Extracellular vesicles from diverse fungal pathogens induce species-specific and endocytosis-dependent immunomodulation

Abstract

Microbial pathogens generate extracellular vesicles (EVs) for intercellular communication and quorum sensing. Microbial EVs also induce inflammatory pathways within host innate immune cells. We previously demonstrated that EVs secreted by Candida albicans trigger type I interferon signaling in host cells specifically via the cGAS-STING innate immune signaling pathway. Here, we show that despite sharing similar properties of morphology and internal DNA content, the interactions between EVs and the innate immune system differ according to the parental fungal species. EVs secreted by C. albicans, Saccharomyces cerevisiae, Cryptococcus neoformans, and Aspergillus fumigatus are differentially endocytosed by murine macrophages triggering varied cytokine responses, innate immune signaling, and subsequent immune cell recruitment. Notably, polysaccharide and hydrophobic protein structures on the outer layers of C. neoformans and A. fumigatus EVs inhibit efficient internalization by macrophages and dampen innate immune activation. Our data uncover the functional consequences of the internalization of diverse fungal EVs by immune cells and reveal novel insights into the early innate immune response to distinct clinically significant fungal pathogens.

Fig 1. Murine macrophage internalization, cytokine production, and neutrophil recruitment differ in responses to different fungal EVs.

A, Percent internalization of Ca EVs, Sc EVs, Cn EVs, and Af EVs by murine macrophages. Significance assessed by an ordinary one-way ANOVA and Dunnet’s multiple comparisons test, ***p = 0.0002 vs PBS treated macrophages. (5x10¹⁰ EVs/mL added per stimulation), n = 3. B, Percent endocytosis of Ca EVs and Sc EVs with murine macrophages treated with DMSO or 100µM dynasore. Significance assessed by a two-way ANOVA and Tukey’s multiple comparisons test, ****p= < 0.0001 vs respective DMSO controls. (5x10¹⁰ EVs/mL added per stimulation), n = 2. C, Induction of MCP-1 (in pg/mL) in WT and Sting^-/- macrophages stimulated with PBS, 2.5µg cGAMP, Ca EVs, Sc EVs, Cn EVs, and Af EVs (1x10¹⁰ EVs/mL added per stimulation). Significance assessed by an unpaired t-test, *p ≤ 0.05, **p ≤ 0.009 vs PBS treated WT macrophages, n = 2. D, Number of neutrophils that transmigrated towards supernatants from WT macrophage stimulated by PBS, 2.5µg cGAMP, Ca EVs, Sc EVs, Cn EVs, and Af EVs (5x10¹⁰ EVs added per stimulation), and 0.1µM fMLP as a positive control. Significance assessed by an ordinary one-way ANOVA and Dunnet’s multiple comparisons test, *p ≤ 0.0203, **p ≤ 0.0073, ****p < 0.0001 vs PBS-treated macrophages, n = 3.

Fig 2. Fungal EVs differentially activate the STING pathway.

A, Representative immunoblots of viperin, phosphorylated IRF3, total IRF3, phosphorylated TBK1, total TBK1, and actin in response to PBS, transfected cGAMP (2.5µg), Ca EVs, Sc EVs, Cn EVs, and Af EVs (5x10¹⁰ EVs/mL added per stimulation) in WT macrophages. B, Densitometry of viperin, p-IRF3, and p-TBK1 relative to appropriate loading controls. Significance assessed by a two-way ANOVA and Dunnet’s multiple comparisons test, *p ≤ 0.0497, **p = 0.0068, ***p = 0.0007, ****p < 0.0001 vs PBS treated macrophages in each respective group, n = 3. C, Immunoblot of viperin in WT macrophages treated with either DMSO (vehicle control) or 100µM dynasore and stimulated with Ca EVs, Sc EVs, Cn EVs, and Af EVs (5x10¹⁰ EVs/mL added per stimulation).

Fig 3. Fungal EVs induce translocation of cGAS from the nuclear membrane to the cytosol.

A, Representative images of the localization of cGAS (green) when cGAS-GFP expressing macrophages are stimulated for 3h with DiI-labeled (red) PBS, Ca EVs, Sc EVs, Cn EVs, or Af EVs (8x10⁹ EVs added per stimulation). Scale bar = 10µm. B, Semi-quantitative analysis of the percentage of cGAS localization that is either nuclear or non-nuclear. cGAS localization analyses were performed on approximately 100 cGAS-GFP expressing macrophages that successfully endocytosed DiI-labeled EVs.

Fig 4. Physical and DNA characteristics of Ca EVs, Sc EVs, Cn EVs, and Af EVs.

A, The -log₁₀ values of EV concentration from standard isolation preps. Significance assessed using an ordinary one-way ANOVA and Tukey’s multiple comparisons test with no significance determined. n ≥ 5. B, Quantification of EV concentration according to diameter for Ca EVs, Sc EVs, Cn EVs, and Af EVs. n ≥ 5. C, The median and mode diameters of Ca EVs, Sc EVs, Cn EVs, and Af EVs. n ≥ 5. D, Representative images from TEM of Ca EVs, Sc EVs, Cn EVs, and Af EVs. E, The average DNA concentration (pg per 5x10¹⁰ EVs) of Ca EVs, Sc EVs, Cn EVs, and Af EVs. Significance assessed using an ordinary one-way ANOVA and Tukey’s multiple comparisons test with no significance determined. n ≥ 3. F, Percent GC content of EV DNA reads that mapped to the reference fungal genome as a truncated violin box plot. G, Percent GC content of the whole fungal genome compared to the EV DNA.

Fig 5. Outer structural layers on EVs inhibit endocytosis and STING pathway activation.

A, Representative confocal microscopy showing GXM layer present on WT C. neoformans (H99) yeast and EVs but not on cap59∆ yeast and EVs. 1x10⁷ WT yeasts, 1x10⁷ cap59∆ yeasts, and 5x10¹⁰ WT and cap59∆ EVs were added to the respective stimulations. Scale bar = 10µm. B, Representative immunoblot of RodA (28.3 kDa) on Af EVs compared to mutant ΔrodA EVs. C, Percent endocytosis of Cn EVs and cap59∆ EVs by WT macrophages. (5x10¹⁰ EVs/mL added per stimulation), n = 2. D, Percent endocytosis of Af EVs and ΔrodA EVs by the total number of WT macrophages (5x10¹⁰ EVs/mL added per stimulation), n = 2. E, Immunoblot of viperin and actin in WT macrophages stimulated by PBS, Cn EVs, cap59∆ EVs, and the positive control for yeast, Ca EVs (5x10¹⁰ EVs/mL were added per stimulation). F, Immunoblot of viperin and actin in WT macrophages stimulated by PBS, Af EVs, and ΔrodA EVs (5x10¹⁰ EVs/mL were added per stimulation), and positive control for Aspergillus: 2.5µg cGAMP.