The pH-responsive PacC transcription factor of Aspergillus fumigatus governs epithelial entry and tissue invasion during pulmonary aspergillosis

Abstract

Destruction of the pulmonary epithelium is a major feature of lung diseases caused by the mould pathogen Aspergillus fumigatus. Although it is widely postulated that tissue invasion is governed by fungal proteases, A. fumigatus mutants lacking individual or multiple enzymes remain fully invasive, suggesting a concomitant requirement for other pathogenic activities during host invasion. In this study we discovered, and exploited, a novel, tissue non-invasive, phenotype in A. fumigatus mutants lacking the pH-responsive transcription factor PacC. Our study revealed a novel mode of epithelial entry, occurring in a cell wall-dependent manner prior to protease production, and via the Dectin-1 β-glucan receptor. ΔpacC mutants are defective in both contact-mediated epithelial entry and protease expression, and significantly attenuated for pathogenicity in leukopenic mice. We combined murine infection modelling, in vivo transcriptomics, and in vitro infections of human alveolar epithelia, to delineate two major, and sequentially acting, PacC-dependent processes impacting epithelial integrity in vitro and tissue invasion in the whole animal. We demonstrate that A. fumigatus spores and germlings are internalised by epithelial cells in a contact-, actin-, cell wall- and Dectin-1 dependent manner and ΔpacC mutants, which aberrantly remodel the cell wall during germinative growth, are unable to gain entry into epithelial cells, both in vitro and in vivo. We further show that PacC acts as a global transcriptional regulator of secreted molecules during growth in the leukopenic mammalian lung, and profile the full cohort of secreted gene products expressed during invasive infection. Our study reveals a combinatorial mode of tissue entry dependent upon sequential, and mechanistically distinct, perturbations of the pulmonary epithelium and demonstrates, for the first time a protective role for Dectin-1 blockade in epithelial defences. Infecting ΔpacC mutants are hypersensitive to cell wall-active antifungal agents highlighting the value of PacC signalling as a target for antifungal therapy.

Figure 1. Impact of pacC deletion upon A. fumigatus growth.

(A and B) Colonial growth phenotypes on supplemented DMEM, 48 hr of growth, 10³ conidia. (C) Immunofluorescence-mediated visualisation of A. fumigatus growth in co-culture with A549 epithelial cells, 16 hours, supplemented DMEM pH 7.4. Fungal cells are labelled with calcoflour white and host cells are labelled with FITC-conjugated concanavalin A, 10⁵ spores/ml, 200× magnification. (D) Quantitation of cell size (μM²) at 16 hr of co-culture with A549 cells, 10⁴ spores/ml, growth conditions as for panel C. (E) Branching frequency of A. fumigatus hyphae, 10⁴ spores/ml, growth conditions as for panel C.

Figure 2. PacC is required for pathogenicity and epithelial invasion in leukopenic mice.

(A) Kaplan-Meier curve for murine survival (n ≥ 9) after infection with 6 × 10⁵ and 5 × 10⁴ spores for ATCC46645 and CEA10 respectively. (B) Histopathology of leukopenic murine lungs after 4, 8, 12, hr of infection with ATCC46645 and ΔpacC ^ATCC (10⁸ spores), Grocott's Methenamine silver (GMS) and light green staining, 200× magnification. (C) ΔpacC ^ATCC hyphae (black arrows) are unable to penetrate the respiratory epithelium (RE), BA indicates bronchial airway space, 20 hr post-infection, 400× magnification.

Figure 3. Quantitative analysis of epithelial integrity.

(A) Percent detachment, relative to PBS challenge, quantified from a similar experiment as that depicted in Figure 1C. Significance was calculated relative to PBS-challenged A549 cells. (B) Quantification of A549 damage by ⁵¹Cr release assay upon infection with A. fumigatus (5 × 10⁵ spores/ml) for 16, 20 and 24 hr. A and B: technical and biological triplicate, unpaired t test. *** p<0.001, ** 0.001 <p<0.01, and * 0.01 <p<0.05.

Figure 4. Distinct and sequential PacC-dependent activities elicit epithelial decay.

(A) Percent detachment, relative to PBS challenge, of A549 cells exposed to antipain-treated and non-treated fungal culture supernatants (48 hr cultures, 10⁵ spores/ml). (B) Integrity of A549 monolayers following exposure to fungal culture supernatants (16 hr cultures, 10⁵ spores/ml). 200× magnification. (C) Percent aberrant morphology, amongst fungus-proximal ECs, per unit of hyphal length, relative to wild-type challenged monolayers. Monolayers were incubated for 18 hr or 36 hr (ΔpacC mutants). A and B: technical and biological triplicates and C: performed twice in triplicate, 1-way ANOVA test, *** p<0.001 and ** 0.001 <p<0.01.

Figure 5. Aberrant morphology of A. fumigatus ΔpacC mutants impacts contact-mediated epithelial decay.

(A) Calcofluor white staining of A. fumigatus strains after growth for 17–18 hr in AMM (pH 6.5) at 28°C. (B) Quantification of calcofluor white-mediated fluorescence (arbitrary units) as measured from panel A. (C) Percentage, in cell wall extracts relative to total weight, of chitin. (D) Percentage, in cell wall extracts relative to total weight, of glucan. (E) Percentage, in cell wall extracts relative to total weight, of mannan. (F) Percent detachment, relative to PBS challenge, of A549 cells after co-incubation with cell wall extracts (10 µg/ml). Significance was calculated relative to PBS-challenged monolayers. (G) Percent aberrant morphology, amongst thimerosal-killed fungus-proximal ECs, per unit of hyphal length, relative to wild-type challenged monolayers. B and C: biological and technical triplicates. D and E: performed twice in triplicate. B, C and D: unpaired t test. E: 1-way ANOVA test. *** p<0.001 and ** 0.001<p<0.01.

Figure 6. ΔpacC mutants are less avidly internalised by A549 and spore internalisation depends on Dectin-1.

(A) Percent internalisation by nystatin protection assay. Epithelial monolayers were incubated with 10⁶ spores/ml for 4 hr, n  =  9 for wild types and ΔpacC. (B) Percent internalisation, when blocking the Dectin-1 receptor (0.3 µg/ml α-Dectin-1 antibody) or actin-mediated internalisation (0.2 µM CD), or in the presence of an isotype control antibody (0.3 µg/ml) technical and biological duplicates. (C) Percent detachment, relative to PBS challenge, of A549 upon infection with A. fumigatus (10⁵ spores/ml) for 16 hr and inhibition of actin-mediated internalisation via Cytochalasin D (CD, 0.2 µM), when blocking the Dectin-1 receptor (0.3 µg/ml α-Dectin-1 antibody) or in the presence of an isotype control antibody (0.3 µg/ml), technical and biological triplicates. (D) Use of a soluble Fc Dectin-1 protein for immunofluorescence-mediated imaging of β-glucan in A. fumigatus spores, 4 hr, supplemented DMEM. (E) A. fumigatus spore size (μM²), 4 hr, supplemented DMEM. A, B, C: unpaired t test (unless otherwise indicated) *** p<0.001, ** 0.001<p<0.01, and * 0.01<p<0.05.

Figure 7. Internalisation and Dectin-1 protect murine pulmonary epithelia from A. fumigatus-mediated damage.

(A) Histopathology of Dectin-1^−/− and Dectin-1^+/+ leukopenic mice (n  =  3) infected with CEA10 (10⁸ spores, 24 hr of infection, GMS and light green staining, 200× magnification). (B) LDH/BCA fold change in BALs from Dectin-1^−/− and Dectin-1^+/+ leukopenic mice (n  =  3) infected as in B, technical triplicates. (C) S100B western blotting on lung homogenates from Dectin-1^−/− and Dectin-1^+/+ leukopenic mice infected as in B. Densitometry plot where pooled densitometry values for S100B were normalised on pooled densitometry values for actin and background subtracted for the saline controls. (D) Macrophages and neutrophils (cells/ml) in BALs from immunocompetent mice infected with ΔpacC ^CEA10 mutant and CEA10 (n  =  4, 10⁶ spores, 24 hr of infection). B and C: unpaired t test, *** p<0.001 and * 0.01<p<0.05.

Figure 8. ΔpacC mutants are hypersensitive to cell wall active drugs.

(A) In vitro susceptibility testing to caspofungin (CFN). Determination of CFN MEC (μg/ml), twice in triplicates, 1.25 × 10⁵ spores, 48 hr. (B) A. fumigatus growth in the presence of 16 µg/ml CFN. 400× magnification. (C) Histopathology from leukopenic mice infected with CEA10 or ΔpacC ^CEA10 (3 × 10⁴ spores) in the presence or absence of 5 mg/kg CFN. Lungs sections at 24 hr of infection, GMS and light green staining, 200× magnification. (D) In vivo susceptibility testing to caspofungin. Fungal burden at 48 hr of infection, calculated by CFUs, from leukopenic mice infected as from C. Results are expressed as CFUs per gram of lung tissue. A and D: unpaired t test, *** p<0.001, **.001<p<0.01, and * 0.01<p<0.05.