Conidial melanin of the human pathogenic fungus Aspergillus fumigatus disrupts cell autonomous defenses in amoebae

The human pathogenic fungus Aspergillus fumigatus is a ubiquitous saprophyte that causes fatal infections in immunocompromised individuals. Following inhalation, conidia are ingested by innate immune cells and can arrest phagolysosome maturation. How such general virulence traits could have been selected for in natural environments is unknown. Here, we used the model amoeba Dictyostelium discoideum to follow the antagonistic interaction of A. fumigatus conidia with environmental phagocytes in real time. We found that conidia covered with the green pigment 1,8-dihydroxynaphthalene-(DHN)-melanin were internalized at far lower rates when compared to those lacking the pigment, despite high rates of initial attachment. Immediately after uptake of the fungal conidia, nascent phagosomes were formed through sequential membrane fusion and fission events. Using single-cell assays supported by a computational model integrating the differential dynamics of internalization and phagolysosome maturation, we could show that acidification of phagolysosomes was transient and was followed by neutralization and, finally, exocytosis of the conidium. For unpigmented conidia, the cycle was completed in less than 1 h, while the process was delayed for conidia covered with DHN-melanin. At later stages of infection, damage to infected phagocytes triggered the ESCRT membrane repair machinery, whose recruitment was also attenuated by DHN-melanin, favoring prolonged persistence and the establishment of an intracellular germination niche in this environmental phagocyte. Increased exposure of DHN-melanin on the conidial surface also improved fungal survival when confronted with the fungivorous predator Protostelium aurantium, demonstrating its universal antiphagocytic properties.

such general virulence traits could have been selected for in natural environments is 23 unknown. Here, we used the model amoeba Dictyostelium discoideum to follow the 24 antagonistic interaction of A. fumigatus conidia with environmental phagocytes in real 25 time. We found that conidia covered with the green pigment 1,8-dihydroxynaphthalene- 26 (DHN)-melanin were internalized at far lower rates when compared to those lacking the 27 pigment, despite high rates of initial attachment. Immediately after uptake of the fungal 28 conidia, nascent phagosomes were formed through sequential membrane fusion and 29 fission events. Using single-cell assays supported by a computational model integrating 30 the differential dynamics of internalization and phagolysosome maturation, we could 31 show that acidification of phagolysosomes was transient and was followed by 32 neutralization and, finally, exocytosis of the conidium. For unpigmented conidia, the 33 cycle was completed in less than 1 h, while the process was delayed for conidia covered 34 with DHN-melanin. At later stages of infection, damage to infected phagocytes triggered 35 the ESCRT membrane repair machinery, whose recruitment was also attenuated by  glycoproteins, and β-1,3-glucan, whose exposure facilitates recognition, phagocytic 69 uptake and killing by immune cells (Chai et al., 2010, Luther et al., 2007. The 70 biochemical fate of fungal melanin following swelling and germination is thus far 71 unknown. 72 In contrast to commensal pathogens such as Candida albicans, A. fumigatus is 73 considered an environmentally acquired pathogen, as it is frequently isolated from 74 natural reservoirs and occupies a well-established niche as a decomposer of organic 75 matter. In its natural environment the fungus is confronted with many abiotic and biotic 76 adverse conditions such as amoebae with some of them being able to ingest and even To initiate phagocytosis, host receptors engage with ligands exposed on the surface of 122 A. fumigatus conidia. This association with its ligand initiates signaling pathways that 123 cause the extension of a lamellipodium, which surrounds the particle and generates the  We first analyzed infection outcomes after co-incubation of D. discoideum with wild-type 129 and pksP mutant conidia of A. fumigatus. After 1 h of co-incubation we found that D. 130 discoideum amoebae had ingested 63% of the melanin-deficient pksP conidia, but only 131 20% of the wild-type conidia ( Figure 1A-B). The phagocytic efficiencies determined for 132 wild-type and ∆pksP conidia were lower and higher than the ones for inert silica 133 particles, respectively ( Figure 1C and S1A). Melanin ghosts obtained after harsh 134 chemical treatment of wild-type conidia were rarely taken up by the amoeba. However, 135 these empty shells of melanin would readily associate with the amoeba cell wall, 136 covering the entire surface ( Figure 1A+B and Figure S1A). Collectively, our results 137 suggested that DHN-melanin might have an impact on the uptake process of conidia.

138
This conclusion was further supported by an experiment with the DHN-melanin 139 monomer 1,8-dihydroxynaphthalene, which also repressed phagocytosis of beads in a 140 dose-dependent manner ( Figure 1D). 141 We were further interested in the intracellular fate of the conidia and frequently observed 142 conidial exocytosis. We thus followed the infection process at the single-cell level and 143 monitored the time of intracellular transit of conidia in D. discoideum. Conidia were 144 stained with FITC (green) and with CF594 (red) for the normalization of signal intensity 145 ( Figure S1B and S2). Ratiometric calculations of the differences between the two dyes, 146 with FITC responding to changes in pH, allowed us to track the phagosomal pH 147 dynamics for conidia over the entire intracellular period ( Figure 1E). These 148 measurements demonstrated that both wildtype and ∆pksP mutant conidia underwent   Figure 2B). Also, following VatB-RFP retrieval, phagosomes took significantly 179 longer to reach pH 6 again when infected with DHN-melanin-covered wild-type conidia 180 when compared to pksP conidia-containing phagosomes ( Figure 2C).

313
We provided further evidence that the first intracellular processing steps in the amoeba, to an influx of nutrients and will thus help the fungus to establish a germination niche.

365
Although this advantage may be restricted to non-specialized phagocytes that are 366 unable to kill the fungus, we also found a protective role for DHN-melanin when   ImageJ. Then average log of these values were plotted on the calibration curve graph.

423
In order to determine pH on the sample image the integrated density were back 424 calculated from the calibration graph.  Phagocytosis of Aspergillus fumigatus conidia by Dictyostelium discoideum. a Resting conidia of the wild type (wt) or the melanin-deficient strain of A. fumigatus (ΔpksP) were added to D. discoideum at MOIs of 5. Silica beads (Bead) and melanin ghosts (Ghost) were added to the amoebae at the same MOI. Images were captured after 1 h of co-incubation. The scale bar is 5 µm. b Cells with phagocytic and attachment events were quantified from images captured at 1 h p. i. The bars represent the mean and SEM from three independent experiments, with n=100 for each experiment. Statistical differences were calculated with a Bonferroni posttest after a two-way ANOVA with asterisks indicating significance (*p<0.05; **p<0.01; ***p<0.001). c Phagocytic ratio for A. fumigatus conidia, silica beads and melanin ghosts. d Wild type amoeba were exposed to silicon beads in the presence of 10 or 50 µM of 1,8-DHN. Imaging and quantification were carried out as in b. e-g Amoebae were infected with resting conidia of the wild type or the ΔpksP strain pre-stained with the pH-sensitive fluorophore (FITC) and the reference fluorophore (CF594) for realtime measurements of acidification and residence time in the amoeba. e Timeline of FITC derived fluorescence intensity indicating pH variations at the conidial surface during phagocytosis. f The intracellular retention time of conidia inside of D. discoideum. Statistical differences were calculated with a t-test. g Time-lapse illustration of major steps during the phagocytic cycle for resting conidia of the ΔpksP mutant.  Internalized conidia and free conidia are indicated by red and white arrows, respectively. b, d Quantification of RITC fluorescence of the two dextrans (B, 70,000 and C, 4,400) as integrated density in conidia containing phagosomes. Values were normalized by background substraction of free conidia. Images were captured after 300 mpi. Data are based on 3 biological replicates with statistical differences calculated in a one way ANOVA with p<0.0001. e, f Schematic representation of size discriminated leakage of dextran from phagolysosomes. g, e Vps32-GFPexpressing cells were infected with dormant (g) or swollen conidia (i) at an MOI of 10 and representative images from 180 m. p. i. are shown. h, j Quantification of Vps32-GFP localization to conidia containing phagosomes. Statistical differences were calculated with a Bonferroni post-hoc test after a two-way ANOVA with asterisks stating significance with *p<0.05, **p<0.01, and ***p<0.001). Scale bars are 5 µm.