Fig 1.
A CEA10-derived strain is killed more than an Af293-derived strain in vivo.
A. Schematic of 2 dpf zebrafish larvae. Area of Aspergillus spore injection (hindbrain) and imaging is indicated. B, C. TBK1.1 (Af293) or TFYL49.1 (CEA10) spores express YFP or GFP and have AlexaFluor594 conjugated to the spore cell wall. 2 dpi larvae were imaged and the percentage of spores killed (GFP+/total) was quantified. Representative images of single z slices (B) and quantification over three experiments (C) are shown. Arrowheads indicate alive spores, asterisks indicate killed spores. Scale bar represents 10 μm. Each symbol represents one larvae, color-coded by experiment. Lines represent lsmeans ± SEM. Af293 n = 51, CEA10 n = 45. D. Fungal burden was monitored by CFU platings from single homogenized larvae infected with TBK1.1 (Af293) or TFYL49.1 (CEA10). CFUs from 24 larvae (3 replicates, 8 larvae each) per condition per day were measured. Average injection CFUs: Af293 = 150, CEA10 = 140. For all analyses, data represent lsmeans ± SEM from three pooled experiments, P values calculated by ANOVA. Asterisks represent significance between strains, # represent significance compared to day 0. See also S1 Fig.
Fig 2.
Inflammatory activation is required for clearance of a CEA10-derived strain.
A, B. NF-κB RE:EGFP larvae were infected with non-fluorescent TJW55.2 (Af293) or CEA17 KU80Δ (CEA10) spores and imaged 2 dpi. Representative images (A) and quantification from three pooled experiments (B) are shown. Scale bar represents 50 μm. CEA10 n = 22; Af293 n = 26. C. Larvae were treated with dexamethasone (DEX) or ethanol (EtOH) vehicle control directly after infection with TFYL81.5 (Af293) or TFYL49.1 (CEA10), and larval survival was monitored. Average injection CFUs: Af293 = 34, CEA10 = 33. D-F. myd88-/- or control larvae were infected with TBK1.1 (Af293) or TFYL49.1 (CEA10) and larval survival and/or fungal burden was monitored. Fungal burden was determined by CFU platings from single homogenized larvae. Asterisks represent significance between myd88-/- and control larvae, # represent significance compared to day 0. Average injection CFUs: D, F: Af293 = 27, CEA10 = 32; E: Af293 = 66. For all analyses, data represent 3 pooled replicates. For CFU experiments, CFUs from 24 larvae (8 larvae per replicate) per condition per day were measured. Intensity values and CFU analyses represent lsmeans ± SEM from three pooled experiments, P values calculated by ANOVA. P values for survival analyses calculated by Cox proportional hazard regression analysis. See also S2 Fig.
Fig 3.
CEA10-induced inflammatory activation is not sufficient for killing of an Af293-derived strain.
A. Fungal burden was monitored by CFU platings from single homogenized larvae infected either singly with TFYL49.1 (CEA10) or TBK5.1 (Af293) or co-infected with both strains. Average injection CFUs: Af293 = 53, CEA10 = 47, co-infection = 21 Af293 + 27 CEA10. CFUs from 24 larvae (8 larvae per replicate) per condition per day were measured; lsmeans ± SEM are shown from three pooled experiments, P values calculated by ANOVA. B, C. Macrophage-membrane labeled larvae (mfap4:tomato-CAAX) were co-infected with YFP-expressing TBK1.1 (Af293) and AlexaFluor633-labeled CEA17 KU80Δ (CEA10) and imaged 1 dpi. Percentage of Af293 spores within 20 μm of a CEA10 spore in each larvae from two replicates was quantified (B). Each point represents one larvae, color-coded by replicate. Single z-slice images from a representative larvae (C) are shown. Scale bar represents 5 μm. See also S3 Fig.
Fig 4.
Macrophages form tight clusters around A. fumigatus.
A-D. Dual macrophage-nuclear (mpeg1:mcherry-H2B) and neutrophil (lyz:BFP) labeled larvae were infected with YFP- or GFP-expressing A. fumigatus TBK1.1 (Af293) or TFYL49.1 (CEA10) strains and imaged days 1–5 post injection. Macrophage (A) and neutrophil (B) recruitment as well as phagocyte cluster size (D) were quantified. Representative images (C, scale bar represents 20 μm) and pooled quantification from 3 experiments (A, B, D) are shown. Each line in graphs represents one larvae followed for the entire course of infection, lines are color-coded by replicate, bars represent pooled lsmeans ± SEM, P values calculated by ANOVA. Af293 n = 25, CEA10 n = 27 for macrophage and cluster area quantification; Af293 n = 17, CEA10 n = 18 for neutrophil quantification. E. Macrophage-membrane (mfap4:tomato-CAAX) labeled larvae were infected with CEA17 KU80Δ (CEA10) and imaged 3 dpi. Asterisks represent cell nuclei, arrows indicate cell-cell junctions. Scale bar represents 10 μm. See also S4 Fig.
Fig 5.
A CEA10-derived strain is more virulent in a neutrophil-defective host.
Neutrophil-defective (mpx:rac2D57N) or control (mpx:rac2WT) larvae were infected with TBK1.1 (Af293) or TFYL49.1 (CEA10). A. CFUs were monitored, average injection CFUs: Af293 = 36, CEA10 = 32. B. Larval survival was monitored, average injection CFUs: Af293 = 35, CEA10 = 27. For all analyses, data shown are from 3 pooled replicates. For CFU analysis, data represent lsmeans ± SEM, P values calculated by ANOVA. CFU data are from 24 larvae (3 replicates, 8 larvae each) per condition per day. For survival analysis, P values were calculated by Cox proportional hazard regression analysis. See also S5 Fig.
Fig 6.
A CEA10-derived strain is cleared more efficiently by macrophage-deficient hosts than wild-type hosts.
A, B. Macrophage-deficient (irf8-/-) or control (irf8+/+) larvae were infected with TBK1.1 (Af293) or TFYL49.1 (CEA10) and CFUs (A) or larval survival (B) were monitored. For CFU experiments, average injection CFUs: Af293 = 23, CEA10 = 23. For survival experiments, average injection CFUs: Af293 = 44, CEA10 = 31. C, D. Macrophage-depleted (clodronate liposomes) or control (PBS liposomes) larvae were infected with TBK1.1 (Af293) or TFYL49.1 (CEA10) and CFUs (C) or larval survival (D) were monitored. Average injection CFUs: Af293 = 37, CEA10 = 40. For all analyses, data shown are from 3 pooled replicates. For CFU analyses, data represent lsmeans ± SEM, P values calculated by ANOVA. All CFU data are from 24 larvae (3 replicates, 8 larvae each) per condition per day. For survival analyses, P values were calculated by Cox proportional hazard regression analysis. See also S6 Fig.
Fig 7.
More neutrophils are recruited to a CEA10-derived strain than an Af293-derived strain in the absence of macrophages.
A-D. Macrophage-deficient (irf8-/-) or control (irf8+/+) larvae were infected with YFP- or GFP-expressing A. fumigatus TBK1.1 (Af293) or TFYL49.1 (CEA10) strains, fixed 1 dpi, and stained for neutrophils (sudan black). Representative images are shown (A, scale bar represents 50 μm) and percent of larvae per experiment with sudan black masses (B), number of neutrophils at infection site (C), and percent of larvae per experiment with A. fumigatus hyphae (D) were quantified. N>9 larvae each condition each replicate. For percentage of larvae with sudan black masses or hyphae (B, D), each symbol represents one replicate, bars represent means ± SEM, P values calculated by t test. For neutrophil number quantification (C), each symbol represents one larvae. Larvae with sudan black masses where neutrophil numbers were difficult to count and represent underestimations are denoted with open symbols. Bars represent lsmeans ± SEM from 3 pooled replicates, P values calculated by ANOVA. All symbols are color-coded by replicate. E. Wild-type or myd88-/- embryos were injected with control (std) or macrophage-depleting (irf8) morpholinos. Larvae were then infected with TFYL49.1 (CEA10) spores and CFUs were measured. Average injection CFUs = 38. CFU data represent lsmeans ± SEM of 3 pooled replicates, n = 24 larvae (3 replicates, 8 larvae each) per condition per day, P values calculated with ANOVA. See also S7 Fig.
Fig 8.
A CEA10-derived strain germinates faster than an Af293-derived strain in vivo.
Wild-type or phagocyte-deficient (pu.1 morpholino) larvae were infected with YFP- or GFP-expressing A. fumigatus TBK1.1 (Af293) or TFYL49.1 (CEA10). All graphs represent data from 3 pooled replicates, Ns and labels noted in (A) are applicable for all data. A. Larval survival was monitored, P values calculated by Cox proportional hazard regression analysis. B-E. Larvae were imaged days 1–5 post injection and fungal growth was measured. Representative z-projection images 1 dpi are shown (B). Scale bar represents 100 μm. Asterisks indicate examples of germinated spores, arrowheads indicate examples of invasive hyphae. Cumulative percent of larvae with germinated spores (C) and with invasive hyphae (D) was calculated, P values were calculated by Cox proportional hazard regression analysis. 2D GFP+ fungal area was measured from maximum intensity projection images of individual larvae for 5 days of infection (or until larvae succumbed) (E). Each line represents one larvae followed for the entire course of infection, lines are color-coded by replicate, bars represent pooled lsmeans ± SEM, P values were calculated by ANOVA. See also S8 Fig.
Fig 9.
Neutrophil-mediated killing requires germination.
A. Macrophage-deficient (irf8-/-) or control (irf8+/+) larvae were injected with live or heat-killed TFYL49.1 (CEA10), fixed 1 dpi, and stained for neutrophils (sudan black). Larvae were then scored for the presence or absence of a sudan black cluster or mass. This experiment was performed twice, data are shown from one representative replicate. B. Wild-type or phagocyte-deficient (pu.1 morphant) larvae were infected with TBK1.1 (Af293 pyrG+), Af293.1 (Af293 pyrG-), TFYL49.1 (CEA10 pyrG+), or CEA17 (CEA10 pyrG-) spores and survival was monitored. Data represent 2 pooled experiments. Average injection CFUs: Af293 pyrG+ = 52, Af293 pyrG- = 52, CEA10 pyrG+ = 38, CEA10 pyrG- = 48. C. Wild-type larvae were infected with TBK1.1 (Af293 pyrG+), Af293.1 (Af293 pyrG-), TFYL49.1 (CEA10 pyrG+), or CEA17 (CEA10 pyrG-) and CFUs were monitored. Average injection CFUs: Af293 pyrG+ = 45, Af293 pyrG- = 48, CEA10 pyrG+ = 42, CEA10 pyrG- = 49. D. Macrophage-deficient (irf8-/-) larvae were infected with CEA17 KU80Δ (CEA10 pyrG+), or CEA17 (CEA10 pyrG-) and CFUs were monitored. Average injection CFUs: CEA10 pyrG+ = 41, CEA10 pyrG- = 45. All CFU data represent lsmeans ± SEM from 3 pooled replicates composed of 24 larvae (8 larvae per replicate) per condition per day. P values were calculated by ANOVA. See also S9 Fig.
Fig 10.
Model of neutrophil and macrophage responses to Aspergillus fumigatus.
Upon infection, A. fumigatus spores are taken up by macrophages, which form dense clusters around the fungus and inhibit spore germination. Macrophages provide a protective niche for spore survival, as neutrophil-mediated killing of A. fumigatus requires germination. As a result, in infections with slow-germinating A. fumigatus strains, the fungus persists for days, while in infections with fast-germinating A. fumigatus strains, germination drives neutrophil recruitment, neutrophil-mediated killing, and faster fungal clearance. Macrophages have some anti-fungal activity, dependent on Myd88-NF-κB signaling, however NF-κB activation is not sufficient to kill slow-germinating strains.
Table 1.
Zebrafish lines used in this study.
Table 2.
Aspergillus fumigatus strains used in this study.