Regulation of Surfactant Protein Gene Expression by Aspergillus fumigatus in NCl-H441 Cells

Aspergillus fumigatus is an opportunistic fungal pathogen that causes serious lung diseases in immunocompromised patients. The lung surfactant produced by alveolar type II and Clara cells in the lungs is an important line of defense against A. fumigatus. The surfactant consists of phospholipids and surfactant proteins (SP-A, SP-B, SP-C and SP-D). The binding to SP-A and SP-D proteins leads to the agglutination and neutralization of lung pathogens as well as the modulation of immune responses. SP-B and SP-C proteins are essential for surfactant metabolism and can modulate the local immune response; however, the molecular mechanisms remain unclear. We investigated changes in the SP gene expression in human lung NCI-H441 cells infected with conidia or treated with culture filtrates obtained from A. fumigatus. To further identify fungal cell wall components that may affect the expression of SP genes, we examined the effect of different A. fumigatus mutant strains, including dihydroxynaphthalene (DHN)-melanin-deficient ΔpksP, galactomannan (GM)-deficient Δugm1 and galactosaminogalactan (GAG)-deficient Δgt4bc strains. Our results show that the tested strains alter the mRNA expression of SP, with the most prominent and consistent downregulation of the lung-specific SP-C. Our findings also suggest that secondary metabolites rather than the membrane composition of conidia/hyphae inhibit SP-C mRNA expression in NCI-H441 cells.


Introduction
Aspergillus fumigatus is a saprotrophic fungus and an opportunistic pathogen that produces thousands of small airborne spores (conidia) [1,2]. In immunocompetent persons, A. fumigatus conidia are cleared from the lung through different mechanisms including ciliary action of the epithelium and phagocytosis [3,4]. In immunocompromised patients, fungal conidia can survive, germinate and form a vegetative mycelium leading to a wide range of pulmonary diseases such as allergic bronchopulmonary aspergillosis, aspergilloma and even invasive pulmonary aspergillosis [3]. Despite current therapy, the mortality of pulmonary aspergillosis is reported to reach 30% to 80% [5].
The fungal cell wall is responsible for interacting with host cells. On one side, the cell wall provides a primary line of fungal defense against a hostile environment. On the other side, the cell wall components are often targets of the host's immune system during fungal infections [6]. The cell wall of A. fumigatus is composed of polysaccharides (glucans), chitin and galactomannans (GMs) [7][8][9]. The outer layer of dormant A. fumigatus conidia has an additional rodlet layer containing hydrophobic RodA proteins and an underlying dihydroxynaphthalene (DHN)-melanin layer [7,10]. The DHN-melanin layer has been revealed to serve as a cover of pathogen-associated molecular patterns (PAMPs), including β-glucan and GM, and further protects fungi against UV irradiation, reactive oxygen species (ROS) and microbial lytic enzymes [11,12]. It was shown that DHN-melanindeficient mutants produce white hydrophilic spores with an altered surface structure and reduced virulence [12][13][14].

Cell Treatments
Swollen conidia and culture filtrates were prepared as described before [42,47]. In short, to obtain swollen conidia, 10 8 freshly isolated conidia/mL were incubated in RPMI plus 3.45% MOPS (Fisher Scientific, Vienna, Austria) plus 2% glucose (Merck KGaA, Darmstadt, Germany) on an environmental shaker-incubator ES-20/60 (BioSan, Riga, Latvia) at 37 • C and maintained at 160 rpm for 2 h. Swollen conidia were further diluted for experiments to the desired concentration, as stated below. To generate fungal culture filtrates, 10 8 conidia/mL were inoculated in 200 mL Sabouraud medium (40 g/L D(+)-glucose (Merck KGaA, Darmstadt, Germany), 10 g/L peptone from casein (Carl Roth, Karlsruhe, Germany)) and kept on an environmental shaker-incubator ES-20/60 (BioSan, Riga, Latvia) for 24 h at 37 • C and 160 rpm. The suspension was further filtered using a sterile Whatman ® paper filter (Merck KGaA, Darmstadt, Germany), and mycelia were washed with double-distilled water, resuspended and incubated in 200 mL cell culture medium for 24 h at 37 • C and 160 rpm. The sterile culture filtrate medium was incubated in parallel as a control. After 24 h, the culture filtrate suspension was sterile-filtered with a 500 mL Nalgene™ Rapid-Flow™ sterile disposable filter unit with 0.2 µm PES membrane (ThermoScientific, Vienna, Austria) and stored in sterile 15 mL tubes (Corning GmbH, Kaiserslautern, Germany) at −20 • C.
For infection experiments, NCI-H441 cells were seeded into 6-well plates (Biologix, Hallbergmoos, Germany) with 10 6 cells per well and cultured for 26-28 h. Before infection with A. fumigatus conidia, cells were washed with serum-and antibiotic-free medium and incubated in this medium overnight. Then, 10 7 swollen conidia/mL were directly applied to the cells and incubated for 4 and 8 h at 37 • C and 5% CO 2 . Untreated cells were incubated in parallel as a control in each experiment. For the experiments with culture filtrates, filtrates diluted 1:5 in serum-free medium were applied on cells in 6-well plates (10 6 cells per well) and incubated for 4 and 8 h at 37 • C and 5% CO 2 . Cells treated with a control filtrate medium in the same dilution were incubated in parallel as a control.

Microscopy
The kinetics of the hyphal growth was examined by time-lapse video microscopy, as previously described [42]. In brief, 10 5 swollen conidia/mL were transferred to the 8-well Nunc™ Lab-Tek™ chamber slide system (Thermo Scientific, Vienna, Austria), and live imaging was performed in a cell incubator at 37 • C and 5% CO 2 using an ioLight Portable Microscope (ioLight, Hampshire, UK) operated with the ioLight app (Vers.: 1.1.1.483) at 37 • C. Video sequences, tracking and quantification of the velocity of hyphal growth were prepared using Fiji software [48] and the "Manual Tracking" plug-in (developed at Institute Curie by F. Cordeli for ImageJ; Orsay, France). Figures were prepared using Fiji software (based on ImageJ 1.51n, maintained by the Laboratory for Optical and Computational Instrumentation at the University of Wisconsin-Madison, Madison, WI, USA) and Adobe Photoshop (Adobe Systems Incorporated, San José, CA, USA).

Real-Time qPCR
RNA was isolated using Tri Reagent™ Solution (Invitrogen, Vienna, Austria) according to the manufacturer's instructions. RNA concentration was measured in duplicates using a NanoPhotometer NP80 (Implen, Munich, Germany). cDNA was prepared using the RevertAid First Strand cDNA Synthesis Kit (Thermo Scientific, Vienna, Austria). TaqMan gene expression master mix and TaqMan gene expression assays were used.

Statistics
All probes were isolated in three independent experiments with two technical repetitions; each experiment included an untreated control performed in the same experiment in parallel. For RT-qPCR experiments, technical duplicates were used for each probe. Data were further analyzed using a multiple unpaired T-test in GraphPad Prisma 9 (GraphPad Software, Boston, MA, USA). Only results with a p-value ≤ 0.05 were considered significant. RT-qPCR data were represented as fold change in gene expression normalized to RPLP0 rRNA expression and relative to untreated NCI-H441 cells (taken as one for each single biological and technical repetition).

Kinetics of Fungal Growth
To exclude significant differences in fungal growth between the four A. fumigatus strains and better define time points for the infection experiments, we quantified conidial germination and hyphal growth dynamics and compared ∆pksP, ∆ugm1 and ∆gt4bc to the DAL strain using time-lapse video microscopy. A total of 10 5 mL pre-swollen conidia were plated on LapTeck slides followed by imaging for 10 h. The start of conidial germination as well as the velocity of hyphal growth (defined as changes in hyphal length per minute) were quantified. Conidial germination started shortly after 5 h of imaging, and a small delay in gemination was detected in ∆ugm1 strain conidia (Figure 1a,b).
When we further tracked hyphal growth in individual conidia and quantified the velocity of hyphae growth, we observed that in the first 2 h after the start of germination, no significant differences in growth velocity were observed when comparing DAL (0.21 µm/min ± 0.03 SD) with ∆pksP (0.19 µm/min ± 0.02), ∆ugm1 (0.20 µm/min ± 0.04) and ∆gt4bc (0.19 µm/min ± 0.01) strains. At later time points, the velocity of growth increased exponentially in all strains and reached 0.64 µm/min ± 0.29 (DAL), 0.46 µm/min ± 0.30 (∆pksP), 0.36 µm/min ± 0.18 (∆ugm1) and 0.47 µm/min ± 0.33 (∆gt4bc) at the end of the video sequence. We observed a slightly increased velocity at several time points when comparing the growth dynamics of DAL with ∆pksP, ∆ugm1 and ∆gt4bc strains (marked with asterisks in Figure 1c); the most prominent difference was observed between the DAL and ∆ugm1 strain (green asterisks in Figure 1c). This observation is in line with previous studies showing a reduced growth phenotype for the ∆ugm1 mutant [50]. Based on the kinetics of fungal growth in our experiments, we defined two time points of special interest for the following RT-qPCR experiments: 4 h (conidia not yet germinated in any of strains) and 8 h (developed germ tubes in all tested strains). Depicted are mean ± SD of mean as a column, mean as value and number of quantified conidia, n; (c) velocity of hyphal growth. For graphical representation, all germinating conidia were plotted to the first tick of the x-axis with 10 min between data points. Each tick depicts velocity mean ± SD of mean in µ m/min. In (b,c) significant differences (p ≤ 0.001) between A. fumigatus strains are marked with *** according to the color scheme.
When we further tracked hyphal growth in individual conidia and quantified the velocity of hyphae growth, we observed that in the first 2 h after the start of germination, no significant differences in growth velocity were observed when comparing DAL (0.21 µ m/min ± 0.03 SD) with ΔpksP (0.19 µ m/min ± 0.02), Δugm1 (0.20 µ m/min ± 0.04) and Δgt4bc (0.19 µ m/min ± 0.01) strains. At later time points, the velocity of growth increased exponentially in all strains and reached 0.64 µ m/min ± 0.29 (DAL), 0.46 µ m/min ± 0.30 (ΔpksP), 0.36 µ m/min ± 0.18 (Δugm1) and 0.47 µ m/min ± 0.33 (Δgt4bc) at the end of the video sequence. We observed a slightly increased velocity at several time points when comparing the growth dynamics of DAL with ΔpksP, Δugm1 and Δgt4bc strains (marked with asterisks in Figure 1c); the most prominent difference was observed between the DAL and Δugm1 strain (green asterisks in Figure 1c). This observation is in line with previous studies showing a reduced growth phenotype for the Δugm1 mutant [50]. Based on the kinetics of fungal growth in our experiments, we defined two time points of special interest for the following RT-qPCR experiments: 4 h (conidia not yet germinated in any of strains) and 8 h (developed germ tubes in all tested strains).

Infection of Human NCI-H441 Cells with A. fumigatus Conidia
The lung adenocarcinoma human NCI-H441 cell line resembles bronchiolar epithelial Clara cells in phenotype. It was previously used to evaluate the protein and gene expression of SP [51,52]. In line with previous studies, we detected high levels of SP-A1, SP- Depicted are mean ± SD of mean as a column, mean as value and number of quantified conidia, n; (c) velocity of hyphal growth. For graphical representation, all germinating conidia were plotted to the first tick of the x-axis with 10 min between data points. Each tick depicts velocity mean ± SD of mean in µm/min. In (b,c) significant differences (p ≤ 0.001) between A. fumigatus strains are marked with *** according to the color scheme.

Infection of Human NCI-H441 Cells with A. fumigatus Conidia
The lung adenocarcinoma human NCI-H441 cell line resembles bronchiolar epithelial Clara cells in phenotype. It was previously used to evaluate the protein and gene expression of SP [51,52]. In line with previous studies, we detected high levels of SP-A1, SP-A2 (Ct values < 19 in control samples) and SP-B (Ct < 23) and low levels of SP-C (Ct < 30) and SP-D (Ct < 29) gene transcripts [51].

Treatment of Human NCI-H441 Cells with A. fumigatus Culture Filtrates
Next, NCI-H441 cells were treated with culture filtrates collected from the DAL A. fumigatus strain. The treatment of cells with undiluted culture filtrates or 1:3 dilution with growth medium caused immediate cell detachment (data not shown). Furthermore, 1:5 dilution of filtrates in the growth medium was well tolerated by cells and, therefore, used in our experiments.
After 4 h of treatment with culture filtrates, a small downregulation of the SP-C gene expression was observed (0.82 ± 0.03, p-value < 0.001), while SP-A1, SP-A2, SP-B and SP-D mRNA were not regulated. However, treatment with culture filtrates for 8 h resulted in a moderate downregulation of all SP mRNA, with SP-C being the most significant ( Figure 3, Table 1). * p ≤ 0.05, ** p ≤ 0.01, *** p ≤ 0.001.

Treatment of Human NCI-H441 Cells with A. fumigatus Culture Filtrates
Next, NCI-H441 cells were treated with culture filtrates collected from the DA fumigatus strain. The treatment of cells with undiluted culture filtrates or 1:3 dilution growth medium caused immediate cell detachment (data not shown). Furthermore dilution of filtrates in the growth medium was well tolerated by cells and, therefore, in our experiments.
After 4 h of treatment with culture filtrates, a small downregulation of the SP-C expression was observed (0.82 ± 0.03, p-value < 0.001), while SP-A1, SP-A2, SP-B and D mRNA were not regulated. However, treatment with culture filtrates for 8 h result a moderate downregulation of all SP mRNA, with SP-C being the most significant (Fi 3, Table 1).  All data are normalized with RPLP0 rRNA expression and given as relative to untreated control. Cells were treated with culture filtrates obtained from A. fumigatus DAL strain for 4 h (left graph) and 8 h (right graph). Columns depict mean ± SE of the mean (* p ≤ 0.05), (** p ≤ 0.01), (*** p ≤ 0.001). See also Table 1.

Discussion
In this study, we evaluated the gene expression of the pulmonary SP (SP-A1, SP-A2, SP-B, SP-C and SP-D) in human NCI-H441 cells following an A. fumigatus infection. We tested the effect of DAL, DHN-melanin-deficient, GAG-deficient and GM-deficient mutant A. fumigatus strains as well as culture filtrates incubated with cells for 4 and 8 h.
Our results show that a short-term infection of NCI-H441 cells did not regulate the SP mRNA. Instead, 8 h of infection caused consistent downregulation of the SP-C gene. Within this time, conidial attachment, germination and hyphal growth took place [47,53]. Thus, in comparison to short-term experiments with conidia, NCI-H441 cells were exposed to another biochemical composition of the fungal cell wall when PAMPs, e.g., β-glucan, chitin, GM and GAG, were exposed to lung cells [10,15]. Surprisingly, we did not observe a significant downregulation of SP-A and SP-D mRNA and no difference was identified between infections with DHN-melanin-, GAG-and GM-deficient strains.
Previous studies demonstrated that infection with A. fumigatus caused the differential regulation of surfactant gene expression in vitro and in vivo. In line with our results, a downregulation of lung-specific SP-C mRNA was observed in cells and mouse models [36,43,44]. A study of A. fumigatus-induced allergic airway inflammation in mice showed decreased mRNA and protein expression of SP-B and SP-C that was accompanied by an upregulated SP-D protein expression without changes in mRNA levels [36]. Protein and mRNA levels of SP-A were not changed in this model [36]. Surprisingly, we did not identify significant differences between the four mutant strains: the downregulation of SP-C mRNA was independent of the fungal strain. A more recent report revealed the downregulation of SP-D mRNA in human alveolar A549 cells after 6 h of incubation with A. fumigatus conidia [44]. In our preliminary experiments, we tested a long-term culture model of A549 cells as a model to consider the deferential regulation of SP in infection settings. This model was shown to express increased levels of SP, as compared to regular A549 cell cultures [54]. Infection of these cells with A. fumigatus swollen conidia did not cause any consistent changes in SP-A and SP-D mRNA in our experiments (unpublished data).
Our data suggest that secondary metabolites of A. fumigatus, rather than contact/binding to the fungal cell wall polysaccharides, may alter the mRNA expression of SP. In line with this hypothesis, we observed that the treatment of NCI-H441 cells with culture filtrates resulted in a moderate decrease in gene expression of all SP mRNA when cells were incubated for 8 h. Short exposure to culture filtrates caused a small inhibition of SP-C mRNA, whereas changes in other SP genes were not statistically significant. Culture filtrates gained from A. fumigatus were previously shown to have a cytotoxic effect on alveolar cells [55] and inhibit a pulmonary immune response [21,56,57]. The effect of the filtrates on SP expression has not been reported before. Importantly, we used diluted filtrates that did not cause cell detachment or apoptosis and observed a moderate inhibiting effect on all tested SP genes.
The major function of SP-C is to reduce the surface tension at the air-liquid alveolar interface by regulating lipid adsorption and transfer of lipids between surfactant membranes, thus providing surfactant film stability [58]. Data from patients and mice models showed that SP-C has an important antimicrobial and anti-inflammatory function in the lung. Thus, patients with mutations in the SP-C gene developed familial interstitial lung disease [34]. Knockout of the SP-C gene in mice resulted in progressive lung inflammation [59]. These mice were also more susceptible to respiratory syncytial virus and Pseudomonas aeruginosa and showed robust inflammation in the lungs after gram-negative bacterial infection [37,60]. The binding of SP-C to bacterial lipopolysaccharide has been observed in vitro [35,59,61]. Several studies showed that exogenous synthetic or natural surfactant treatment reduced symptoms of lung infection and improved bacterial clearance [62].
In vitro experiments demonstrated that treatment with Survanta (a natural surfactant product containing SP-B and SP-C and phospholipids) inhibited proinflammatory cytokine release from LPS-stimulated human alveolar macrophages [63]. Thus, decreased levels of SP-C may impair the resolution of lung inflammation and stimulate progressive interstitial disease [59]. The molecular mechanisms of SP-C inhibition by A. fumigatus observed in our study and by others remain unclear.
To summarize, our data show consistent downregulation of lung-specific SP-C mRNA in NCI-H441 cells caused by A. fumigatus. Further elucidating this inhibitory effect in more detail is important to better understand how potentially infectious pathogens overcome the defense mechanisms in the human lung. Moreover, data on the combinational use of pulmonary surfactants and antifungal agents against A.fumigatus have only started to arise [42,64] and should be investigated in future studies.