Cystic Fibrosis Aggregate Biofilm Model to Study Infection-relevant Gene Expression

Summary

This study outlines a framework for developing an aggregate polymicrobial biofilm model and optimizing an RNA extraction protocol to assess Pseudomonas aeruginosa gene expression reflective of the cystic fibrosis lung environment. Applications include evaluating antimicrobial effects and studying antibiotic alternatives under conditions relevant to cystic fibrosis.

Abstract

Standard pre-clinical testing methods for novel antimicrobial therapeutics used to treat chronic lung infections in people with cystic fibrosis do not reflect the environmental conditions of the hostile lung niche. Current reductionist testing conditions can lead to the progression of compounds along a preclinical pipeline without evidence of their activity under cystic fibrosis lung niche-appropriate conditions. Several approaches used to study traditional antimicrobials may not be suitable for antibiotic alternatives, including anti-virulence therapeutics like anti-quorum sensing agents and siderophore inhibitors. This protocol documents an aggregate biofilm model of Pseudomonas aeruginosa to compare resistance and infection-relevant gene expression in single-species and multi-species cultures (Staphylococcus aureus and Candida albicans), examining colony-forming unit (CFU) reductions and changes in gene expression, using algD as an exemplar. The model was optimized for small, static volumes of bacterial cultures to allow the study of novel compounds in the discovery phase of the drug development pipeline, where compound quantities may be limited. Single-species P. aeruginosa biofilms were formed in Synthetic Cystic Fibrosis Medium 2 (SCFM2) for 24 h before treatment with meropenem at different concentrations (1, 16, and 256 µg/mL) for a further 24 h. Polymicrobial biofilms were established by growing Staphylococcus aureus and Candida albicans together in SCFM2, then inoculating with P. aeruginosa for an additional 24 h and treating with meropenem. The lack of a direct connection between compound efficacy measures in pre-clinical testing and clinical trial results has cast doubt on the applicability of current laboratory screening tools. This model allows us to understand the impact of relevant factors on P. aeruginosa gene expression, including genes contributing to resistance and virulence, thereby bridging this gap.

Introduction

Cystic fibrosis (CF) is the most common autosomal recessive disorder within Caucasian populations¹. Mutations in the Cystic Fibrosis Transmembrane Conductance Regulator (CFTR) gene result in an array of clinical symptoms, including the overproduction of mucous within the respiratory tract. Impaired ion transport results in this mucous becoming dehydrated, as a consequence of which, it is not adequately cleared by the mucociliary system, increasing susceptibility to chronic pulmonary infections¹. The microbiology of the CF lung consists of a distinct group of opportunistic pathogens, with Haemophilus influenzae and Staphylococcus aureus being considered primary organisms infecting the lungs of infants and children with CF, followed by a decline of these organisms and subsequent colonization with Burkholderia spp., or Pseudomonas aeruginosa during adulthood^1,2,3,4. In addition to bacterial species, fungal microorganisms are also common in the CF lung environment, with Candida spp and Aspergillus spp being amongst the most commonly isolated fungal species^3,4,5. Although these organisms play important roles in CF lung infections, advancements in culturing techniques and culture-independent methods have shown that CF lung infections are rarely limited to a single bacterial species, thus highlighting the complex, polymicrobial nature of the CF lung^1,6.

Despite this knowledge, current in vitro and in vivo methods for the development and testing of antibiotics for use in chronic lung infections lack complete resemblance to the CF lung niche⁷, with focus being placed on determining the minimum inhibitory concentration (MIC) of the drug against single-species cultures. This provides limited consideration of the infectious microenvironment, which is a key component of CF lung pathology⁸. As such, methods typically involve the growth of bacteria as planktonic cultures, a growth state that likely does not reflect that of bacteria within the CF lung. This is particularly important for P. aeruginosa, which typically presents as biofilm aggregates that often additionally contain other bacterial and fungal species⁹. Such factors present challenges when testing antibiotics for chronic lung infections and highlight a disparity between antibiotic susceptibility testing methods and antimicrobial activity within the CF lung environment^9,10. These challenges are exacerbated when screening for antibiotic alternatives, such as anti-virulence therapeutics, where MIC assays are not an indicator of activity and do not provide an understanding of host-pathogen interactions¹¹.

Alginate, an extracellular polysaccharide, is a key component of P. aeruginosa biofilms within the lung niche and increases bacterial tolerance to antibiotics and immune agents, as well as enhancing surface adhesion¹². Alginate is an attractive target for anti-virulence therapy development, and several therapeutics targeting alginate biosynthesis have been developed, such as those targeting algD, a key gene in alginate biosynthesis^13,14. Such therapeutics could be used in conjunction with antibiotic therapies in people with CF to disrupt P. aeruginosa biofilms, converting bacteria into a planktonic state. This increases antibiotic efficacy, as disruption of the biofilm extracellular matrix allows for an increased ability of the antibiotic to reach the bacterial target¹⁵. However, despite evidence of in vitro activity, the action of these therapeutics within the CF lung niche remains unclear.

The aggregate biofilm model presented here acts to address these issues through the addition of S. aureus and C. albicans, which are common co-colonizers with P. aeruginosa within the CF lung. It also allows P. aeruginosa to grow in biofilm aggregates, as seen in the respiratory environment. The use of the sputum mimicking media Synthetic Cystic Fibrosis Media 2 (SCFM2) as the nutritional environment for growth allows for the addition of host influences, including mucin, extracellular (e)DNA, and an amino acid profile that replicates that seen within the CF lung¹⁶.

To further capture the conditions of the CF lung, biofilms are exposed to meropenem, an antibiotic that is frequently prescribed to people with CF during periods of pulmonary exacerbation. The use of an antibiotic within this model is recommended as a means of replicating the clinical conditions in which an anti-virulence agent may be prescribed, e.g., as a combination therapy coupled with antibiotic treatment to manage respiratory symptoms during exacerbation. Furthermore, the inclusion of meropenem allows assessment of the P. aeruginosa transcriptional response to antibiotic exposure under infection-relevant conditions. This model could be used to provide information on which factors would be appropriate drug targets within the CF lung. The model could also be amended to include other relevant antibiotics, such as tobramycin, and to assess the expression of other genes of interest.

Here, this aggregate polymicrobial model is utilized to assess the effects of the presence of Staphylococcus aureus and Candida albicans and varying concentrations of meropenem (1, 16, and 256 µg/mL) on the expression of algD by P. aeruginosa. Meropenem is prescribed to treat primarily P. aeruginosa lung infections and has no direct effect on S. aureus or C. albicans, where the MIC for both of these co-habiting species is >256 µg/mL.

This work will increase understanding of how relevant factors, such as host factors within SCFM2 and co-colonizing microorganisms, can influence the expression of P. aeruginosa infection-relevant genes and will provide the foundations for the development and testing of anti-virulence therapeutics for use in chronic CF respiratory infections.

Protocol

1. Preparation of single-species biofilms

1. Revive P. aeruginosa PAO1 from frozen bead stock cultures by streaking a single bead onto Lysogeny Broth (LB) agar and incubating for 24 h at 37 °C.
2. Prepare overnight cultures of PAO1 by taking a single colony from the streak plate and adding it to 5 mL LB medium. Incubate overnight (18 h) then dilute to OD₆₀₀ of 0.05 (equivalent to 1 x 10⁸ CFU/mL), and further dilute 1:100 in SCFM2 prepared according to Palmer, et al.¹⁶.
3. Add 180 µL of the inoculum to wells of a round-bottomed 96-well microtiter plate. Prepare biofilms in biological triplicate for enumeration by colony-forming units (CFUs). A total of 3 replicates per biological replicate are performed for qPCR only. To 3 separate wells, add 180 µL of sterile SCFM2 to be used for the media controls.
   NOTE: Round-bottom plates were selected as opposed to traditional flat-bottom plates due to their superiority in allowing homogeneous exposure of the entire biofilm to the surrounding growth media¹⁷.
4. Allow the biofilms to form over 24 h, incubating at 37 °C and shaking at 75 rpm.

2. Preparation of multi-species biofilms

1. Revive P. aeruginosa PAO1 and S. aureus SH1000 from frozen bead stock cultures as described above. Revive C. albicans CAF2.1 from frozen bead stock cultures by streaking a single bead onto Saboraud dextrose agar and incubating for 24 h at 37 °C.
2. Prepare overnight cultures of SH1000 and CAF2.1 by taking a single colony from the streak plate and adding 5 mL of LB (separate overnight cultures for each organism). Incubate overnight for 18 h. Standardize overnight cultures of S. aureus SH1000 and C. albicans CAF2.1 in LB to OD₆₀₀ values of 0.1, which equates to 1 x 10⁹ and 1.2 x 10⁷ CFU/mL, respectively^17,18.
3. Dilute the S. aureus and C. albicans to form a single inoculum containing both species at concentrations of 1 x 10⁶ and 1 x 10⁵ CFU/mL, respectively, in SCFM2.
4. Add 162 µL of the inoculum to wells of a 96-well microtiter plate. Inoculate biofilms in biological triplicate for enumeration by colony-forming units (CFUs). A total of 3 replicates per biological replicate are performed for qPCR only. To 3 separate wells, add 180 µL sterile SCFM2 to be used for the media controls.
5. Allow the biofilms to form for 24 h, incubating at 37 °C and shaking at 75 rpm.
6. Prepare an overnight culture of PAO1 by taking a single colony from the streak plate and adding it to 5 mL of LB medium. Incubate overnight (18 h), then add at OD₆₀₀ of 0.05 (equivalent to 1 x 10⁸ CFU/mL), and further dilute to 1 x 10⁵ CFU/mL in SCFM2.
7. Add 14.2 µL of the PAO1 inoculum to wells of the 96-well microtiter plate, resulting in the final inoculum of the PAO1 being 1 x 10⁴ CFU/mL.
   NOTE: This variation in the concentration of the inoculate between the different species is to ensure that the S. aureus and C. albicans are not outcompeted by the P. aeruginosa and all are recoverable by spot plating from the 24 h biofilm where no antibiotics are present.
8. Allow the polymicrobial biofilms to form over 24 h, incubating at 37 °C and shaking at 75 rpm.

3. DNase and antibiotic treatment of single-species and multi-species biofilms

1. Add 2 µL of DNase I directly to the wells containing the biofilms at a final concentration of 50 µg/mL, incubate for 1 h at 37 °C, and shake at 75 rpm.
   NOTE: This is added to thin the viscous biofilms to allow better penetration of antimicrobials, reflecting the dosage of rhDNase (recombinant human DNase) that is taken by many people with cystic fibrosis as a management strategy for pulmonary exacerbations.
2. Prepare an antibiotic stock of meropenem to 2.56 mg/mL, using distilled water as the diluent. Serially dilute by a factor of 2 to achieve concentrations 0.01 mg/mL - 2.56 mg/mL.
3. Add 20 µL of each meropenem dilution to the biofilms to achieve a dosage range of 1 µg/mL - 256 µg/mL. Add 20 µL of distilled water to negative control biofilms in place of the antibiotic.
4. Seal the 96-well microtiter plate with transparent film and incubate at 37 °C for 24 h, shaking at 75 rpm.

4. Enumeration of viable cells and metabolic activity

NOTE: The following steps are for the biofilms destined for CFU quantification only.

1. Add 10 µL of cellulase (100 mg/mL, 0.05 M citrate buffer) to biofilms and incubate at 37 °C for 1 h, shaking at 140 rpm. Cellulase is regularly used to disrupt the cellulose skeletons within biofilms^19,20,21. Further disrupt the biofilms by pipetting up and down if required until no visible clumps are present.
2. Add 10 µL of 0.02% resazurin solution to assess the metabolic activity of the test biofilms and control biofilms and incubate at 37 °C for 1 h, shaking at 140 rpm.
3. Measure the fluorescence of each well using an excitation wavelength of 540 nm and an emission wavelength of 590 nm in the microplate reader.
4. Correct for background fluorescence by subtracting the fluorescence values from the blank media samples and calculate percentage reduction as (mean of F_{antibiotic treated biofilms}/mean F_{negative control}) x 100 %.
5. Enumerate the contents of the wells, containing both the disrupted biofilms and the planktonic cells, by serially diluting and spot plating onto selective agar as follows:
   Pseudomonas selective agar supplemented with 1% glycerol and CN supplement for enumeration of P. aeruginosa
   Mannitol salt agar for the enumeration of S. aureus|
   Yeast Peptone Dextrose agar supplemented with penicillin (250 U/mL), streptomycin (250 µg/mL, gentamycin sulfate (30 µg/mL), and vancomycin (3 µg/mL) for the enumeration of C. albicans.
   NOTE: The entire contents of the wells are enumerated due to the challenges of separating the free-floating aggregate biofilm from the planktonic cells. However, this approach more accurately reflects the conditions present in the cystic fibrosis lung. Condensation may be present on the lid after incubation. To ensure no contamination between wells, spot plate the sterile controls onto selective agar.
6. After spot plating onto the selective agar, incubate the plates for 24 h at 37 °C prior to enumeration.

5. Preparation of biofilms for RNA extraction

NOTE: All steps from this point forward are for biofilms destined for RT-qPCR analysis only.

1. Remove the 96-well microtiter plate containing the biofilms from the 37 °C incubator and transfer them to a 1 mL microcentrifuge tube. When transferring, ensure that the biofilm replicates are pooled for each biological repeat performed. This ensures that there is enough biomass for RNA extraction.
2. Centrifuge the 1.5 mL tubes containing the biofilms at 16,000 x g for 5 min. If performing the RNA extraction on the same day, proceed to step 6.2.1. If RNA extraction is not performed on the same day, remove the supernatant and store the pellet in 250 µL of RNA stabilization solution at -20 °C. This solution has been used to store bacterial pellets in several studies^22,23.

6. RNA extraction

1. Extract bacterial RNA single-species and multi-species biofilms using a commercial kit. Prepare the buffers as instructed in the manufacturer's guidelines.
2. Remove the culture media supernatant and discard it. If the pellet was stored in a stabilization solution, thaw it to room temperature and centrifuge at 16,000 x g for 5 min. Resuspend the pellet in 600 µL of RNA extraction reagent, then manually disrupt it using a 0.2 mm needle attached to a 2 mL syringe (aspirate 5x-10x or until the pellet is completely disrupted).
3. Perform the RNA purification steps as instructed in the manufacturer's guidelines.
4. Quantify the RNA using a UV spectroscope, an RNA BR Assay Kit and tubes. The detectable limits of RNA using the Assay kit is 10 - 1200 ng/µL, therefore the minimum acceptable yield of RNA for this protocol is 10 ng/µL.
   1. Set up the required number of tubes for samples and standards (2 standards are required).
   2. Prepare the working solution by adding 199 µL of buffer and 1 µL of reagent per number of samples (+ 2 standards and 2 for error).
   3. Add 190 µL of working solution and 10 µL of Standard 1 and Standard 2 into separate tubes.
   4. Add 199 µL of working solution and 1 µL of RNA sample into separate tubes. Vortex for 30 s and then store in a dark location for 5 min.
   5. On the Fluorometer, select RNA > Broad range RNA and follow the instructions.
   6. Assess the quality of the RNA using a spectrophotometer. Blank the spectrophotometer by pipetting 1 µL of DNase/RNase- Free Water onto the spectrophotometer. Wipe away the water using a lint-free wipe and set the ratio of absorbance to 260/280 nM. Pipette 1 µL of RNA onto the spectrophotometer. If the RNA is pure, a single peak will be observed between 260 and 280 nM, and a ratio of ~ 2.0 will be shown.
5. Convert extracted RNA to complementary DNA (cDNA) using a cDNA Synthesis Kit. For any RNA samples exceeding 50 ng/µL, first dilute to 50 ng/µL using DNase/RNase-Free Water.
   1. Using tubes, prepare the cDNA reaction mix according to Table 1.
   2. Place the tubes in a Thermal Cycler using the following conditions: 25 °C for 5 min, 42 °C for 30 min, 85 °C for 5 min. Either use the cDNA immediately or store it frozen at -20 °C.

7. qPCR

1. Perform the qPCR using a qPCR Master Mix. Use the following algD and housekeeping gene (16S) primers (Table 2).
   NOTE: The primers were designed using the NCBI-primer Blast online tool with reference to the whole genome sequence of P. aeruginosa, and then synthesized through a commercial vendor. The primers were validated for specificity in the polymicrobial biofilms through sequence matching using the NCBI-primer Blast tool.
2. Dilute the primers to 4 µM (4 pm/µL) by diluting 20 µL of primer in 480 µL of DNase/RNase-Free Water. These can be frozen for multiple uses.
3. Prepare the working solution of the qPCR master mix according to the instructions in Table 3.
   NOTE: The algD working solution contains enough solution for 3x biological repeats and 3x technical repeats per meropenem concentration, 6x negative controls, and 2x error. The 16S working solutions contain enough solution for 2x standards of known RNA concentration and 2x error.
4. Add 18 µL of algD working solution to the appropriate qPCR reaction tubes. This will include a total of 36 tubes for algD expression analysis and 6 tubes for negative controls.
5. Add 2 µL of RNA to each of the qPCR reaction tubes or 2 µL of DNase/RNase-Free Water to the negative controls.
6. Add 18 µL of 16S working solution to the appropriate qPCR reaction tubes. This will include a total of 2 tubes. Add 2 µL of RNA of known concentration to the qPCR reaction tubes.
7. Perform the qPCR according to the following conditions: 2 min at 95 °C, 15 s at 95 °C (40 cycles), 1 min at 60 °C.
8. Analyze the qPCR results using the 2-ΔΔCt method, as described earlier²⁴. When analyzing algD expression, use algD expression at 0 µg/mL as the control for the respective biofilm compositions and consider fold change expression relevant to untreated cultures

Results

The results shown in Figure 1A highlight the resilience of the single-species Pseudomonas biofilms when subjected to treatment with meropenem, as no significant decrease was observed at any dosages tested. The negative control, which was not exposed to meropenem, showed a recovery of 8.31 ± 0.07 Log₁₀CFU/mL, while the largest reduction of viable cells occurred at 16 µg/mL, resulting in a recovery of 8.06 ± 0.10 Log₁₀CFU/mL. However, this level of resistance is expected, as it is shown in the literature that P. aeruginosa biofilms are much more resistant to meropenem than their planktonic counterparts, with the concentration requiring >256 µg/mL to be bactericidal against the biofilms^25,26, compared to the planktonic cells, which have an MIC of 2 µg/mL.

Figure 1B shows the average percentage reduction in metabolic activity of the mono-culture P. aeruginosa biofilms of each antibiotic concentration compared to the control biofilm, which was not exposed to meropenem. A reduction in fluorescence translates to less metabolic activity taking place, taken as a proxy for a reduction in viable cells. Meropenem-treated cultures showed increased metabolic activity (denoted by a negative percentage reduction), indicating no (or very limited) bactericidal activity.

The increased metabolic activity without increased CFU recovery may indicate the metabolic response of the bacteria to antibiotic exposure. Between the single-species and multi-species biofilms (Figure 2), the overall growth of P. aeruginosa differed only slightly, with the recovery of viable cells from the single-species biofilm being 0.74 ± 0.17 Log₁₀CFU/mL higher than that seen in the polymicrobial biofilm when no antibiotic was present. S. aureus was recovered from the polymicrobial biofilm at 6.95 ± 0.11 Log₁₀CFU/mL, and C. albicans was recovered at 1.66 ± 2.88 Log₁₀CFU/mL. C. albicans, however, is present only where no meropenem has been added and was not recovered from any biofilms where the antibiotic had been administered. As meropenem has no direct effect on Candida, which has been shown in previous MIC assays conducted in the laboratory on C. albicans monoculture where the MIC value of meropenem was >256 µg/mL, it is probable that it is the bacterial stress response of the other cohabiting species, S. aureus and P. aeruginosa, to the antibiotic exposure that results in hostile conditions for the fungus.

As P. aeruginosa biofilms, in both monoculture and with co-colonizing pathogens, show no reduction in viable cell recovery in meropenem concentrations up to 256 µg/mL, this highlights the need for alternative therapeutics, such as anti-virulence therapies, to combat such chronic infections in the clinic, which may work in conjunction with antimicrobials to effectively reduce the bacterial load.

algD expression in P. aeruginosa PAO1 single-species biofilms
The expression of algD in P. aeruginosa PAO1 exposed to different concentrations of meropenem (1, 16, and 256 µg/mL) grown in SCFM2 was assessed using SYBR-Green RT-qPCR. The expression was then analyzed using the 2-ΔΔCt method, with 16S used as the control gene (Figure 3). The expression of algD did not significantly differ between the concentrations of meropenem used, suggesting that when grown as a single-species biofilm, PAO1 algD expression is not heavily affected by the presence of meropenem.

algD expression in P. aeruginosa PAO1 multi-species biofilms
The expression of algD was also assessed in P. aeruginosa PAO1 grown in biofilms consisting of S. aureus SH100 and C. albicans CAF2.1 to assess if the presence of co-colonizing organisms influenced the expression of algD or the response to meropenem in PAO1. No significant differences in algD expression were observed between 0-16 µg/mL meropenem exposure. However, at 256 µg/mL, the expression of algD was significantly higher compared with the other antibiotic concentrations (0 µg/mL versus 256 µg/mLp = 0.0075, 1 µg/mL versus 256 µg/mL p = 0.035, 16 µg/mL versus 256 µg/mL p = 0.0048) at approximately 7-fold higher than the control condition (Figure 4). This highlights that, when grown in the presence of S. aureus and C. albicans, there is an increased expression of algD at the highest concentration of meropenem, in comparison to when P. aeruginosa is grown alone, suggesting an increase in alginate production in the polymicrobial setting. Despite demonstrating differences in algD expression between the single and multi-species biofilms, it is unclear if this increase in expression is driven by the presence of a specific species (S. aureus or C. albicans) or if interactions involving both microorganisms are driving this gene expression change.

These results highlight the importance of understanding the impact of therapeutics and environment on gene expression, particularly in the context of future testing of virulence-targeting therapeutics.

Figure 1: Single-species biofilms consisting of Pseudomonas aeruginosa PAO1. The biofilms were cultured for 24 h in SCFM2, treated with 50 µg/mL DNase I for 1 h, and treated with meropenem at various concentrations for a further 24 h. (A) Average Log₁₀CFU/mL. The dotted line denotes a 2-log reduction (equivalent to 99 % reduction) in P. aeruginosa recovery from the negative control. (B) Percentage reduction in relative fluorescence units. The dotted line denotes a 50% reduction in fluorescence, the first point above which is considered to be the MIC50 value. Please click here to view a larger version of this figure.

Figure 2: Multi-species biofilms consisting of Pseudomonas aeruginosa PAO1, Staphylococcus aureus SH1000, and Candida albicans CAF2.1. The biofilms were cultured for 24 h in SCFM2, treated with 50 µg/mL DNase I for 1 h, and treated with meropenem at various concentrations for a further 24 h. The dotted line denotes a 2-log reduction (equivalent to 99 % reduction) in P. aeruginosa recovery from the negative control. Significance was determined by two-way ANOVA where **** p < 0.0001. Please click here to view a larger version of this figure.

Figure 3: Expression of algD extracted from single species consisting of P. aeruginosa PAO1. The biofilms were cultured for 24 h in SCFM2, treated with 50 µg/mL DNase I for 1 h, and exposed to varying concentrations of meropenem (0, 1, 16, 256 µg/mL). Gene expression was analyzed using the 2-ΔΔCt method. The housekeeping gene 16S was used in the analysis. Significance determined by Tukey's multiple comparison test. Please click here to view a larger version of this figure.

Figure 4: Expression of algD extracted from multi-species biofilms consisting of P. aeruginosa PAO1, S. aureus SH1000, and C. albicans CAF2.1. The biofilms were cultured in SCFM2, treated with 50 µg/mL DNase I for 1 h, and treated with meropenem at various concentrations (0, 1, 16, 256 µg/mL) for a further 24 h. Gene expression was analyzed using the 2-ΔΔCt method. The housekeeping gene 16S was used in the analysis. **p < 0.01 as determined by Tukey's multiple comparison test. Please click here to view a larger version of this figure.

	Volume per cDNA reaction (μL)
5X iscript Reaction mix	4
iscript Reverse Transcriptase	1
DNase/RNase-Free Water	14
2.5ng RNA	1
Total reaction volume	20

Table 1: Working solutions of cDNA master mix. Components and volumes required per cDNA reaction.

Gene	Forward primer	Reverse primer
algD	CGCCGAGATGATCAAGTACA	TGTAGTAGCGCGACAGGTTG
16S	ACCGCATACGTCCTACGG	CGAAGACCTTCTTCACACACG

Table 2: algD and 16S forward and reverse qPCR primers.

	Volume per reaction (µL)	algD working volume (µL)	16S working volume (µL)
GoTaq qPCR Master Mix (2X)	10	440	40
Forward primer	1	44	4
Reverse primer	1	44	4
DNase/RNase-Free Water	6	264	24
Total	18	792	72

Table 3: Working solutions of qPCR master mix. Components and volumes required for the algD and 16S qPCR master mix solutions.

Supplementary Figure 1: Multi-species biofilms consisting of Pseudomonas aeruginosa LESB58, Staphylococcus aureus SH1000, and Candida albicans CAF2.1. The biofilms were cultured for 24 h (n = 3) in SCFM2, treated with 50 µg/mL DNase I for 1 h, and treated with meropenem at various concentrations for a further 24 h. The dotted line denotes a 2-log reduction (equivalent to 99% reduction) in P. aeruginsoa recovery from the negative control. Statistical analysis was carried out by two-way ANOVA. *p=0.0242, ** p=0.0013, **** p <0.0001. Please click here to download this figure.

Discussion

These results show that despite its frequent use in managing pulmonary exacerbations in people with CF, meropenem did not significantly influence the growth of P. aeruginosa in the single-species or multi-species biofilm model. A slight increase in P. aeruginosa recovery was observed in both models following treatment with 128 µg/mL and 256 µg/mL meropenem, as also seen for S. aureus in the multi-species biofilm; however, these increases were insignificant. The decrease in C. albicans when subjected to meropenem may be due to hostile conditions for the fungus being created by the bacteria in response to the stress of the antimicrobial. No significant differences were observed in algD expression between meropenem concentrations when P. aeruginosa was grown as a single-species biofilm. However, algD expression was significantly higher in multi-species biofilms when treated with 256 µg/mL meropenem, suggesting alginate production may be increased under these conditions. The differences in algD expression observed between biofilm compositions and meropenem concentrations show how the presence of relevant factors can influence infection-relevant gene expression and highlights the importance of considering antimicrobial usage and polymicrobial infection status when developing and testing antibiotic alternatives.

Despite the usefulness of the models, there are several considerations to ensure successful application. The establishment of S. aureus and C. albicans within the biofilms is essential before infection with P. aeruginosa, due to the risks of S. aureus and C. albicans being outcompeted by the Gram negative. In this biofilm model, C. albicans appears to be a poor partner due to its significantly lower survival rates compared to S. aureus and P. aeruginosa, even in the absence of antibiotics. This suggests that the yeast is either outcompeted for resources or is negatively impacted by interspecies interactions within the biofilm^27,28. The absence of C. albicans in the presence of meropenem, despite the antibiotic having no direct antifungal activity, supports the idea that the stress response of S. aureus to meropenem may create an environment that is inhospitable to C. albicans^29,30. However, it is important to note that strain choice can have an effect on the amount of C. albicans that can be recovered from the polymicrobial biofilm model. Different trends are observed when using P. aeruginosa LESB58 instead of PAO1 (Supplementary Figure 1). With LESB58, C. albicans is recoverable at all concentrations of meropenem and actually shows a slight increase as the P. aeruginosa concentration decreases as a result of the antimicrobial, allowing the community composition to change. However, despite these differences, PAO1 is used in this model due to being a laboratory reference strain, which provides a consistent and well-characterized baseline for comparative studies. This consistency is crucial for reproducibility and for drawing reliable conclusions about the interactions within the biofilm.

It can be challenging to extract high quantities of RNA from bacteria grown in SCFM2, due to the viscous nature of the media. Therefore, the disruption of the biofilm with 0.2 mm needles prior to RNA extraction is an important step to increase RNA yield and ensure sufficient concentrations (≥10 ng/µL) are extracted for RT-qPCR. RNA is highly susceptible to degradation³¹, therefore following extraction, ensure all downstream activities that use RNA are performed on ice. Regular freeze-thawing of RNA, cDNA, and qPCR primers can increase the likelihood of product degradation. Therefore, it should be limited as much as possible. Another common challenge in the application of these models is the risk of contamination during the RNA extraction or qPCR assay. This risk can be minimized by spraying equipment with DNA degradation solution prior to RNA extraction and by using DNase/RNase-Free pipette tips throughout the procedure. The regular replacement of DNase/RNase-Free water will also help to reduce the risk of contamination within the qPCR assay.

This protocol chose to focus on algD as the gene of interest due to its necessity in the production of alginate, which is described as a key virulence factor of P. aeruginosa during both acute and chronic infections³². Thus, alginate is increasingly being studied as a target for anti-virulence therapeutics^33,34. Although this model has been used to assess the expression of algD, it can be applied to assess the expression of other P. aeruginosa genes for different anti-virulence drug targets. Meropenem was utilized in this study due to evidence suggesting the antibiotic influences the expression of alginate biosynthesis genes, particularly when P. aeruginosa is grown in the presence of C. albicans³⁵. Furthermore, meropenem is often clinically prescribed to treat and manage P. aeruginosa respiratory infections, highlighting its relevant employment within this model³⁶. However, the addition of different antibiotics, or even dual treatments, is also possible with these models. The incorporation of antibiotics within this model would also allow for this model to be used to study evolutionary changes in bacterial resistance mechanisms, such as the development of porin-mediated resistance by P. aeruginosa to carbapenems³⁷. This polymicrobial biofilm model could be adapted to include other relevant species, for example commensal bacteria such as Rothia sp. and Streptococcus sp. or substituted with other infection-causing microbes such as Aspergillus sp. to reflect the requirements of the drug candidate being examined.

Polymicrobial biofilm models that mimic the CF lung have been developed previously³⁸. The model described here builds on previous work with a view to increasing throughput by using a setup amenable to small-volume microbial cultures. This is advantageous for testing novel compounds where the available material may be limited. As the model is small-scale and static, it is possible to grow the biofilms in different oxygen availabilities to more closely reflect the microaerophilic environment in the lungs^39,40. For ease of reproducibility and for reasons of cost-effectiveness, the biofilms reported in this study were grown aerobically.

Despite the valuable insights gained from this study, several limitations must be acknowledged. Firstly, the absence of host cells limits its utility for the study of virulence factors that have a direct mammalian cellular target. This outlines the challenges of establishing in vitro models suitable for the study of virulence factors and highlights the need for further development of models that replicate the hostile environment of chronic infections. Furthermore, although regarded as a virulence factor, alginate is a key feature of biofilms and is produced by P. aeruginosa in both the laboratory and the environment. Using in vitro models to study antibiofilm agents may be less challenging than using the models to study other anti-virulence therapies, such as anti-exotoxin agents, where the target has direct action on eukaryotic cells⁴¹.

While the choice of P. aeruginosa strain of PAO1 is suitable for reproducibility and standardization of the in vitro model, it may be desirable to include isolates from chronic infection. Future experimentation using this model with a strain isolated from patient samples is to be carried out to observe the effect of antimicrobials on strains that are adapted to the respiratory niche.

The use of SCFM2 as a sputum mimic in this in vitro model allows for standardized and replicable observation of the behavioral changes of aggregate biofilms cultured in an airway-relevant environment during the course of antibiotic treatment. SCFM2 was selected over the original version of SCFM due to the inclusion of extracellular DNA (eDNA) and mucin, which are important factors in supporting the formation of biofilms. However, it is important to note that other variants of the synthetic sputum media are also available, such as SCFM3, which contains p-aminobenzoic acid, NAD+, adenine, guanine, xanthine, and hypoxanthine⁴², and SCFM4, which is an iron-deficient version. Furthermore, the use of synthetic sputum media, such as cystic fibrosis lung media (CFLM) and cystic fibrosis sinus media (CFSM), developed by Ruhluel and colleagues⁴³, would allow for compartmentalization of the airways to visualize the behavior and effect of anti-virulence therapeutics on polymicrobial biofilms in different environments.

While AlgD is a key enzyme in the biosynthetic pathway of alginate, its expression level alone should not be taken as an indication of the production of alginate. Other factors, such as post-transcriptional and post-translational modifications, as well as regulatory mechanisms, can influence the amount of alginate produced, leading to reduced biofilm formation, increased susceptibility to environmental stress, and enhanced antibiotic sensitivity where alginate production is reduced^12,14. There are several anti-virulence therapeutics in development that target the alginate protein where this model would not be appropriate for determining their efficacy and where it would be more beneficial to directly measure alginate production or a functional read-out⁴⁴. Also, some anti-virulence therapies may target different stages of alginate biosynthesis or regulation, so measuring alginate production directly captures the net effect of the therapy on the virulence factor⁴⁴. Nevertheless, several of the anti-virulence therapies in current studies do target key alginate genes, and gene expression models such as these would be appropriate to study the effects of such therapies^13,14. Techniques such as this can also be easily adapted for high-throughput screening, allowing for the rapid assessment of multiple samples or conditions, which is beneficial in large-scale studies involving initial screening of potential anti-virulence compounds.

Overall, single-species and multi-species biofilm models represent valuable tools for studying the expression of infection-relevant genes of P. aeruginosa under CF-appropriate conditions. The large-scale rollout of modulator therapy has revolutionized the landscape of CF health care. However, despite the array of clinical benefits of CF modulators, studies have shown that clonal lineages of Pseudomonas are still circulating, highlighting that new therapies to manage P. aeruginosa lung infections are still required^45,46. Despite the discussed limitations, the application of these models could be useful within the discovery phase of the drug development pipeline to assess the effectiveness of compounds within the CF lung niche and the interactions of novel anti-virulence compounds with known antibiotics within this environment.

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Breathe-easy sealing membrane for multiwell plates	Sigma Aldrich	Z763624
Bunsen burner
Cellulase (100 mg/mL)	Cambridge Bioscience	HY-B2220-5g
CN supplement	Sigma Aldrich	1.07624
Direct-zol RNA Miniprep kit	Zymo Research	R2050
Dnase I	Thermo Fisher Scientific	EN0521
DNase/Rnase-Free water	VWR	J60610.XC
Gentamycin sulfate (30 µg/mL)	Sigma Aldrich	G3632
Glycerol (1 %)	Thermo Fisher Scientific	15514029
GoTaq qPCR Master Mix	Promega	A6101
iScript Cdna Synthesis Kit	Bio-Rad	1708890
Penicillin (250 U/mL)	Scientfic Laboratory Supplies	P3032-10MU
Qubit 4 Fluorometer	Thermo Fisher Scientific	Q33238
Qubit RNA BR Assay Kit	Thermo Fisher Scientific	Q10210
Qubit tubes	Thermo Fisher Scientific	Q32856
Resazurin solution (0.02 %)	Thermo Fisher Scientific	R12204
RNAlater Soln	Invitrogen	AM7021
Streptomycin (250 µg/mL)	Sigma Aldrich	S9137
Thermal cycler (we used the T100 Thermal Cycler)	Biorad	1861096
TRIzol Reagent	Invitrogen	15596026
Vancomycin (3 µg/mL)	Sigma Aldrich	1709007