Microbial strains, base media and culture conditions

All P. aeruginosa isolates used in this work are shown in Table 1. Bacteria were routinely streaked onto Tryptone Soy Agar (TSA) (Sigma-Aldrich) from bead stocks and restreaked every other day to maintain a fresh stock or from bead stocks, as required. After 18 hours of growth on agar, overnight cultures were prepared in 5ml Lysogeny Broth (LB) (Neogene) at 37°C and incubated for 18 hrs at 180 rpm. M9 medium (1 litre) was prepared by mixing 700 ml M9 salts (15 g/l KH ₂PO ₄, 64 g/l Na ₂HPO ₄, 2.5 g/l NaCl, 5.0 g/l NH ₄Cl) (Sigma-Aldrich) (autoclaved), 2 ml MgSO ₄ (autoclaved) (Sigma-Aldrich) and 20 ml 20% glycerol (Sigma-Aldrich) (filter-sterilized), in order. Sterilized water was then used to make up the volume to 1 L.

Single chemical growth rate assays

Each chemical under consideration for inclusion in the respiratory tract-mimicking media was first assessed by supplementation individually into M9 media at ‘low’, ‘ideal’ or ‘high’ concentrations. Ideal concentrations represent those estimated to most closely reflect the physiological concentration found in each niche, in either health or disease. The relevance of the final list of chemicals included in the final media, in the context of bacterial adaptation to airway environments, is summarised in Table 2. The tested concentration ranges were determined by reference to the literature or from experimental measurement of metabolites ( Table 3). Low and high concentration values were set at 2-fold below and above the ideal concentration of each chemical. Growth rates of PAO1 and LESB65 were assessed in M9 media supplemented with each chemical, using a microplate reader (Varioskan ^® LUX, Thermo Scientific) (Underlying data Figures 1-4). 2 μl of bacteria from overnight cultures were added to 198 μl of M9 media supplemented with the appropriate chemical. The bacteria were then incubated in sterile 96-well plates (U-bottom) (Greiner) for 20 hours and the growth rates assessed at OD _600nm at 10 minutes intervals. Incubation conditions were 37°C/5% CO ₂ for lung conditions and 34°C/0% CO ₂ for sinus conditions. Three biological replicates were performed, per bacterial isolate, per condition. A single working concentration to be used in the final pooled media was then chosen for each chemical, per condition and growth rate assays were performed again with the chosen concentration before the chemicals were combined into a single pooled media. The ‘ideal’ concentration was chosen for the final media unless found to cause complete inhibition of bacterial growth.

Media synonyms:

Healthy sinus media (HSM),

Healthy lung media (HLM),

CF sinus media (CFSM),

CF lung media (CFLM).

Media components were pooled together at concentrations given in Table 3, yielding four different respiratory media: Healthy sinus media (HSM), healthy lung media (HLM), CF sinus media (CFSM) and CF lung media (CFLM). Briefly, eDNA and mucin solutions were prepared separately in distilled water by constant stirring overnight at 4 °C. Next day, the solutions were autoclaved at 121 °C. Other components of the media were prepared in a single solution by firstly adding the M9 media components (water, 20% glucose and 1M MgSO ₄). Stock solutions of all the media components except glucose, succinate, N-acetyl glucosamine (Glc-NAc), bile, NaCl, albumin and amino acids were prepared and added to the media in solution form. Aforementioned chemicals were then added in powder form, slowly, under constant stirring at room temperature. Once all the components were fully mixed, the solution was sterilized by filtration using Vacuubrand ME 2 diaphragm vacuum pump and Millipore Steritop filter units with a pore and neck size of 0.22 μm. Sterile eDNA and mucin was then added to the filtered media and the pH of media was adjusted to 6.8-6.9 using sodium hydroxide (NaOH) (Sigma-Aldrich). The volume of the solution was brought to 1 litre by the addition of autoclaved distilled water. The media was then aliquoted to 50ml falcon tubes and stored at -80°C until further use.

M9 media was prepared without the addition of M9 salts because exogenous salts were added in the form of NaCl and bile salts, as defined in Table 3. Salinity of the media was adjusted according to the salinity of ASM (2%) (measured in this study) for CF media. Salinity of healthy media was kept lower than CF media. Briefly, 1ml of media (ASM, HSM, HLM, CFSM, CFLM) was deposited to the surface of refractometer (V-RESOURCING) after it was calibrated with sterile water. Salinity of CF media was adjusted to 2%, salinity of healthy media was kept less than 2% as CF airways is known to have higher salt content than healthy airways, although salinity in health is not well defined.

Viscosity of media was measured before and after allowing bacteria to form biofilms for three days, to mimic chronic infection. Viscosity measurements (mPa.s) were performed using a rheometer (Anton-Paar) with cone plate (Serial number: 44806). 50-100 1/s range was used for all sterile media and healthy media during infection, and 0.01-1000 1/s range was used for CF media during infection. Changes in media viscosity was observed under each shear rate point.

Stability of developed media was tested over a 30-day period by performing crystal violet stain assays (as described below) using media stored at 4°C for 30 days. This experiment was designed to determine whether media would continue to support biofilm formation after prolonged storage.

Colony forming unit (CFU) assays were performed to observe the growth of PAO1 and LESB65 in the pooled media. Bacterial cultures grown in LB for 18 hours were centrifuged for 12 minutes at 3220rcf at room temperature. After the LB suspension was removed, bacterial pellets were resuspended in 5 ml of sterile phosphate buffer saline (PBS) (Sigma-Aldrich). The suspension was homogenized by vortexing and bacterial optical density was adjusted to OD 0.05-0.06 in PBS using an optical spectrometer (HANNA instruments). 50 ul of this suspension was then added to 4950 ul HSM, HLM, CFSM or CFLM and cultures were incubated under niche-appropriate conditions: sinus media at 34°C with no CO2, lung media at 37°C with 5% CO ₂. Anaerobic jars were used to increase the CO ₂ concentration for lung conditions. CFU was measured at pre-determined time intervals by serial-dilution onto agar. Briefly, 10-fold serial dilutions were prepared in 96 well microplates (Greiner U-bottom) using PBS as the diluent. 20 ul of bacterial dilution was then plated to TSA plates covering the dilutions from 10 ^-2 to 10 ^-9. The plates were air dried and then incubated at 37°C until the colonies were clearly visible (18 hours for PAO1 and 24 hours for LESB65).

Measurement of attached biofilm biomass by crystal violet staining

Starting from an overnight liquid culture in LB (incubated for 18 hours), bacterial cultures were diluted 1:100 in the developed media, or in LB, or M9, two widely used laboratory bacterial culture media. To minimise edge-effect, non-perimeter wells of U-bottomed polystyrene 96-well microtiter plate (Greiner U-bottomed) were inoculated with 180 μl of these dilutions and perimeter wells were used as a negative control (sterile medium). Following 3 days of growth at 37°C, the supernatant (containing non-adhered cells) was removed from each well and plates were rinsed using 200 μl sterile PBS, twice. Subsequently, 200 μl of Crystal Violet (CV) was added to non-perimeter wells and plates were incubated at room temperature for minimum 20 minutes. After 20 minutes, the excess CV was removed by washing the plates under running tap water. Finally, bound CV was released by addition of 200 μl of 95-100% ethanol (Sigma-Aldrich) to the wells with bacteria. After incubating with ethanol for 30 minutes, the absorbance was measured at 600 nm using a BMG plate reader. 95-100% ethanol was used as a blank.

Measurement of cell viability and free-floating biofilm formation by resazurin assays

Overnight cultures of PAO1 and LESB65 were diluted in LB to an OD 600 of 0.05 (±0.01). These cultures were then further diluted in the developed media (1:100) to a total volume of 10ml in glass universal tubes. Free floating biofilm formation in respiratory tract-mimicking media was compared to that observed in LB and M9. Diluted developed media cultures were incubated under conditions appropriate for each respiratory niche, as described above. Cultures were incubated for 3 days while shaking at 75 rpm. Three glass universal tubes containing bacteria-free respiratory media, LB or M9 were used as a negative control. After the incubation, biofilms were disrupted using 250 μl of 100 mg/ml cellulase (diluted in 0.05 M citrate buffer [9.6 g/l Citrate.H ₂O (VWR)] in water and pH to 4.6 with NaOH) and further incubated under aerobic conditions at 37 °C, while shaking at 150 rpm for 1 h. After 30 minutes incubation, biofilms were further disrupted by manual pipetting to ensure complete disruption of biofilms. A portion of disrupted biofilms were then transferred to 96-well plates (Greiner U-bottomed) and to determine the metabolic activity of the bacterial cells released from the disrupted biofilms, 10 μl of 0.02 % (v/v) resazurin (Sigma-Aldrich) (diluted in distilled water) was added to each well of the 96-well plates and incubated for 1-2 hours at niche specific temperature, while shaking at 150 rpm. Following incubation with resazurin, the fluorescence of each well was measured using an excitation wavelength of 540 nm and an emission wavelength of 590 nm in a Fluostar Omega microplate reader and the MARS Data Analysis Software.

The bacteria were grown in 5ml of each developed media or LB until early stationary phase (12 hrs for PAO1 and 20 hrs for LESB65 in the developed media, and 6hrs for PAO1 and 18hrs for LESB65 in LB). To extract RNA, TRI reagent (ZYMO Research) was added and mixtures were incubated for 5–10 min at room temperature. Bacteria were harvested by centrifugation for 30 min at 13,000 rpm and 4 °C. Supernatant was removed and bacterial pellets were used for subsequent RNA isolation or stored at −80 °C. RNA from bacterial cultures was isolated using the Direct Zol RNA Microprep kit (ZYMO Research: R2061) according to the manufacturer’s instructions. DNase digestion treatment was performed using DNAse 1 (ZYMO Research) as per the manufacturer’s protocol. Isolation was performed in an RNase-free environment using RNase-free consumables and reagents. Purified RNA was eluted with RNase-free water (ZYMO Research). Quantification of RNA was performed by measuring the absorbance at 260 nm wavelength using the NanoDrop8000 UV–vis Spectrophotometer (Thermo Scientific). and purity of RNA was analysed by determination of the 260/280 nm ratio. Nuclease free water (Invitrogen) was used as a blank. The 260/280 nm ratio obtained for all samples was between 1.8 and 2.0.

First-strand cDNA synthesis was performed using the iScript cDNA synthesis kit (BIO-RAD: 1708891) according to the manufacturer’s instructions in an RNase-free environment using RNase-free consumables and reagents. For each sample, 2.5 ng RNA was added and incubated in a thermocycler (Applied Biosystem) under the following protocol: 5 min at 25 °C, 30 min at 42 °C and then 5 min at 85 °C. The cDNA was then stored at −20 °C until further use.

cDNA was used for quantitative real-time PCR (qPCR) in duplicate reactions using the GoTaq ^® qPCR Master Mix (Promega: A6001) as per the manufacturer’s instructions. Each reaction contained 2 μl of cDNA (or nuclease free water for the no template control) and 0.2 μM of forward and reverse primers (Eurofins). Primer sequences are provided in Table 4. qPCR was performed using the BioRad CFX Connect Real Time PCR System (BIORAD) using MicroAmp™ Optical 96-Well Reaction Plate (Applied Biosystem) under the following conditions; 2 min at 95 °C followed by 40 cycles of 15 sec at 95 °C and 1 min at 60 °C.

Each qPCR was performed with three biological replicates and in duplicate on each run. Gene rpoD, encoding sigma factor RpoD, was used as a reference gene. cDNA obtained from cultures grown in LB was used as a control. Analysis of relative gene expression in each developed media and SCFM2 versus (vs) LB was performed using the 2 ^−ΔΔCt relative quantification method as described previously. ³⁹

PAO1 was cultured in the developed media for forty days, sub-cultured to fresh media every 2 days and plated on Tryptone Congo red/Coomassie blue Agar (TCCA) prepared by mixing 20 mg/L Coomassie blue (Sigma-Aldrich), 40 mg/L Congo red (Sigma-Aldrich), 10 g/L Tryptone (Sigma-Aldrich) and 12 g/L Bacto agar (Fisher-Scientific) in distilled water and autoclaving at 121°C for 15 minutes. 100 μl of bacterial cultures were spread on to the plates in serial dilutions (10 ^-4, 10 ^-6) every ten days. Plates were incubated at 37°C for 24 hours and for a further 48 hours at room temperature to allow the colonies to uptake the two dyes in the media.

All assays were carried out at least three times independently. Statistical significance was evaluated using One-way ANOVA (*p<0.05, **p<0.01, ***p<0.001, **** p<0.0001). All statistical analysis was done using Graphpad Prism versions 8 and 9 (similar analysis could be performed in Microsoft Excel as a free to use alternative). For single chemical growth rate assays, R studio version 3.6.2 and GrowthCurver (version 0.3.1) was used to plot the area under the curve (AUC_e) values.

Here we describe, step-by step, the procedure used to prepare 1L of each of the four developed media. Reagents and equipment used in this study are listed in Table 16.

Preparation of the base (same for both CF media) 1.

In a 1 litre Duran bottle, add 300 ml distilled water, 20 ml 20% glucose and 2 ml 1 M MgSO ₄.

Preparation of CF sinus media 1.

Add 4.8 g eDNA in 300 ml water under continuous stirring with a magnetic stirrer at room temperature. Add the eDNA very slowly as it can form clumps in the water.

Preparation of CF lung media 1.

Follow steps 1-3 of CFSM protocol for eDNA and mucin stock solution preparation.

Healthy sinuses media

Preparation of the base media (Same for both healthy media) 1.

In a 1 litre Duran bottle, add 250 ml distilled water, 20 ml 20% glucose and 2 ml 1 M MgSO ₄.

Preparation of healthy sinuses media 1.

Add 1 g eDNA in 500 ml water under continuous stirring with a magnetic stirrer at room temperature. Add the eDNA very slowly as it can form clumps in the water.

Healthy lungs media

Preparation of healthy lungs media 1.

Follow steps 1-3 of HSM protocol for eDNA and mucin stock solution preparation.

Incubation conditions for the media

Incubation conditions differ between lung- and sinus media. The incubation conditions for each media are given in Table 13. Conditions of 5% CO ₂ can be achieved by incubating the media in GasPak Jars with candles, using Campygen packs or microaerobic incubators

Notes on handling of media components and media preparation •

Break bovine serum albumin into smaller pieces before adding to media, as it can take time to fully dissolve.

Testing the effect of individual chemicals on bacterial growth

For each chemical under consideration for inclusion in respiratory-mimicking media, the effect on P. aeruginosa growth was determined at three different concentrations. Experiments were performed over 20 hours of growth for PAO1 and LESB65, using M9 media as a base to support minimal bacterial growth (Underlying data Figures 1-4). PAO1 was chosen as a non-CF laboratory reference strain, to act as a comparator to the CF-adapted strain LESB65. Chosen concentrations of all chemicals were higher in CF conditions vs health conditions. Most chemicals were included in both health and CF conditions, with the exception of bile salts and glucose, which were only tested in CF conditions. Airway glucose is significantly elevated in CF related diabetes and bile in the lower airways is associated with CF, as a co-morbidity of gastroesophageal reflux disease. ^{28 , 40}

Results from these experiments showed that, when used individually, most of the tested chemicals enhanced growth of both PAO1 and LESB65 (Underlying data Figures 1-4). The highest concentrations of bile salts, succinate and extracellular DNA (eDNA) used in CF conditions were toxic to bacteria (Underlying data Figures 1-4 (CF lung and CF sinus graphs)). The antimicrobial activity of lysozyme and lactoferrin was also observed to be both niche and concentration dependent, potentially due to decreased antimicrobial activity in lower temperatures, charge interference at high concentrations or induction of bacterial defence mechanism.

PAO1 and LESB65 can be maintained in respiratory-mimicking media for at least three days

Next, the chosen concentration of each chemical was used to prepare pooled media. Growth of PAO1 and LESB65 was determined in the respiratory media, by CFU determination over 72 hours ( Figure 1). Data from these experiments show that viable PAO1 and LESB65 populations were maintained in all media for at least three days, with mid exponential growth of bacteria occurring at ~1×10 ⁸ CFU/ml ( Figure 1), similar to the findings of Palmer et al. for growth of P. aeruginosa in CF sputum. ²¹ PAO1 was able to grow to a higher density in CF sinus and lung media than in either of the equivalent conditions in health. In addition, PAO1 grew faster than LESB65 in all media and reached early stationary phase after 12-14 hours of growth, whilst LESB65 reached stationary phase only after 24 hours of growth in each media. Although LESB65 reached a similar density in all four media, CFU began to decrease in healthy sinus and lung media after 34 hours.

Both attached and free-floating biofilm structures are thought to contribute to success of P. aeruginosa in the CF airways. Mature biofilms are commonly observed as free floating structures in mucus, whereas surface attached biofilms are suggested to dominate the early stages of infection, as physical contact of bacteria and epithelium is necessary as an initial survival strategy. ^{41 , 42} Crystal violet (CV) staining assays showed that there are niche, disease state and strain specific influences on biofilm formation ( Figure 2). PAO1 formed attached biofilms more readily than LESB65 in all airway niches and conditions. This may be due to the differences in growth rate of PAO1 and LESB65. In particular, PAO1 was able to form significantly more attached biofilm in sinus media compared to lung media. No statistically significant differences in biofilm formation were observed with LESB65 between healthy airway niches, but biofilm formed more readily in CF sinus media than in CF lung media. These results further highlight the importance of studying different airway niches and conditions in isolation. The media were compared with two widely-used laboratory media (LB and M9), and attached biofilm formation was shown to form most readily in LB.

Free-floating biofilm formation, associated with chronic infection, was assessed in respiratory-mimicking media, LB and M9, Figure 3. Free-floating biofilms formed more readily in LB and M9 than in any of the four respiratory media, for both isolates, suggesting that the respiratory media alter bacterial growth and biofilm formation, compared to standard laboratory media. Biofilms grew well in CF sinus and lung media, and in healthy lung media, but only minimal biofilm was observed in healthy sinus media. PAO1 outperformed LESB65 in the attached biofilm formation assays but not in the free-floating biofilm formation assays. As an airway adapted strain from chronic infection, LESB65 may be more adept at forming free-floating biofilms.

Viscosity measurements were taken in the four developed media, pre and post biofilm formation ( Figure 4). Sterile HSM and HLM media were shown to have Newtonian liquid characteristics with very similar viscosity measurements to water (~1.0 mPa·s viscosity for water, 1.02 mPa·s for HSM and 1.03 mPa·s for HLM) ( Figure 4A). Viscosity changed only minimally in healthy media after bacterial growth (1.06 mPa·s for HSM and 1.05 mPa·s for HLM) ( Figure 4A). Both CF media were also determined to be Newtonian fluids pre-biofilm formation but had higher viscosity measurements compared to healthy media (1.37 mPa·s for CFSM and 1.27 mPa·s for CFLM) ( Figure 4A). When the biofilm was established in CF media, both media became more viscous, as compared to both their sterile controls and compared to post-biofilm formation healthy media. Infected CF media showed clear non-Newtonian fluid characteristics, with shear thinning properties (viscosity decreases as the shear rate increases) ( Figure 4B). Viscosity of CFSM media was approximately 16 mPa·s during infection at low shear rate ranges and decreased down to 1.57 mPa·s at the highest shear rate (1000 1/s). CFLM media was the most viscous media during infection, with approximately 2400 mPa·s viscosity at low shear rates, decreasing down to 1.8 mPa·s at 1000 1/s shear rate. These results are in agreement with previous literature, where the viscosity of CF sputum was shown to be higher than healthy controls. ⁴³

Biofilm formation of PAO1 was measured over a month in the developed media stored at 4°C to test the stability of the airway mimicking media. Over a 30-day period, there was no significant change in the amount of biofilm formed by PAO1 in any of the four airway mimicking media ( Figure 5).

Four P. aeruginosa genes were selected, to determine whether growth in respiratory-mimicking media induced CF infection-relevant gene expression. The major virulence regulator, extracytoplasmic sigma factor algU, and mexB (part of a tripartite multidrug efflux system) genes were selected as they are virulence genes of P. aeruginosa known to contribute to chronic infection. ^{44 , 45} PA2911 and PA2382 genes were also selected, as expression of these genes in clinical isolates in CF sputum is observed but expression has not previously been captured in in vitro CF models. ⁴⁶ algU is responsible for alginate overproduction, causing mucoidy and contributing to chronic infection in PwCF. ⁴⁴ MexB is responsible for exporting biological metabolites and antimicrobial compounds, playing a crucial role in antimicrobial resistance in the airways. ⁴⁵ PA2911 is predicted to be a Ton-B dependent receptor located in the outer membrane and responsible for sideophore transport. ⁴⁷ PA2382 ( lldA) is a L-lactate dehydrogenase, important for pyruvate metabolism. ⁴⁸ Results from qPCR studies showed that expression of all four genes were increased in all the airway-mimicking media when compared to expression in LB. Expression of the four genes was also determined in P. aeruginosa (PAO1 and LESB65) grown in SCFM2. Expression of algU and mexB was comparable in airway media and SCFM2, but expression of PA2911 and PA2382 was higher in respiratory-mimicking media than in SCFM2, although did not reach statistical significance, Figure 6.

PAO1 growing in respiratory-mimicking media show changes in colony morphology

PAO1 was grown for a period of 40 days in each of the four media. Every ten days, PAO1 populations cultured under different media conditions were streaked onto TCCA plates to observe colony morphotypes ( Figure 7). Only shiny, circular, concave colonies were observed in the starting populations. Wrinkly colonies began to appear in all media conditions over time. These wrinkly structures have been suggested to be the result of redox-driven adaptation that maximizes oxygen accessibility in biofilms, to increase their surface area in oxidant limitation. ¹⁷ Colony morphotypes in Figure 7D and E were visible only after transfer 10 and were unique to CF media ( Table 14).

Figshare: Underlying data for “Development of liquid culture media mimicking the conditions of sinuses and lungs in cystic fibrosis and health”, https://doi.org/10.6084/m9.figshare.20463159.v6. ⁸¹

This project contains the following underlying data: -

Biofilm assay OD values and fluorescent intensity.xlsx

Data are available under the terms of the Creative Commons Attribution 4.0 International license (CC-BY 4.0).