A bacterial endosymbiont of the fungus Rhizopus microsporus drives phagocyte evasion and opportunistic virulence

Abstract

Opportunistic infections by environmental fungi are a growing clinical problem, driven by an increasing population of people with immunocompromising conditions. Spores of the Mucorales order are ubiquitous in the environment but can also cause acute invasive infections in humans through germination and evasion of the mammalian host immune system. How they achieve this and the evolutionary drivers underlying the acquisition of virulence mechanisms are poorly understood. Here, we show that a clinical isolate of Rhizopus microsporus contains a Ralstonia pickettii bacterial endosymbiont required for virulence in both zebrafish and mice and that this endosymbiosis enables the secretion of factors that potently suppress growth of the soil amoeba Dictyostelium discoideum, as well as their ability to engulf and kill other microbes. As amoebas are natural environmental predators of both bacteria and fungi, we propose that this tri-kingdom interaction contributes to establishing endosymbiosis and the acquisition of anti-phagocyte activity. Importantly, we show that this activity also protects fungal spores from phagocytosis and clearance by human macrophages, and endosymbiont removal renders the fungal spores avirulent in vivo. Together, these findings describe a new role for a bacterial endosymbiont in Rhizopus microsporus pathogenesis in animals and suggest a mechanism of virulence acquisition through environmental interactions with amoebas.

Keywords: Dictyostelium; Murcomycete; Ralstonia; Rhizopus; endosymbiosis; evolution; fungal pathogenesis; soil microbiology.

Figure 1

Swelling of R. microsporus FP469-12 spores inhibits phagocytosis (A) Phagocytosis of resting spores, or those allowed to swell for 2 or 4 h, by J447.2 macrophages. (B) Effect of swollen spore supernatant on phagocytic uptake of naive, resting R. microsporus spores by macrophages. (C) Effect of R. microsporus FP469-12 conditioned medium (Dulbecco’s modified eagle medium [DMEM]) on phagocytosis of C. albicans and S. cerevisiae. For all assays, the number of macrophages containing at least one spore were counted after 1 h. Counts were normalized to uptake of untreated resting spores in each replicate. n = 3 biological replicates of >1,000 macrophages each, error bars represent SEM. ^∗p < 0.05, ^∗∗p < 0.001, ^∗∗∗p < 0.0001, one-way ANOVA with Tukey’s correction for multiple comparisons. See also Figure S1.

Figure 2

R. microsporus FP469-12 contains a bacterial endosymbiont required for anti-phagocytic activity (A) PCR screen for the presence of bacterial 16S rDNA. Genomic DNA was isolated from wild-type and ciprofloxacin-treated R. microsporus FP469-12 cells, as well as the isolated endosymbiont alone. Presence of 16S rDNA is indicated by the presence of a 1.5-kb PCR product. (B and C) (B) Phylogenetic comparison of R. microsporus FP469-12 and (C) its R. pickettii endosymbiont based on 28S and 16S sequences respectively. Both were aligned with MUSCLE, bootstrapped and produced with RAxML. Strain R. mircosporus var. microsporus is CBS 699.68, shown to harbor M. rhizoxinica. Strains CCF4531 and VPCI are clinical isolates from a nasal and pulmonary mass respectively. (D) SYTO9 staining of R. microsporus FP469-12 mycelium for bacterial endosymbionts. Spores of parent and ciprofloxacin-treated cells were fermented in VK medium, the mycelial pellet submerged in NaCl and then stained with SYTO9 prior to brightfield and fluorescence imaging. (E) High resolution confocal imaging of endosymbionts in both spores and hyphae allowed to germinate for 4 h then stained with SYTO9. R. pickettii alone was grown for 16 h, then stained with SYTO9. Top row shows the parental and ciprofloxacin-treated cells, and isolated R. pickettii, scale bars represent 10 μm. The boxed regions are enlarged below showing both hyphae (middle row) and spores (bottom) row, for each fungal strain (scale bars, 5 μm). (F) Phagocytosis of parental, and cirprofloxacin-teated (endosymbiont-free) R. microsporus FP469-12 spores by J774.2 macrophages, upon swelling. (G) Contributions of the fungi and bacteria to the secreted anti-phagocytic activity. J774.2 cells were incubated for 1 h with resting spores in medium conditioned by either parental spores (with endosymbionts), endosymbiont-free spores, or the isolated R. pickettii endosymbiont alone. (H) Effect of media conditioned by co-cultures of bacterial symbionts and endosymbiont-free fungal spores grown on phagocytosis of R. microsporus resting spores. Each graph shows the mean and SEM of 3 independent experiments. ^∗p < 0.05, ^∗∗p < 0.001, ^∗∗∗p < 0.000, one-way ANOVA with Tukey’s correction for multiple comparisons. See also Figure S2.

Figure 3

Fungal-bacterial endosymbiosis inhibits amoeba phagocytosis and growth (A) Phagocytosis of heat-killed S. cerevisiae by D. discoideum in either normal medium or medium pre-conditioned by R. microsporus FP469-12. (B) DIC images of D. discoideum cells 0 and 15 min after addition of conditioned medium. Black arrowheads indicate the large swollen vacuoles induced, yellow asterisks mark forming protrusions. (C) Dose-dependent inhibition of D. discoideum growth by R. microsporus FP469-12 conditioned medium. Amoebas were incubated in different concentrations of conditioned medium diluted in fresh medium as indicated, and cells counted at each time point. (D) Fluid uptake (macropinocytosis) by D. discoideum cells in medium conditioned by either parental, or endosymbiont-free R. microsporus FP469-12 spores. Cells were incubated in TRITC-dextran containing medium, and fluorescent dye uptake measured by flow cytometry. (E) Effect of endosymbiont removal on the ability of R. microsporus FP469-12 conditioned medium to inhibit D. discoideum growth. (F) Effect of R. pickettii-conditioned medium on D. discoideum growth. HL5 medium was conditioned for 4 h by addition of the indicated dilutions of an overnight R. pickettii culture, before bacteria were removed. Each graph shows the mean and SEM of 3 independent experiments. ^∗p < 0.05, ^∗∗p < 0.01, ^∗∗∗p < 0.005, ANOVA with Tukey’s correction for multiple comparisons. See also Figure S3.

Figure 4

The secreted activity inhibits phagosome maturation by a novel mechanism (A) Phagosomal proteolysis of D. discoideum cells incubated in fungal-conditioned medium, in the presence, or absence of the endosymbiont. Measured by increasing fluorescence of DQ-BSA-conjugated beads after engulfment. (B) Kaplan-Meijer survival curve of GFP-expressing K. pneumonia after engulfment by D. discoideum in R. microsporus FP469-12-conditioned medium. Phagocytosis was observed by time-lapse fluorescence microscopy, and the point of bacterial death inferred from the quenching of GFP-fluorescence (n > 100 for each condition, ^∗∗∗p < 0.001, log rank Mantel-Cox test). (C and D) (C) Maximum intensity projections of D. discoideum cells expressing GFP α-tubulin after 20-min treatment with either conditioned medium or (D) the indicated concentrations of rhizoxin D. (E) Quantification of the proportion of cytoplasm covered by the microtubule array in cells treated as in (C) and (D) (^∗∗∗p < 0.001, t test). (F) Effect of rhizoxin D treatment on D. discoideum growth. Generation times calculated from growth curves obtained over 72 h (^∗∗∗p < 0.001, paired t test). Unless otherwise indicated all graphs show the mean and standard deviations of 3 independent experiments.

Figure 5

Endosymbiosis with R. pickettii influences the fungal cell wall and stress tolerance (A) Survival (CFUs) of R. microsporus FP469-12 resting spores after incubation with J774.1 macrophages in the presence and absence of endosymbiont. (B and C) Impact of the endosymbiont on R. microsporus FP469-12 spore survival under cell wall, nitrosative, oxidative, and antifungal stresses (AmB: Amphotericin B). Either resting (B) or swollen (C) spores were incubated at the indicated concentrations for 24 h, prior to CFU determination. (D) TEM images of spores swollen for 4 h in the presence or absence of the endosymbiont. Lower panels show enlargements of representative cell wall regions. Corresponding regions with comparable density are indicated by the different colored bars. (E–H) Show comparative changes in cell wall composition upon swelling of the parental FP469-12 strain (+Endo) and endosymbiont-free fungal spores (−Endo). Staining intensities were quantified by fluorescence microscopy. (E) shows total chitin (calcofluor white), (F) exposed chitin (wheat-germ agglutinin), (G) total protein (FITC), and (H) mannan (concanavalin A) (n = 300 for each repeat). All graphs show mean ± SEM of 3 repeats (^∗∗p < 0.001, ^∗∗∗p = 0.0001, ^∗∗∗∗p < 0.00001, one-way ANOVA with Tukey’s correction for multiple comparisons). See also Figure S4.

Figure 6

Effect of endosymbiont status on fungal infection of zebrafish (A) Survival of AB wild-type zebrafish injected via the hindbrain with resting or swollen spores of R. microsporus FP469-12 in the presence or absence of endosymbiotic bacteria. PVP indicates mock-injected fish. Three biological replicates of populations of 10 fish each were examined (n = 30). Statistical differences were determined using Mantel-Cox with Bonferroni’s correction for multiple comparisons (5% family-wise significance threshold = 0.025). (B and C) Effect of endosymbiont status on fungal survival (CFUs) following hindbrain injections of AB wild-type zebrafish with (B) resting or (C) swollen R. microsporus FP469-12 spores. Three biological replicates of 5 fish per condition were examined (n = 15). (D and E) Effect of the endosymbiont on in vivo recruitment of macrophages and neutrophils to the site of infection. Figure S5 shows representative images and Figure S6 shows equivalent data with swollen spores. Statistical significance was assessed by two-way ANOVA with Tukey’s correction for multiple comparisons or pairwise t tests where sample number was unequal due to fish death, ^∗p < 0.05; ^∗∗p < 0.001; ^∗∗∗p < 0.0001 unless otherwise indicated.

Figure 7

Ciprofloxacin treatment can modify R. microsporus FP469-12 infection in mice. (A) Fungal survival (CFUs) at 4 and 48 following intra-tracheal infection of mice hours with resting spores. Where indicated, spores were pre-treated with 60 μg/mL ciprofloxacin for 3 h prior to infection. (n = 5, significance determined by Mann-Whitney test.) (B and C) Proportion of mice positive or negative (below the detection limit) for fungal CFU’s 48 hpi. (B) Data from infection with resting spores with and without ciprofloxacin pre-treatment, (C) is the same experiment performed with swollen spores. n = 5 mice per condition, statistical significance assessed using a two-sided chi-squared test, and attributable risk was assessed using Newcombe/Wilson with continuity correction, ^∗p = 0.384.