were measured in culture supernatants. (D) Human peripheral blood-derived dendritic cells were prepared from healthy donors determined to be homozygous for the CARD9^S12N A (risk) or G (protective) alleles. Cells were stimulated (MOI 10) with the indicated yeasts or LPS (100 ng/ml) for 24 hours and production of TNF-α, IL-8, IL-1β, IL-10, and IL-6 (n=6–15) was measured. Each dot is an individual patient (measured in duplicate), and the boxes indicate means and standard deviations. (*, p<0.05; **, p<0.01; ***, p<0.001; one-way ANOVA with Tukey’s multiple comparison test). Data in A and B are representative of three independent experiments.

Figure. 7.. In vivo effects of M. restricta require CARD9

(A, B) Colon length upon termination of experiment in which Card9^−/− mice or wild type littermates were gavaged with the indicated live yeast and exposed to DSS in their drinking water for 7 days and then without DSS for 5 days (n=6–7/group). (C, D) Disease activity over duration of experiment. (E) Fecal lipocalin-2 levels were measured by ELISA on day 6. (*, p<0.05; **, p<0.01; ***, p<0.001; one-way ANOVA with Tukey’s multiple comparison test). Each dot represents an individual mouse. Data are representative of at least two independent experiments.