Abstract
Aspergillus fungi are opportunistic pathogens that affect a large number of people worldwide. Many aspects of Aspergillus spp. pathogenesis toward humans are known, but their ability to enhance their infectious potential by manipulating the environmental pH of its host has not been considered yet. In this study, we tested the hypothesis that by producing oxalic acid, Aspergillus niger can manipulate pH during lung infection and thus, interfering with this process could limit pathogenicity. To test this hypothesis, we co-cultured A. niger with oxalotrophic bacteria in increasingly complex testing systems (Petri dishes and 3D-cell cultures systems). In in vitro tests, oxalotrophic bacteria limit oxalic acid production and suppressed the pH shift induced by A. niger. In 3D-cell cultures (Transwells® and Bronchioles-on-a-chip), A. niger also modified pH, Ca2+ and oxalic acid concentrations. Co-inoculation with as little as 10 cells of the oxalatrophic bacterium strongly inhibited the germination and development of A. niger and returned each of the three parameters to the baseline physiological values of uninfected cells. This biocontrol interaction between oxalotrophic bacteria and oxalate-producing A. niger could represent a paradigm shift in the fight against opportunistic fungal pathogens, where the host environment is rendered less permissive to fungal development.
Introduction
Fungal diseases are estimated to kill more than 1.5 million people every year (1, 2). Over the last three decades, the increase in the number of at-risk individuals has correlated with an intensification in the burden of fungal disease on human health (3). As the size of at-risk populations (e.g. immunosuppressed patients) is expected to keep increasing in the future (2), tackling fungal pathogenesis is urgent. However, only a very limited number of antifungal drugs are used nowadays to control fungal pathogens. Because of this restricted chemical arsenal, the same classes of molecules are used in human health, animal husbandry, and agriculture, leading to the rise, and rapid spread, of resistance in fungal pathogens that can affect both plant and animal hosts (4). In spite of this, tackling fungal diseases has been largely neglected up to now (5).
The most prevalent fungal pathogens affecting humans are airborne opportunists such as Aspergillus spp., Cryptococcus spp., Pneumocystis spp., and human-associated commensals like Candida albicans (6, 7). These fungal species are responsible for approximately 90% of the deaths due to fungal infection (1). The ecology of all these organisms plays a very significant role in their ability to transition to pathogenic lifestyles. For instance, the remarkable plasticity in the ecology and stress-response of Aspergillus spp. is believed to form the basis of its success as an opportunistic pathogen (8, 9).
Although many aspects of the ecology of Aspergillus spp. have been connected to pathogenicity (9), its ability to manipulate pH via the secretion of low molecular weight organic acids (LMWOA), and in particular, oxalic acid, has been largely ignored in the context of human pathogenesis. Several clinical reports have shown the presence of calcium oxalate (CaOx) crystals in pulmonary aspergillosis in animals and humans (10-15). Oxalic acid and oxalate crystals are thought to directly cause damage to the host tissues (including pulmonary blood vessels), and to generate free radicals which can harm cells indirectly (14). A recent case report of invasive pulmonary aspergillosis in a 69-year old man with lymphoma and pneumonia indicated the presence of CaOx crystals around blood vessels and within the blood vessel walls. This suggests a potential mechanical role of oxalate crystals in the angioinvasion of Aspergillus (15). However, a link between oxalic acid production and pathogenicity has not yet been made in fungi from this genus or in any other fungal human pathogen. On the contrary, oxalic acid production has been widely acknowledged as a pathogenicity factor in fungal plant pathogens, such as Sclerotinia sclerotiorum and Botrytis cinerea (16, 17) where pH manipulation and calcium chelation plays a direct role in pathogenesis (18-20). Both acidification and cation complexation in the local environment is exploited by plant pathogens to weaken the cell wall structure, facilitate infection, inhibit plant defenses, and induce programmed cell death (19, 21).
Oxalic acid is a ubiquitous compound in the environment and is thought to have a central role in fungal metabolism (19). Its production and consumption by microorganisms have been directly associated with pH regulation in soil (22). In such ecosystems, oxalic acid is often found complexed with divalent cations, especially calcium (23). Despite its chemical stability (Ksp CaOx monohydrate = 2.32 × 10−9), CaOx rarely accumulates (24, 25). This is because oxalate is used by soil oxalotrophic bacteria as carbon and energy sources. The transformation of a strong organic acid into a weaker one (oxalic acid versus carbonic acid; pKa = 1.25 and 4.14, respectively), leads to a local increase in soil pH. This overall process has been coined out in the oxalate-carbonate pathway (26). Moreover, oxalic acid plays a key role in bacterial:fungal interactions acting as a signaling cue by bacteria in order to localize fungi and to establish different types of trophic interactions with them (27-29).
In this study we wanted to evaluate whether the metabolic processes associated to oxalic acid during Aspergillus spp. infection could parallel those occurring during its natural cycling in the oxalate-carbonate pathway in soils. Indeed, humans can be seen as a complex ecosystem governed by the same ecological principles affecting any other ecosystem (30). Thus, the ability of oxalotrophic bacteria to degrade oxalic acid produced by Aspergillus spp., and control the subsequent pH manipulation and calcium chelation, would result in a mechanism to control biologically this opportunistic fungal pathogen (Fig. 1). To test our hypothesis, we selected Aspergillus niger as a model. This organism is regarded as a safe relative of the more infectious Aspergillus fumigatus (31), and has been extensively studied for its ability to produce oxalic acid (32, 33). It is also a known agent of aspergillosis in humans, being responsible for 5% of the cases (9, 34). We first confirmed that A. niger secretes oxalic acid and assessed the effect of co-cultivation with oxalotrophic bacteria on pH and oxalic acid concentration, and on the inhibition of A. niger’s growth. This was done in vitro and on human bronchial epithelial cell (HBEC) cultures using two complementary 3D-cell cultures systems (Transwell® inserts and bronchioles-on-a-chip (BoC)).
Results
Detection of oxalic acid produced by Aspergillus niger in different culture conditions
Since the metabolism of fungi can change significantly depending on the nutritional conditions of the growth medium, we first tested whether our fungal strain produced oxalic acid only or a mixture of different organic acids in culture media differing in their trophic conditions. While A. niger is known to produce high amounts of oxalic acid, it is also well known to produce other LMWOA such as citric acid (32, 35-37). An Ultra-High-Performance Liquid Chromatography (UHPLC) analysis revealed that in the conditions we tested, A. niger produced oxalic acid, only (Fig. S1A). The absence of other organic acids is not surprising as the conditions for other LMWOA production are highly specific (e.g. carbon concentration above 50 g/L carbon and pH < 3; (38, 39) and are not provided in the conditions we tested. Acidification and the presence of CaOx crystals could also be detected in water yeast agar (WYA) (Fig. S1B). We thus concluded that our A. niger strain consistently produced oxalic acid and acidified the pH of its medium under laboratory growth conditions.
Confrontation assays between A. niger and selected bacterial strains
We assessed how the presence of oxalotrophic bacteria impacted A. niger growth and pH evolution in different growth media. Two oxalotrophic bacteria (Cupriavidus necator and C. oxalaticus) and one non-oxalotrophic bacterium (Pseudomonas putida) were used in confrontation assays. In malt agar diluted 10 times (MA 1/10), A. niger colonized the entire Petri dish, including the area in which the bacterial inocula were applied. All three inoculated bacteria did not survive in the area of interaction with the fungus (Fig. S2A). In Reasoner’s 2 agar (R2A), A. niger mycelia did not colonize the area beyond the barrier formed by the bacterial inocula, but some hyphae were still able to grow beyond the bacterial inoculation zone and develop into microcolonies in the co-culture with P. putida (Fig. S2B). In WYA, mycelial growth was also restricted to the area delimited by the inocula in the co-culture with the two oxalotrophic bacteria. On the other hand, P. putida did not survive the interaction with the fungus, which instead, colonized the entire plate. Moreover, in WYA medium, which contained a pH indicator, the acidification of the medium by the fungus grown alone or in co-culture with the non-oxalotrophic bacterium was clearly visible. In contrast, in the co-cultures with oxalotrophic bacteria, the pH of the medium did not change. The control over fungal growth was particularly remarkable in the case of C. oxalaticus (Fig. S2C). Given the impact of the medium on the control of fungal growth, we repeated the confrontation assays in Air-Liquid Interface (ALI) medium, which is the medium used for differentiation of lung cells (40). The non-oxalotrophic bacterial model (P. putida) acidified the medium when grown alone, while the oxalotrophic bacteria (C. necator and C. oxalaticus) did not acidify the medium (Fig S3A). C. oxalaticus was found to not only control mycelial growth but also inhibit conidia germination (Fig. S3B). Moreover, the presence of C. oxalaticus in co-culture with the fungus stabilized the pH of the culture medium at a neutral pH (Fig. S3C), consistent with decreased oxalic acid concentration in the medium (Fig. S3D). We therefore conclude that oxalotrophic bacteria have a significant inhibitory effect on A. niger growth in all media tested.
Establishment of a dose-response curve on submerged cultures
In order to obtain evidence that oxalotrophic bacteria control the growth of A. niger in an animal-free lung infection model that is compatible with the 3R principles of animal experimentation (41), we set up experiments on human bronchial epithelial cell (HBEC) cultures. We established a dose-response curve for increasing conidial and/or bacterial loads on HBECs in submerged cultures, to identify the optimal load to perform experiments in Transwells® and BoC systems. After 24 h, the overall size, shape and integrity of the lung cells changed at an absolute load of 500 conidia and above. The HBECs shrank in size due to actin agglomeration. Moreover, from a conidial load of >=1000, fungal growth also had an adverse effect on tissue integrity (Fig. S4). The same experiment was performed with C. oxalaticus and P. putida. A strong morphological change was induced by a total load of 500 bacterial cells and above for the former (Fig. S5), and as little as 10 cells for the latter (Fig. S6), which was not used further. The HBECs co-cultured with C. oxalaticus became rounder, and actin agglomeration increased compared with the cells-only control (Fig. S5).
To analyze the effect of co-culturing A. niger with the oxalotrophic bacterium on HBECs integrity, we performed a test with 10 and 500 conidia confronted with 10 bacterial cells. After 72h, A. niger induced morphological changes (size reduction and actin agglomeration), with a stronger effect for 500 conidia, confirming the results obtained at 24h. With the co-inoculation of as few as 10 C. oxalaticus cells, the morphology of the HBECs was similar to the morphology of HBECs of the bacteria-only control, suggesting the inhibition of fungal development (Fig. S7). We concluded that a conidial and bacterial load of 10 conidia/cells was ideal to monitor the interaction of A. niger and C. oxalaticus in differentiated HBECs in Transwells® and BoC systems.
Biocontrol assay of A. niger bronchial cells infection by C. oxalaticus
After establishing a dose-response curve on HBECs in submerged cultures, the effect of inoculation of 10 A. niger conidia alone or in co-culture with 10 C. oxalaticus cells was assessed in differentiated bronchial tissue in Transwell® inserts and BoC systems. In the presence of the fungus alone, changes in three key environmental factors were observed: pH, Ca2+ concentration, and concentration of soluble oxalic acid. The pH dropped significantly from 7.5 down to 4.5 in Transwells® and from 7.3 to 6.8 in BoC systems (Fig. 3A). Ca2+ concentrations changed from 1 mM to around 0.2 mM in both culture systems (Fig. 3B). The level of soluble oxalic acid produced by A. niger dropped from 500 μM (cells alone) to 75 μM (cells inoculated with the fungus) (Fig. 3C). In contrast, pH, Ca2+ and free oxalic acid levels were statistically indistinguishable when oxalotrophic bacteria were co-cultured with the fungus compared to the controls with lung cells alone or with bacteria. In addition, CaOx crystals were observed in the cultures in which the fungus developed (Fig. 3D), but not when the fungus was co-cultured with oxalotrophic bacteria (Fig. 3E). This suggests that the lower levels of soluble oxalic acid measured in the treatment with the fungus were likely the result of complexation of oxalic acid and Ca2+, and corroborates the pH and Ca2+ concentration data. Moreover, the absence of CaOx crystals when the fungus was in co-culture with C. oxalaticus agrees with the pH, Ca2+ concentration and soluble oxalic concentrations measured. These results validated our hypothesis that oxalotrophic bacteria can be used to manipulate the microenvironment created by A. niger. In addition to the changes in the environmental parameters measured above, we also observed a cytopathic effect when conidia of A. niger developed into mycelia (Fig. 4A and B). This cytopathic effect resulted in the destruction of the bronchial epithelium. Lactate dehydrogenase (LDH) activity were elevated in response to the presence of the foreign oxalotrophic bacteria (Fig. 4C), something that needs to be addressed for any future therapeutic application.
Genomic potential for oxalic acid production in other Aspergillus spp
We performed a genomic screening of orthologous genes to the oxaloacetate acetylhydrolase (oahA) and the oxalate/formate antiporter (genes involved in oxalic acid production in A. niger) in genomes available in the Aspergillus Genome Database (AspGD). The genomic screening revealed that orthologs of both genes are found in multiple Aspergillus spp. and are highly conserved across diverse species (Fig. 5). In the case of the orthologs to the oxalate/formate antiporter, they were conserved to a lesser extent (Fig. S8). This genomic analysis confirmed that diverse Aspergillus spp. possesses the genes necessary to produce and secrete oxalic acid.
Discussion
Here we present a biological interaction between A. niger and oxalotrophic bacteria that results in the biological control of A. niger, preventing infection in 3D-lung cell tissues (Transwells® and BoC). The direct consequence of acidification through oxalic acid production by A. niger was the decrease in free Ca2+ and subsequent precipitation of CaOx crystals. CaOx crystals are well known to occur in lung tissues upon infection by A. niger (10, 11, 13). Presumably, by consuming the oxalate produced by A. niger, the oxalotrophic bacterial species C. oxalaticus blocks the subsequent decrease in pH and formation of CaOx crystals observed in the absence of the bacterium. To obtain a direct confirmation of the role of oxalic acid in the manipulation of pH during lung infection, the use of a non-oxalate-producer A. niger mutant would be indispensable. Such mutants (oahA gene) are described in the literature and exhibited a decreased acidification of the culture medium and reduced extracellular protease activity (32, 42). After multiple failed attempts to obtain the published mutants by addressing the corresponding scientific teams, we attempted to construct a non-oxalate-producer mutant of our A. niger strain using CRISPR-Cas9 gene editing. However, this was unsuccessful due to multiple targets of the sgRNA probes and thus could not be included in this study.
Oxalate-degrading bacteria are known inhabitants of the human gut, where they perform the key function of degrading dietary oxalate (43). These species have also been used as probiotics for the treatment of hyperoxaluria (high oxalate in urine) and the management of kidney stones (43, 44). While oxalate-degrading bacteria are well characterized in the gut, this is not the case of the lung. Although considered sterile for a long time, the lung is now known to harbor a diverse microbiota (45, 46). Oxalate-degrading capabilities have been previously reported in strains of the genera Lactobacillus (47), Streptococcus (48), Prevotella (49, 50) and Veillonella (50), all of which are reported as components of the lung microbiota. However, assessing the oxalotrophic potential of the lung microbiome is something that still need to be accomplished.
The genomic analysis of multiple Aspergillus spp. suggests that oxalotrophy could also be relevant to other Aspergillus causing pulmonary aspergillosis (5). The presence of CaOx crystals during infection by A. fumigatus has been reported in the literature (8, 11, 14, 51). Accordingly, we found orthologs of the oxaloacetate acetylhydrolase (OAH) and the oxalate/formate antiporter of A. niger in the genomes of two well characterized model A. fumigatus strains Af293 and A1163 (52), suggesting the production of oxalic acid by this pathogen and the potential of using oxalotrophic bacteria in fungal species more relevant for human health. To conclude, the results presented here represent a stepping stone towards developing an alternative approach to control the development of oxalate-producing Aspergillus spp. based on the manipulation of the lung environment using bacterial:fungal interactions.
Materials and Methods
Bacterial and Fungal cultures
All bacterial and fungal strains come from the collection of the Laboratory of Microbiology of the University of Neuchâtel (LAMUN; Table 1). P. putida KT2440 was kindly provided by Dr. Arnaud Dechesne (Technical University of Denmark). C. necator JMP289 was kindly provided by Prof. Jan van der Meer (University of Lausanne). C. oxalaticus Ox1 was tagged in-house using insertion with a MiniTn7 system. Table 2 summarizes all the media used. Bacterial strains were routinely cultured on NA medium. Aspergillus niger was routinely cultured on MA medium. PDA was used for A. niger conidia production. BHIA was used to have mycelium-only colony edge without any conidia in order to prevent unwanted conidia dispersal during confrontations with bacteria.
LMWOA detection by UHPLC and by colorimetric pH indicator-based Petri dish assay
For the UHPLC analysis, 500 µl of 30 mM H2SO4 were added to 1 mL of a two-week liquid culture in malt 1/10, Reasoner’s 2, and ALI liquid media in triplicate, to obtain 20 mM H2SO4 final concentration in order to obtain a low pH for the extraction of LMWOAs and to dissolve any precipitated crystals. The samples were incubated at 60°C for two hours to dissolve precipitated metal oxalate crystals, and then centrifuged at 3000 g for 10 min. All the samples were filtered at 0.22 µm (13mm syringe filters, PTFE, hydrophilic) and 200 µl were added into HPLC vials with 250 µl conical inserts. UHPLC (Ultimate 3000 RS-Dionex, Thermo Fisher Scientific, USA) was coupled with DAD detector set at 210 ± 2 nm. A 5 µL of sample was injected onto a Prevail− organic acid column (5 µm particle size, 150 x 4.6 mm, Grace Davison Discovery Sciences, USA) with the temperature kept at 40°C. The mobile phase consisted of 50 mM phosphate buffer adjusted to pH 2.5 with phosphoric acid with a flow rate of 1 mL/min. Pure oxalic acid (Merck, Germany) was identified by the retention time and was quantified by an external standard curve, linear regression from five calibration points (0.2 to 5 mg/mL). For the culture-based assay, WYA supplemented with bromocresol purple (WYA+BP) was used as a pH indicator-containing medium. After one week of incubation at room temperature (RT), the presence of typical bi-pyramidal shaped CaOx crystals was assessed by observing a thin slice of agar medium sampled at the edge of the colony and stained with lactophenol cotton blue under a Leica DM4 B optical microscope connected to a Leica DFC7000 T camera.
Confrontation assays on solid media
Confrontations assays were performed between A. niger and P. putida, C. necator and C. oxalaticus (Table 1), on three culture media (MA 1/10, R2A and WYA+BP). A plug coming from the apical part of an actively growing A. niger colony was sampled using the wider end of a Pasteur pipette and inoculated in the center of the plates. The bacterial strains were inoculated from fresh plates as opposite lines on either side of the fungal inoculum. Plates were incubated at RT for 20 days, and pictures of the plates at 20 days were taken. Pictures of the bacterial inocula were taken using a Nikon SMZ18 epifluorescence stereoscope, connected to a Nikon DS-Ri2 camera, in order to assess the viability of the bacterial strains thanks to constitutively expressed fluorescent proteins.
Growth tests and confrontation assay in Air-Liquid Interface (ALI) medium
Bacterial growth in ALI medium was tested for 3 days at RT. To produce conidial suspensions, A. niger was cultured on PDA for 10 days at RT. Conidia were harvested using Dulbecco’s Phosphate Buffer Saline (DPBS) supplemented with 0.01% (v/v) Tween 80. Harvested conidia were washed three times with DPBS following centrifugation at 2000xg for 5 min at RT. Finally, conidia were resuspended in 2 mL DPBS and quantified with an Improved Neubauer counting chamber. Two μL of 16’000 conidia/bacterial cell per μL suspensions were inoculated in 200 μL ALI medium. To test mycelial growth in the different media, small agar plugs (approximately 3×3 mm) coming from the edge of a colony of A. niger on BHIA were used. The 96-well plate was incubated at RT for 7 days and growth was visually assessed.
Confrontation of A. niger with C. oxalaticus was performed in 100-mm Corning® tissue culture plate containing 12 mL ALI medium to allow the fungus to attach during growth. The fungal inoculum was taken from the apical part of an active colony grown on BHIA by using the wider end of a Pasteur pipette. For the confrontation assay, the fungal inoculum was placed in the medium after addition and mixing of 100 μL inoculum of an overnight culture of C. oxalaticus in ALI medium. A fungus-only plate was used as control. The plates were incubated at RT for 7 days. Oxalic acid concentration was quantified by using the Oxalic Acid Colorimetric Assay Kit (Sigma-Aldrich, Germany), following the manufacturer instructions.
Preparation and sterilization of the bronchiole-on-a-chip (BoC)
The chip was assembled as described in Hsieh et al. (53). Each unit of the culture platform was fabricated by using a layer-by-layer stacking technique (54). The devices were designed using Solid Edge 2D software (ST9, Siemens PLM Software), and each layer of the prelaminated polymeric sheet was obtained using a CO2 laser cutter (Universal Laser System). The prelaminated polymeric sheets were combined with biocompatible adhesive tapes (9122, 3M Company) with PMMA (1.5 and 3 mm thick) or PET (0.1 and 0.25mm thick). After cutting, each layer was aligned and assembled using a seam roller to complete the devices. The culture chip includes a Y-shaped apical and a basal part separated by a porous PET membrane (pore size = 0.4μm) prepared as described in Arefin et al. (55). The PET membrane was sandwiched between two PET sheets using adhesive transfer tape to create the cell culture surface. This allows nutrients to pass from the media to the cells through the porous membrane. The open design of the tissue chip makes cell seeding procedure easy and accessible.
For sterilization, each chip was placed in a 100 mm Petri dish and sterilized with 5% H2O2 solution for 1 h. The chips were then rinsed 2-3 times with sterile deionized water for 15 min between each rinse. Once all liquid was removed, the chips were let dry overnight under a laminar flow hood. The next day, the inlet and outlet of the chip were connected with a sterile tubing and rinsed 3 more times with sterile deionized water as explained before. After the last rinse, 200 μl sterile DPBS was added in the channel (apical part) and 5 mL in the basolateral part (bottom part) of the chip, and the chip was placed in a humidified incubator at 37°C with 5% CO2 overnight. The next morning, peroxide contamination was checked in each chip using a CG8+ i-STAT cartridge (Abbott, USA). Rinses with sterile deionized water and overnight incubation with sterile DPBS were repeated until peroxide was no longer detected.
Primary normal human bronchial epithelial cell culture
Primary normal human bronchial epithelial cells (Lifeline Cell Technology, USA) were expanded in a T-75 cell culture flask with vent cap (Corning, USA) in BronchiaLife™ B/T complete medium (Lifeline Cell Technology, USA) supplemented with 0.5% Phenol Red solution (Sigma-Aldrich, USA, 15 mg/L final concentration) to 70-80% confluence in a humidified incubator at 37°C with 5% CO2. Culture medium was changed every other day. Cells were used until passage 2 for all experiments. Cells were harvested by trypsinization with 0.05% Trypsin / 0.02% EDTA (Lifeline Cell Technology, USA), followed by the addition of Trypsin Neutralizing buffer (Lifeline Cell Technology, USA), and counted using a hemocytometer after centrifugation at 100xg for 5 min and resuspension of the cell pellet in BronchiaLife™ medium.
Cells were seeded at a density of 3×104 cells/well in 200 μl BronchiaLife™ medium for submerged undifferentiated tissue culture in 96-well plates, and 5×104 cells and 8.6×104 cells in 200 μL BronchiaLife™ medium for air-lifted differentiated tissue culture in the apical side of Transwell® inserts in 24-well plates (Corning, USA) and BoC devices, respectively. Transwell® inserts and BoC were first coated with collagen (30 μg/mL) prior seeding of the cells in order to allow proper cell attachment onto the porous membrane, as described in Arefin et al. (55). 600 μL of BronchiaLife™ medium was added to the basolateral side of Transwell® inserts in the 24-well plates and 3 mL in the basolateral side of the BoC device. 96-well plates, Transwell® inserts, and BoC devices were placed in a humidified incubator at 37°C with 5% CO2 for 2-3 days until confluence and formation of a monolayer of bronchial cells. For differentiated bronchial cell tissues (Transwells® and BoCs), cells were shifted to air-liquid interface by removing carefully the BronchiaLife™ medium from the apical side and replacing it by Air-Liquid Interface (ALI) Epithelial Differentiation Medium (Lifeline Cell Technology, USA) supplemented with 0.5% Phenol Red solution (Sigma-Aldrich, USA, 15 mg/L final concentration). The same was done for the medium on the basolateral side. Finally, the medium on the apical side was removed and the inserts and devices were placed in a humidified incubator at 37°C with 5% CO2 for 21 days. Medium was changed every other day as described previously. The cultures were observed daily using an EVOS™ XL Core bright field inverted microscope (Thermo Fisher Scientific, USA).
Determination of conidial and bacterial load and confrontation assay on submerged undifferentiated bronchial epithelial cell cultures
In order to determine the optimal conidial and bacterial load to be used for confrontation on bronchial tissue cultures, increasing conidial and bacterial loads were tested to assess their effect on the morphology of bronchial epithelial cells in submerged cultures. A. niger was cultured on PDA for 7 days at 37°C in order to produce conidia. A. niger conidia were then harvested as already described. Finally, conidia were resuspended in 2 mL DPBS and quantified with an Improved Neubauer counting chamber. Bacteria were cultured in BronchiaLife™ medium at 37°C overnight and quantified with an Improved Neubauer counting chamber. A stock suspension of A. niger conidia was made at 106 conidia/mL that was diluted further to obtain suspensions at 5×105 to 5×103 and 103 conidia/mL. The same was done for C. oxalaticus from a stock suspension at 105 bacterial cells/mL diluted until 103 bacterial cells/mL. 10 μL of each suspension was added to submerged undifferentiated bronchial tissue in a 96-well plate in order to have 104 to 10 conidia/well (200 μL) for A. niger, and 103 to 10 bacterial cells/well (200 μL) for C. oxalaticus. The plates were placed in a humidified incubator at 37°C with 5% CO2 for 24h. For the confrontations assay, cells were infected with 10 or 500 A. niger conidia and were put in confrontation with 10 bacterial cells. The plate was placed in a humidified incubator at 37°C with 5% CO2 for 72h. After incubation, cells were fixed, stained (actin and nucleus), and imaged as described below (Immunofluorescence staining).
Confrontations on differentiated bronchial tissues in Transwell® inserts and BoC devices
Differentiated bronchial tissues in Transwell® inserts and in BoC devices were infected with 10 μL of 103 conidia or bacterial cell to get 10 conidia or bacterial cells. Fungus and bacteria were co-inoculated for the confrontation and controls with only medium, cells, fungus or bacteria were included. Each condition was done in triplicate for the Transwell® inserts and one unique replicate for each condition was done for the BoC devices. All Transwell® inserts and BoC devices were incubated in a humidified incubator at 37°C with 5% CO2 for 72h.
Immunofluorescence staining
Undifferentiated and differentiated bronchial tissues (Transwell® and BoCs) were fixed with 100 μL 4% paraformaldehyde in DPBS for 15 min at RT. Cells were then rinsed 3 times with 200 μL DPBS, with 2 min waiting time between each rinse. Cells were permeabilized with 100 μL 0.5% Triton X-100 in DPBS for 15 min at RT and rinsed 3 times with 200 μL DPBS, with 2 min waiting time between each rinse. After that, cells were blocked with 100 μL 3% BSA in DPBS for 1h at RT. Anti-Mucin 5AC mouse monoclonal antibody (Abcam, USA, Cat.# ab218466) was prepared in DPBS (1/100). The actin stain (ActinGreen™ 488 ReadyProbes™ Reagent) and the nuclei counterstain (NucBlue™ Live ReadyProbes™ Reagent) were added to the same buffer (2 drops/mL and 1 drop/mL, respectively). Anti-Aspergillus rabbit polyclonal antibody (Abcam, Cat.# ab20419) was also added (1/200) in the staining buffer for the conditions where A. niger conidia were inoculated. Fixed cells were incubated with 100 μL buffer containing the stains and Anti-Mucin 5AC and Anti-Aspergillus antibodies overnight at 4°C. The next day, fixed cells were washed 3 times with DPBS and secondary antibodies were applied. Goat anti-Mouse IgG antibody (1/250) conjugated with Alexa Fluor 546 (Thermo Fisher Scientific, USA, Cat.# A-11003,) directed against Anti-Mucin 5AC antibody and Goat anti-Rabbit IgG antibody (1/500) conjugated with Alexa Fluor 594 (Thermo Fisher Scientific, USA, Cat.# A-11012) directed against Anti-Aspergillus antibody were prepared in DPBS. Fixed cells were incubated with 100 μL buffer containing the secondary antibodies overnight at 4°C. The following day, fixed cells were once again washed 3 times with DPBS and the membranes from the inserts were carefully cut out with a sharp knife. The membranes were mounted on a glass slide using Fluoromount-G™ Mounting Medium (Thermo Fisher Scientific, USA) and imaged with a Zeiss Axio Observer Z1 fluorescence inverted microscope (Carl Zeiss AG, Germany).
pH measurements and quantification of calcium, oxalic acid and lactate dehydrogenase (LDH)
pH measurements of the culture medium after 72h incubation were done directly after taking the samples out of the incubator using pH-indicator strips (Merck, Germany) for the Transwell® inserts samples, and CG8+ i-STAT cartridges for the BoC devices, as the pH change of the culture medium indicator was less visible. Free calcium (Calcium Colorimetric Assay, Sigma-Aldrich, Germany), free oxalic acid (Oxalic Acid Colorimetric Assay Kit), and LDH (ScienCell Research Laboratories, USA) were quantified in the culture medium using colorimetric assay kits following the manufacturer instructions.
Statistical analyses
Statistical significance of the data from the confrontations on differentiated bronchial tissues in Transwell® inserts (3 replicates, n = 3) was tested with unpaired two-tailed Student t-tests in Microsoft® Excel (Version 16.37). The statistical significance threshold was set to 5%.
Funding
This work was supported by the Novartis Foundation (FreeNovation program), the Gebert Rüf Stiftung (Grant agreement GRS-064/18) and the U.S. Department of Energy, Office of Science, Biological and Environmental Research Division, under award number LANLF59T.
Data and materials availability
All data is available in the main text or the supplementary materials.
Conflict of interests
Authors declare no conflict of interests.
Acknowledgments
We would like to thank Diego Gonzales and Ted Turlings for critical review of the paper, and Pulak Nath from the Materials Physics and Applications Division of the Los Alamos National Laboratory for providing the equipment and laboratory infrastructure for the BoC devices fabrication. BoC devices were developed under Defense Threat Reduction Agency (DTRA) interagency agreement CBMXCEL-XL1-2-0001.
Footnotes
Title, Abstract, Introduction, Results, Discussion, Materials & Methods, and References have been updated.