Inbred BALB/c female mice were purchased from Charles River (Sulzfeld, Germany) and maintained in the pathogen-free animal facility of the Institute for Molecular Infection Biology (IMIB), University of Würzburg, Germany. Firefly luciferase transgenic BALB/c.L2G85 female mice had been backcrossed from FVB/N.L2G85 mice for more than 12 generations (Cao et al., 2004; Beilhack et al., 2005). All experiments were performed with 8–12-week-old female mice. All animal experiments were carried out according to German guidelines for animal experimentation. The responsible authority (Regierung von Unterfranken; Permit Number 55.2-2531.01-86-13) approved the study.

In the cyclophosphamide and corticosteroid treated (CCT) model, mice were intraperitoneally injected with 150 mg kg^-1 cyclophosphamide (Sigma–Aldrich, Munich, Germany) and subcutaneously (s.c.) with 112 mg kg^-1 hydrocortisone acetate (Sigma–Aldrich) on days -3 and -1 before A. fumigatus infection. In the corticosteroid treated (CT) model, mice were s.c. injected with 112 mg kg^-1 hydrocortisone acetate on days -3 and -1 before infection.

The clinical isolate of A. fumigatus ATCC46645 strain (Hearn and Mackenzie, 1980) was routinely used. Fluorescent A. fumigatus strain Afu-TdTomato (Lother et al., 2014) generated from ATCC46645 (kindly provided by Dr. Sven Krappmann) was used to determine fungal developmental stages inside lung tissue. All the fungal strains were cultivated on defined minimal medium (Amich et al., 2013) under standard culture conditions and handled according to German laboratory safety guidelines. Conidia were harvested from sporulating mycelium using the standard saline/0.01% tween solution, filtered through cell strainer and finally washed with sterile saline. Mice were anesthetized by intraperitoneal injection of ketamine (50 μg/g bodyweight) and xylacine (5 μg/g bodyweight) in 0.1 M phosphate-buffered saline (PBS) in a total volume of 10 μl/g bodyweight and intra-nasally infected with 1 × 10⁶ conidia suspended in 50 μl saline/0.01% tween. All infected mice were monitored carefully according to the standard guidelines; briefly, mice were regularly observed twice a day and carefully monitored for weight loss and disease symptoms. In the immune cell recruitment studies at least n = 3/group of mice were used in each independent experiment and data are pooled from three different experiments. In survival studies at least n = 5/group of mice were used in each independent experiment.

Single cell suspensions were prepared from lungs according to the previously described protocol (Stockmann et al., 2010) with some modifications. Briefly, left and right lung lobes were dissected from euthanized mice and finely minced using surgical blades in six well-tissue culture plates containing RPMI medium (Life Technologies, USA), then enzymatically digested for 30 min at 37°C in presence of 2 mg/ml Collagenase D and 0.1 mg/ml DNase I (Roche, Mannheim, Germany), diluted with PBS + 0.5 % BSA, filtered through a 70 μm cell strainer (Greiner bio-one, Frickenhausen, Germany) and centrifuged at 1200 rpm for 5 min. The lung cell pellet was re-suspended in erythrocyte lysis buffer (168 mM NH₄Cl, 10 mM KHCO₃, 0.1 mM ethylene diamine tetra acetic acid (EDTA)) for 2 min, and immediately diluted with double the volume of PBS and centrifuged. Finally, single cell suspensions were diluted to desired volumes suitable for flow cytometry analyses.

Cells were blocked with normal rat serum (1: 20 in PBS) and stained with appropriate antibodies at 4°C for 30 min. To discriminate live/dead cells, they were stained with LIVE/DEAD fixable violet dead cell stain kit (Invitrogen). All the antibodies used were from Biolegend (Uithoorn, The Netherlands). Antibodies (clones) utilized are listed below: CD90.2-PE (30-H12), CD4-APC/Cy7 (GK1.5), CD8-APC/Cy7 (53–6.7), CD11b-perCP-Cy5.5 (M1/70), CD11b-PE (M1/70) CD11c-FITC (N418), I-A/I-E-PE/Cy7 (M5/114.15.2), SiglecF-APC (E50-2440), Ly-6G-APC (1A8), FITC-Ly-6C (HK1.4), Ly6C-PerCP-Cy5.5 (HK1.4), F4/80-APC/Cy7 (BM8), CD49b-PE/Cy7 (DX5). All experiments were carried out using a BD fluorescence-activated cell sorting (FACS) Canto II (BD Biosciences) and data was recorded using BD FACS Diva software and analyzed using FlowJo software version 8.0 (Tree Star, Ashland, OR, USA).

Cryo-embedded lung tissues were cut into 8 μm thick sections on a Leica CM1900 cryostat (Leica Microsystems, Wetzlar, Germany) and mounted onto frosted slides. Slides were air-dried and fixed with acetone at room temperature for 7 min. Slides were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) and mounted with mounting medium (Vector Laboratories, Peterborough, UK) or stained with hematoxylin and eosin. Luciferase expressing CD11b⁺ myeloid cells were stained with an anti-luciferase antibody (Abcam, USA) and the secondary Goat anti-Rabbit IgG, FITC-conjugated antibody (Abcam, USA) according to the manufacturer’s instructions. To detect apoptotic cells TUNEL staining was performed with a commercial kit (Roche Diagnostics, Mannheim, Germany) according to manufacturer’s instructions. Images were taken using Z1 fluorescence microscope (Carl Zeiss, Gottingen, Germany) and evaluated with Zeiss AxioVision software (Carl Zeiss).

Lungs were homogenized in 500 μl PBS using Precellys ceramic kit 1.4 mm in a Precellys 24 homogenizer. Serum was separated from cell debris by 10 min centrifugation at 13000 rpm 4°C and immediately stored at -80°C until further use. Cytokine/chemokine concentrations were measured using BD Cytometric Bead Array Kit (BD Pharmingen, Heidelberg, Germany) or Biolegend Multiplex assay kit (Biolegend, Uithoorn, The Netherlands) according to the manufacturer’s instructions. Data were analyzed by FCAP Array v2.0 software.

Mouse CD11b⁺ myeloid cells were enriched from bone marrow (flushed from femur and tibia bones with PBS) of healthy untreated or hydrocortisone-treated BALB/c mice, using myeloid cell enrichment kit (STEMCELL Technologies, Cologne, Germany) according to manufacturer’s instructions. Cell purity was confirmed by post-enrichment FACS analysis (>90%) in all the experiments. Enriched cells were adoptively transferred via retro-orbital i.v. injection after mice were anesthetized by intraperitoneal injection of ketamine (50 μg/g bodyweight) and xylacine (5 μg/g bodyweight) in 0.1 M Phosphate-Buffered Saline (PBS) in a total volume of 10 μl/g bodyweight.

Ex vivo lung bioluminescence imaging was performed as previously described (Chopra et al., 2013, 2015). Briefly, mice were injected with 300 mg/kg D-luciferin and euthanized after 10 min to prepare lungs and immediately subjected to ex vivo bioluminescence imaging with an IVIS Spectrum imaging system (Perkin–Elmer/Caliper Life Sciences, Mainz, Germany). Images were evaluated with Living Image 4.0 software (Caliper Life Sciences).

All the measurements are expressed as the mean ± standard deviation (SD). Statistical analyses were performed using Graph Pad Prism 6 (Groningen, The Netherlands) software. To compare cell numbers between the two different groups the unpaired Mann–Whitney u-test was applied. Significant differences are marked as follows: ^∗P < 0.05; ^∗∗P < 0.01; ^∗∗∗P < 0.001. To compare survival curves of infected mice, the Log-rank (Mantel-Cox) test was utilized.

To explore how immunosuppressive therapy affects pulmonary control of A. fumigatus infection, we compared immunocompetent mice with two different immunosuppressed mouse models (Figure 1A). Firstly, cyclophosphamide and cortisone treated (CCT) mice and, secondly, corticosteroid treated (CT) mice to investigate pulmonary host immune responses following respiratory A. fumigatus infection. We examined different morphotypes of fungal developmental stages in infected lung sections at 4, 16, and 40 h post-infection (p.i.) time points with immunofluorescence microscopy. We observed fungal differentiation from conidia at 4 h p.i. to germlings at 16 h p.i., and hyphae at 40 h p.i. in CT infected lung sections and CCT infected lung sections (Figure 1B). However, we observed elongated filaments (hyphal growth) in CCT mice at 16 h and more clearly at 40 h (Figure 1B) compared to CT infected mice. Strikingly, these results suggested that different numbers or types of immune cells might have been recruited to lungs of CT infected mice to restrict the hyphal growth. Qualitative hematoxylin & eosin staining of lung sections from immunocompetent, CT and CCT mice at 40 h p.i. exhibited different levels of infiltrated lung immune cells. Lung sections from immunocompetent mice revealed a strong pulmonary immune cell infiltration, CT mice showed less infiltration compared to immunocompetent mice, whereas CCT mice showed fewer infiltrating immune cells (Figure 1C). Next, we infected immunocompetent, CCT and CT mice with A. fumigatus conidia to determine their survival after A. fumigatus infection. Immunocompetent mice were resistant to infection, whereas CCT mice survived until 4 days p.i. and CT mice survived until 7 days p.i. (Figure 1D). We hypothesized that some immune cells would have been recruited to the infected lungs to fight against infection in these immunocompromised mouse models.

To determine the timing and magnitude of immune cell recruitment at different stages of A. fumigatus infection in immunocompromised CCT mice, we infected them with 1 × 10⁶ A. fumigatus conidia intranasally and analyzed defined immune cell populations in the lungs at 4, 16, and 40 h p.i. by flow cytometry (Supplementary Figure S1). As determined previously, at these selected time points the fungus had evolved through different morphotypes (conidia, germlings, and hyphae, respectively) that would likely trigger distinct types of immune responses. All immune populations were strongly reduced in lungs of CCT mice when compared to the immune cells in lungs of immunocompetent mice at steady-state-conditions (Figure 2). Upon infection, myeloid cells, especially neutrophils (Figure 2A) and macrophages (Figure 2B) were significantly recruited to the lungs of CCT mice at 4 h p.i. However, cell numbers did not surmount numbers of non-infected immunocompetent mice under steady-state-conditions, suggesting that there were not sufficient cells to fight against infection. Despite their low number, these cells were strongly recruited at the 4 h p.i. time point; but not at 16 and 40 h p.i. (Figures 2C,D). We did not observe recruitment of other myeloid cells particularly monocytes, dendritic cells and eosinophils in CCT mice upon A. fumigatus infection (Supplementary Figure S2). Lymphoid cells, particularly NK cells, CD4⁺ T cells and CD8⁺ T cells were strongly reduced in the lungs of CCT mice and were not recruited upon A. fumigatus infection (Figure 2E–G), suggesting that lymphoid populations cannot play a pivotal role in the defense against A. fumigatus under these immunosuppressive conditions. To investigate underlying factors behind minute immune cell infiltration in CCT mice particularly at 16 and 40 h p.i., we performed TUNEL staining to observe apoptotic cells (Supplementary Figure S3). TUNEL positive cells appeared in CCT mice at 16 and 40 h p.i., whereas in IC mice no TUNEL positive cells were observed at 40 h p.i. Despite low or reduced recruitment at 16 and 40 h after infection, fungus growth was not controlled in CCT infected mice. Growing hyphae in the lung tissue might lead to apoptosis of some of the cells in CCT mice even at 40 h after infection, whereas in IC infected mice until 40 h fungus might have been cleared and no apoptotic cells were found in TUNEL staining.

Corticosteroids are widely used immunomodulatory drugs in patients for a variety of clinical conditions (Shaikh et al., 2012). Corticosteroid treated mouse models are also employed to determine virulence of A. fumigatus mutants (Grahl et al., 2011). The phagocyte recruitment in corticosteroid treated mice after A. fumigatus infection had been previously studied (Duong et al., 1998; Balloy et al., 2005); however, the temporal kinetics of this dynamic immune cell response after A. fumigatus infection remains poorly defined. To determine the local host immune responses against A. fumigatus infection in CT mice, we infected CT mice with A. fumigatus conidia and analyzed immune cell recruitment at the above-specified time points of fungal development. Myeloid cells, particularly neutrophils (Figure 3A), macrophages (Figure 3B), dendritic cells (Figure 3C) and monocytes (Figure 3D) were recruited to the lungs of CT infected mice at 4 h p.i. Myeloid cell recruitment to lungs of infected mice was high at 4 h p.i. and low at 16 and 40 h p.i. (Figures 3E–H). Lymphoid cells were significantly reduced under these conditions and not recruited upon A. fumigatus infection (Supplementary Figure S4).

Myeloid cells were strongly recruited to the infected lungs in corticosteroid treated mice. To determine the lung cytokine environment at different time points after A. fumigatus infection in corticosteroid treated mice, we measured inflammatory cytokines in lung homogenates of immunocompetent, CT infected and non-infected mice. At 4 h p.i. the amount of the inflammatory cytokines MCP-1, IFN-γ, TNF-α, IL-6, and IL-12 in CT infected mice significantly exceeded cytokine levels in CT non-infected mice (Figure 4A). At 40 h p.i. the levels of lung inflammatory cytokines, except IFN-γ were similar in both CT infected and non-infected mice. However, the amount of the anti-inflammatory cytokine IL-10 was significantly higher in CT infected mice compared to non-infected mice at both 4 and 40 h after infection (Figure 4B). In contrast to CT mice, inflammatory or anti-inflammatory cytokines were below detection limits to determine in lungs of CCT mice with or without infection by the multiplex assay. Strikingly, these results suggest that increased inflammatory response in CT mice after infection is accompanied by high lung myeloid cell recruitment to the CT infected lungs.

Regardless of the immune status of mice, myeloid but not lymphoid cells were recruited to the site of infection. Despite their strongly reduced immune cell numbers, this was also true for the lungs of CCT mice after A. fumigatus infection. To determine whether myeloid cells alone can rescue immunosuppressed mice from lethal A. fumigatus infection we adoptively transferred CD11b⁺ myeloid cells into immunosuppressed mice that had been treated with cyclophosphamide (150 mg/kg) on days -3 and -1 (Figure 5A) alone, since corticosteroid models might interfere with antifungal functions of myeloid cells, as CT infected mice were not resistant to infection irrespective to strong myeloid cell recruitment to the lungs. CD11b⁺ myeloid cells were enriched from bone marrow of BALB/c donor mice (Figure 5B) and transfused intravenously to cyclophosphamide immunosuppressed (C IS) mice on day 0. This CD11b⁺ population consisted of CD11b⁺Ly6G^high neutrophils (70 ± 1%), CD11b⁺Ly6G^dim cells (5 ± 0.5%), CD11b⁺Ly6G^-Ly6C⁺ monocytes (7 ± 1%) and CD11b⁺Ly6G^-Ly6C^- non-differentiated neutrophilic and monocytic precursor cells (18 ± 4%, Supplementary Figure S5). On day +1 we infected mice with a lethal dose of 2 × 10⁵ A. fumigatus conidia and monitored their survival (Figure 5C). C IS mice, which had received an adoptive CD11b⁺ myeloid cell transfer, were resistant to a lethal infection dose, whereas, immunosuppressed and infected (control) mice were unable to clear the infection and died within 4 days after infection (Figure 5C). To determine whether transfused CD11b⁺ myeloid cells recruit to the infected lungs and directly impair A. fumigatus growth, we performed an adoptive cellular transfer experiment with transgenic firefly luciferase expressing CD11b+ myeloid cells enriched from a BALB/c.L2G85 luciferase reporter mouse (Cao et al., 2004; Beilhack et al., 2005) and infected with TdTomato expressing A. fumigatus conidia with the same experimental settings as described in Figure 5. The transfused CD11b⁺ cells were detected in C IS infected and not infected lungs 3 days p.i. with ex vivo bioluminescence imaging (Chopra et al., 2015). Lungs from transfused and infected C IS mice contained many CD11b⁺ cells, whereas lungs from transfused and not infected C IS mice did not show CD11b⁺ myeloid cells (Figure 5D). To determine whether recruited CD11b⁺ myeloid cells interacted with A. fumigatus, we performed fluorescence microscopy on C IS transfused and infected lung sections. Luciferase expressing CD11b⁺ cells were found in close proximity to A. fumigatus and fungal hyphal formation was impaired at 3 days p.i. (Figure 5E). These results indicate that adoptively transferred CD11b⁺ cells recruit to the infected lungs and support locally the control of A. fumigatus fungal growth. In contrast, cortisone and cyclophosphamide immunosuppressed (CC IS) mice, which had received adoptively transferred CD11b⁺ myeloid cells, could not clear the infection and died within 5 days after infection (Figures 6A,B). CC IS mice which had received luciferase expressing CD11b⁺ myeloid cells showed strong influx of these cells to the lungs upon infection (Figure 6C). However, these strongly recruited cells failed to control of A. fumigatus growth in CC IS infected lungs (Figure 6D). These striking results indicate that corticosteroid treatment might either have caused tissue damage to recipient mice or affected the protective function of adoptively transferred CD11b⁺ myeloid cells. However, myeloid cells significantly contributed to the host anti-A. fumigatus defense as adoptive transfer of CD11b⁺ myeloid cells alone rescued cyclophosphamide immunosuppressed mice from lethal A. fumigatus infection.