Pseudomonas aeruginosa polysaccharide Psl supports airway microbial community development

Pseudomonas aeruginosa dominates the complex polymicrobial cystic fibrosis (CF) airway and is a leading cause of death in persons with CF. Oral streptococcal colonization has been associated with stable CF lung function. However, no studies have demonstrated how Streptococcus salivarius, the most abundant streptococcal species found in individuals with stable CF lung disease, potentially improves lung function or becomes incorporated into the CF airway biofilm. By utilizing a two-species biofilm model to probe interactions between S. salivarius and P. aeruginosa, we discovered that the P. aeruginosa exopolysaccharide Psl promoted S. salivarius biofilm formation. Further, we identified a S. salivarius maltose-binding protein (MalE) that is required for promotion of biofilm formation both in vitro and in a Drosophila melanogaster co-infection model. Finally, we demonstrate that promotion of dual biofilm formation with S. salivarius is common among environmental and clinical P. aeruginosa isolates. Overall, our data supports a model in which S. salivarius uses a sugar-binding protein to interact with P. aeruginosa exopolysaccharide, which may be a strategy by which S. salivarius establishes itself within the CF airway microbial community.


INTRODUCTION
Cystic fibrosis (CF) is the most common lethal genetic disorder in Caucasian populations [1]. Individuals with this disorder accumulate thick mucus in the lungs, and the inability to clear this mucus from the airways facilitates the colonization of microbes [1]. The bacterial species Pseudomonas aeruginosa is the leading cause of death for individuals with CF [2]. The prevalence of P. aeruginosa colonization in the CF population increases with age, with up to 60% of CF adults being colonized in their lifetime [3]. P. aeruginosa uses various strategies to cause persistent infections in the lung, including evasion of the host immune system, conversion to a mucoid phenotype, and biofilm formation [2]. These adaptation mechanisms, particularly the ability to form recalcitrant biofilms, render P. aeruginosa difficult to treat with antibiotics and permit the development of lifelong chronic infections that lead to rapid lung deterioration and mortality [4].
Biofilms are defined as a community of microbes that are attached to a surface and embedded in a protective extracellular matrix [5]. P. aeruginosa produces multiple exopolysaccharides that comprise its biofilm matrix. Non-mucoid strains of P. aeruginosa, which typically colonize the CF lung initially, produce the exopolysaccharides Pel and Psl [6]. Over the course of infection, non-mucoid strains will accumulate mutations and convert to a mucoid phenotype by switching to production of the exopolysaccharide alginate [7]. Pel, Psl, and alginate play an important role in antimicrobial resistance by preventing penetration of antibiotics into the P. aeruginosa biofilm [8,9]. Additionally, Psl is important for P. aeruginosa integration into polymicrobial biofilms [10].
Only in recent years have researchers began to study how crossspecies interactions in biofilms influence the composition of polymicrobial communities [10]. The importance of studying P. aeruginosa interactions with other microbes in the CF airway polymicrobial community is becoming increasingly recognized [11,12]. For instance, P. aeruginosa has been found to synthesize glutamate from precursor molecules secreted by Rothia mucilaginosa, another common microbe found in the CF lung [13]. Additionally, Psl produced by P. aeruginosa has been shown to interact with the staphylococcal protein A of Staphylococcus aureus to increase P. aeruginosa resistance to antibiotics [14]. Lastly, colonization of Stenotrophomonas maltophilia, an emerging CF pathogen, is promoted in murine lungs through integration into P. aeruginosa biofilms [15]. Relevant to our study, oral streptococci are increasingly recognized as core residents of the CF lung microbiota. Historically, these bacterial taxa have been thought to reside solely in the oral cavity and any detection of these microbes outside of this environment, particularly in the CF lung, was thought to be transient or attributed to oral contamination during sample collection. Multiple independent microbiome studies using sputum and bronchoalveolar lavage fluid have confirmed the presence of mitis and salivarius group oral streptococci in the CF airway [16][17][18][19]. Oral commensal streptococci have been shown to be associated with lung stability and increased microbial diversity in CF individuals. Streptococcus salivarius is the most prevalent streptococcal species found in the lungs of individuals with stable CF lung disease [16]. We previously reported that the second most abundant oral commensal found in the CF airway, Streptococcus parasanguinis, adheres to the mucoid P. aeruginosa exopolysaccharide alginate, resulting in the promotion of S. parasanguinis biofilm formation [20]. Therefore, it is important to understand how these commensals are incorporated into the CF polymicrobial community and potentially impact lung function in the CF population.
S. salivarius has been shown to commonly colonize the upper respiratory tract of infant children [21,22]. Because many individuals with CF become colonized with P. aeruginosa during adulthood, it is likely that S. salivarius is present in the lungs during the early stages of infection with non-mucoid P. aeruginosa [3]. In an effort to better understand how S. salivarius incorporates into biofilms with the major CF pathogen P. aeruginosa, we utilized two species in vitro and in vivo biofilm models to identify mechanisms that facilitate S. salivarius colonization. Here, we report that S. salivarius exploits the nonmucoid P. aeruginosa exopolysaccharide Psl to promote streptococcal biofilm formation. This enhanced biofilm phenotype was consistent when P. aeruginosa environmental and clinical isolates were grown with S. salivarius. Moreover, we found that the presence of a streptococcal maltose-binding surface protein, MalE, potentially facilitates the interaction between S. salivarius and P. aeruginosa Psl. Finally, we show that P. aeruginosa promotes S. salivarius colonization in a Drosophila melanogaster in vivo model of co-infection. Taken together, our study highlights a unique mechanism by which S. salivarius utilizes P. aeruginosa extracellular components to influence the CF airway microbial community by initiating and sustaining streptococcal colonization within the CF lung.

RESULTS
Non-mucoid Pseudomonas aeruginosa promotes Streptococcus salivarius biofilm formation To characterize interactions between the oral commensal S. salivarius and the CF lung pathogen P. aeruginosa, we co-cultured S. salivarius with an acute wound isolate (PAO1) and chronic CF isolate (FRD1). S. salivarius and the acute isolate PAO1 formed significantly more biofilm biomass when co-cultured compared to single species controls (Fig. 1). When we co-cultured S. salivarius with the mucoid isolate FRD1, biofilm formation did not increase compared to the single species controls (Fig. 1). To examine the relative species contribution in dual species biofilms with S. salivarius and PAO1, both planktonic and biofilm colony forming units (CFUs) were enumerated ( Fig. 2A). When co-cultured in TSBYE, both S. salivarius and P. aeruginosa biofilm CFUs increased significantly compared to single species controls. In contrast, no change in planktonic cell number was observed for either S. salivarius or P. aeruginosa in the presence of the other species (Fig S1A). We also co-cultured S. salivarius and P. aeruginosa in a synthetic cystic fibrosis sputum medium known as SCFM2, which mimics the nutrient profile found in the sputum of persons with CF [23]. When cultured in SCFM2, we observed an even greater increase in S. salivarius biofilm CFUs in the presence of P. aeruginosa (Fig. 2B). However, P. aeruginosa planktonic and biofilm cell number did not increase in the presence of S. salivarius in SCFM2, suggesting that this interspecies interaction exclusively enhances S. salivarius biofilm growth in synthetic CF sputum ( Fig. 2 (Fig S1B)), which better recapitulates the nutritional environment in the CF airway [23]. We confirmed these observations via confocal laser microscopy and observed an increase in S. salivarius and P. aeruginosa biofilm formation in TSBYE when co-cultured, and an increase only in S. salivarius biofilm formation in SCFM2 (Fig. 2C). Additionally, propidium iodide staining, which stains dead cells or nucleic acids [24,25] was prominent in our dual biofilm TSBYE group ( Fig S2). We did not observe a reduction in planktonic cells for either species in the dual biofilm group, suggesting that this staining is likely due to extracellular DNA, an important component of the biofilm matrix [24,26]. Enhanced S. salivarius biofilm formation in the presence of P. aeruginosa strain PAO1, but not FRD1 suggests that PAO1 possesses a distinct feature that promotes S. salivarius biofilm formation.
P. aeruginosa exopolysaccharide Psl enhances S. salivarius biofilm formation One major difference between the two strains is the exopolysaccharides produced within their respective biofilm matrices. Initial CF infecting P. aeruginosa strains that are acquired from the environment resemble PAO1 and produce Pel and Psl, whereas CF adapted strains like FRD1 overproduce alginate, largely due to accumulation of mutations in the mucA gene [6,7]. To investigate whether production of the exopolysaccharide Psl, the most prominent exopolysaccharide made by PAO1, promotes S. salivarius biofilm formation, we co-cultured S. salivarius biofilms with a PAO1 mutant deficient in Psl (PAO1ΔpslA), and PAO1ΔpslA complemented with a wild-type copy of pslA (PAO1pslA + ) [27]. When S. salivarius was co-cultured with PAO1ΔpslA, we observed significantly less dual biofilm formation compared to wild-type PAO1 co-cultures with S. salivarius. Co-culturing of S. salivarius with PAO1pslA + restored the dual species biofilm to levels similar to the wild-type PAO1 dual biofilm (Fig. 3A). When quantifying S. salivarius biofilm CFUs in both TSBYE and SCFM2, we found that co-culture of S. salivarius with PAO1ΔpslA resulted in a significantly decreased S. salivarius biofilm cell number compared to S. salivarius co-cultured with wild-type PAO1 (Fig. 3B). When co-cultured with PAO1pslA + , S. salivarius biofilm was restored to levels similar to wild-type PAO1 in SCFM2 and partially restored in TSBYE, further confirming that Psl enhances S. salivarius biofilm formation. We quantified and compared planktonic CFUs of PAO1ΔpslA and PAO1pslA + to wildtype PAO1 planktonic CFUs to confirm the lack of S. salivarius biofilm formation in the presence of PAO1ΔpslA was not caused by growth defects and lower cell density in the mutant strain co-culture (Figs. S3, S4). Additionally, we saw no growth advantage over time when we performed 16h growth curves of wild-type PAO1 cultured with and without S. salivarius, further confirming that S. salivarius biofilm promotion is not due to increased growth of P. aeruginosa (Fig. S5). To further confirm that P. aeruginosa Psl does indeed contribute to aeruginosa produce an enhanced biofilm in a dual species model. S. salivarius (Ss) was co-cultured with either of two P. aeruginosa strains (PAO1 and FRD1) in TSBYE medium in a 96-well plate for 16 h at 37°C (n = 3 biological replicates, 6 technical). Biofilm biomass was then measured using crystal violet staining. One-way ANOVA with Šίdák's multiple comparisons test. Error bars indicate standard deviation. ***p < 0.001, ****p < 0.0001. enhanced biofilm formation by S. salivarius, we co-cultured S. salivarius with two P. aeruginosa isolates known to produce little to no Psl and measured changes in biofilm biomass [28,29]. P. aeruginosa exopolysaccharide biosynthetic genes compete for the same precursor sugars, therefore, induction of one polysaccharide decreases the production of alternative exopolysaccharides [30]. Hence, overproduction of alginate in the mucA deficient strain of PAO1 results in a reduction in Psl production [29,30]. Significantly less dual biofilm formation was observed with the PAO1ΔmucA-alginate overproducing strain in comparison to wild-type PAO1 (Fig. S6A). We observed a similar result when co-culturing S. salivarius with the P. aeruginosa strain PA14, which lacks essential Psl biosynthetic genes and solely produces the exopolysaccharide Pel ( Fig. S6B) [31]. Additionally, purified Psl significantly increased the single species S. salivarius biofilm, further confirming the role of Psl in promoting S. salivarius (Fig. 3C). Lastly, pslA, a gene required for Psl production, was shown to be significantly upregulated in the presence of S. salivarius (Fig. 3D).
To understand the spatial relationship between S. salivarius and Psl within a dual biofilm, we performed confocal laser scanning microscopy on single and dual biofilms of S. salivarius and P. aeruginosa strains PAO1, PAO1ΔpslA, and PAO1pslA + stained with a FITC-conjugated α-Psl antibody to further characterize the role of Psl in promotion of S. salivarius biofilm formation (Fig. 3E, F). Consistent with our biofilm CFU quantifications, we observed a significant increase in S. salivarius biofilm in the presence of wildtype PAO1 as well as PAO1pslA + . When co-cultured with PAO1ΔpslA, no increase in S. salivarius biofilm formation was observed. Both S. salivarius and Psl were dispersed throughout the wild-type PAO1 and PAO1pslA + dual biofilms. Additionally, changes in S. salivarius biofilm architecture were observed in the presence of Psl (Fig. 3F). Overall, our findings show that Psl not only promotes S. salivarius biofilm development, but also modifies S. salivarius biofilm structure in dual species biofilms with P. aeruginosa.
To determine whether S. salivarius utilizes Psl only as a biofilm matrix scaffold or also metabolizes Psl, we tested whether S. salivarius could utilize purified Psl as a carbon source by monitoring planktonic growth in full-strength and 1:1 diluted THB media that was supplemented with Psl or glucose (Fig. S7). S. salivarius grew similarly in the presence of Psl compared to the nosugar control in full strength THB. Conversely, S. salivarius growth was enhanced in the presence of purified Psl in diluted THB media compared to the no-sugar control. This finding suggests that S. salivarius metabolizes Psl under conditions in which preferred carbon sources are limited. Our results demonstrate that Psl promotes S. salivarius biofilm formation, as well as S. salivarius planktonic growth via metabolism of Psl under specific nutrientlimited conditions. Non-mucoid P. aeruginosa strain PAO1 promotes S. salivarius biofilm formation. A Quantification of Ss and PAO1 biofilm CFUs was performed in a 6-h, 6-well mono-and dual-species biofilm model. Samples were cultured in TSBYE medium and B synthetic CF sputum (SCFM2), serially diluted and plated on THB agar. Data represent three biological replicates performed in triplicate. Student's t test. Error bars indicate mean ± SD. C Confocal microscopy was performed on Ss and PAO1-GFP single and dual species biofilms in both TSBYE and SCFM2. Ss was stained with hexidium iodide. **p < 0.01, ***p < 0.001. Ss was co-cultured with P. aeruginosa PAO1 strains (A) PAO1ΔpslA and PAO1 pslA + in TSBYE medium with 1% sucrose in a 96-well plate for 16 h at 37°C with 5% CO 2 (n = 3 biological replicates, 3 technical). Biofilm biomass was then measured using crystal violet staining. One-way ANOVA with Dunnett's multiple comparisons test. B Quantification of Ss biofilm-forming cells after co-culturing with PAO1, PAO1ΔpslA, and PAO1 pslA + in TSBYE (left) and SCFM2 [50] in a 6-h, 6-well model at 37°C with 5% CO 2 (n = 3 biological replicates, each with 3 technical replicates). One-way ANOVA with Šίdák's multiple comparisons test. C 0.5 mg/mL purified Psl was added to Ss single cultures in TSBYE with 1% sucrose in a 96-well 16-h biofilm. Crystal violet staining was used to quantify biofilm biomass. D qPCR quantification of P. aeruginosa pslA expression compared to 16S rRNA control. Student's t test. Fluorescence microscopy images at 60× magnification of 16-h single (E) and dual species (F) biofilms of Ss and PAO1, PAO1ΔpslA, and PAO1pslA + cultured in TSBYE supplemented with 1% sucrose. Ss was stained with hexidium iodide, and Psl was stained with a FITCconjugated α-Psl monoclonal antibody. Scale bar: 20 μm. *p < 0.05, **p < 0.01, ****p < 0.0001.

S. salivarius maltose-binding protein MalE plays a role in promotion of S. salivarius biofilm formation both in vitro and in vivo
To identify candidate S. salivarius proteins that could be involved in P. aeruginosa-dependent biofilm promotion, we examined the protein profile of whole-cell lysates of S. salivarius and P. aeruginosa single and dual cultures (Fig. 4A). Liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis revealed the overexpression of one~50 kDa protein that was identified as the S. salivarius maltose-binding protein MalE. MalE was overexpressed in dual cultures with P. aeruginosa, but not in single S. salivarius cultures. To examine whether MalE is involved in P. aeruginosa-dependent promotion of S. salivarius biofilm formation, we added anti-MalE antibodies to single and dual cultures of S. salivarius and P. aeruginosa. We found that anti-MalE antibodies significantly inhibited dual biofilm formation in a dosedependent manner, while growth of single species biofilms was unaffected (Fig. 4B). These findings suggest that MalE is involved in promotion of S. salivarius biofilm formation in the presence of P. aeruginosa. To determine whether P. aeruginosa promotes S. salivarius colonization in an in vivo model of co-infection, Drosophila melanogaster were co-infected with subcultures of S. salivarius and P. aeruginosa, and bacterial CFUs were enumerated after 24 h. Colonization was performed with and without 10 µg/mL of α-MalE antibodies to test whether colonization of S. salivarius is MalE-dependent. S. salivarius colonization significantly increased in the presence of P. aeruginosa, while no change in P. aeruginosa colonization between groups was observed (Fig. 5A). Consistent with our in vitro data, the addition of 10 µg/mL α-MalE antibodies caused a significant decrease in S. salivarius colonization in the dual infection group but had no effect on colonization during single species infections (Fig. 5A). These findings suggest that MalE is required for increased colonization by S. salivarius in the presence of P. aeruginosa.
To examine whether the promotion of S. salivarius colonization is also Psl-dependent in an in vivo model of infection, we infected flies with subcultures of PAO1 and PAO1ΔpslA with and without S. salivarius, then quantified CFUs per fly after 24 h. The presence of wild-type PAO1 significantly promoted S. salivarius colonization, while the presence of PAO1ΔpslA marginally, but not significantly, promoted S. salivarius colonization (Fig. 5B). Additionally, S. salivarius did not promote P. aeruginosa colonization. Our results demonstrate that in vivo colonization of S. salivarius is both MalEdependent and Psl-dependent.
Environmental and clinical non-mucoid isolates of P. aeruginosa enhance dual biofilms with S. salivarius To test whether our enhanced dual biofilm phenotype was consistent with non-lab adapted strains of P. aeruginosa, we isolated ten non-mucoid environmental P. aeruginosa isolates from local  water sources in Birmingham, Alabama. Biofilm formation significantly increased to levels similar to that of PAO1 when all ten isolates were co-cultured with S. salivarius (Fig. 6A). S. salivarius was also co-cultured with fifteen non-CF clinical P. aeruginosa isolates. Out of fifteen isolates, nine produced an enhanced dual biofilm in combination with S. salivarius (Fig. 6B). Lastly, three non-mucoid and three mucoid CF clinical P. aeruginosa isolates were co-cultured with S. salivarius (Fig. 6C). One out of the three non-mucoid CF isolates promoted biofilm formation to a greater degree than PAO1, while one out of the three mucoid isolates significantly promoted dual biofilm formation to a lesser extent than PAO1. Overall, these findings indicate that the ability to promote S. salivarius biofilm formation is common among both environmental and clinical isolates of P. aeruginosa.

DISCUSSION
Research in recent years has emphasized that a large percentage of bacteria present within any given environment exists in biofilms, rather than a free-floating, planktonic state. Additionally, bacterial species usually exist within polymicrobial communities rather than mono-species infections [32]. However, few studies have examined the role of polysaccharides in the development of mixed-species biofilms. In this paper, we demonstrate a mechanism by which a commensal bacterium utilizes a polysaccharide from a pathogenic species to promote its own biofilm formation. Our results are significant because this mechanism could potentially change the microbial composition of the CF lung by promoting commensal streptococci colonization.
The importance of studying bacterial infections in the context of the CF lung polymicrobial community is becoming increasingly recognized. A seminal study that reported oral commensal streptococci were associated with improved lung function identified S. salivarius as the most abundant oral streptococcus found in the lungs of clinically stable individuals with CF [16]. However, no studies have examined how S. salivarius colonizes the CF lung polymicrobial environment and impacts lung function. In this study, we demonstrate S. salivarius may incorporate itself into polymicrobial biofilms by using a streptococcal sugar-binding protein, MalE, to interact with the P. aeruginosa exopolysaccharide Psl in both in vitro and in vivo fly models of co-infection.
Our findings are significant because interactions between S. salivarius and P. aeruginosa may have an impact on P. aeruginosa virulence and pathogenesis. Although P. aeruginosa viability was not reduced in our in vitro and in vivo models of co-infection, S. salivarius may potentially interfere with P. aeruginosa pathogenesis and impact CF lung function by disrupting Psl availability to P. aeruginosa. Psl is a polysaccharide known for its role in resistance against antibiotic treatment and persistence of lung infections [9,33,34]. Our findings demonstrate that S. salivarius not only uses Psl as a biofilm scaffold, but can also metabolize Psl. This sequestering of Psl by S. salivarius could interfere with the ability of P. aeruginosa to use Psl as a mechanism for persistence.
We have previously shown that oral commensal streptococci can use cell surface adhesins to bind to the P. aeruginosa exopolysaccharide alginate to promote its own biofilm formation [20]. MalE has been shown to be anchored to the cell surface of Gram-positive bacteria, therefore allowing it to interact with extracellular molecules [35]. MalE in Group A Streptococcus has a wide range of sugar substrates, which has been implicated in helping Streptococcus species adapt to different host environments and support colonization [36]. Additionally, the maltosebinding protein of Streptococcus mutans along with many other bacterial species has been shown to also bind and transport sucrose in addition to maltose [37]. The Psl structure contains multiple sugars, including galactose, mannose, rhamnose, and Fig. 6 Environmental and clinical isolates of P. aeruginosa promote dual biofilm formation. Ss was co-cultured with A environmental, B non-CF acute, or C CF non-mucoid and mucoid isolates of P. aeruginosa in TSBYE with 1% sucrose in a 96-well 16-h biofilm model. Biofilms were stained with crystal violet to measure biofilm biomass (n = 3 biological, 3 technical). Error bars indicate mean ± SD. One-way ANOVA with Dunnett's multiple comparisons test (A + B). One-way ANOVA with Holm-Šίdák's multiple comparisons test (C). **p < 0.01, ***p < 0.001, ****p < 0.0001.
glucose [38], which S. salivarius MalE may be utilizing to promote biofilm formation. Our previous findings and current results suggest MalE as a possible candidate protein that facilitates oral commensal biofilm promotion by binding to exopolysaccharides produced by P. aeruginosa.
Numerous studies have demonstrated the potential health benefits of S. salivarius. In combination with Streptococcus oralis, S. salivarius has been shown to inhibit biofilm formation by six pathogens that commonly infect the upper respiratory tract, including Staphylococcus aureus, Staphylococcus epidermidis, Streptococcus pyogenes, Streptococcus pneumoniae, Propionibacterium acnes, and Moraxella catarrhalis [39]. Persons with CF infected with P. aeruginosa experience respiratory exacerbation episodes characterized by a large inflammatory response associated with increased pro-inflammatory cytokines IL-1, IL-6, IL-8, and TNF-α [40]. S. salivarius downregulates the innate immune response to P. aeruginosa in infected human epithelial cells [41] and inhibits both the pro-inflammatory NF-κB pathway in vitro and inflammation in an in vivo colitis mouse model [42]. Additionally, S. salivarius has been shown to downregulate IL-8 production induced by P. aeruginosa in human bronchial epithelial cells [41]. These observations warrant further studying of interactions between S. salivarius and P. aeruginosa to understand how S. salivarius affects the inflammatory host response to P. aeruginosa and how, in turn, this affects CF lung tissue damage.
Because S. salivarius colonizes the upper respiratory tract early in life and P. aeruginosa colonizes individuals with CF more commonly as they age, we would expect S. salivarius to be present in the CF lung during the early stages of infection when nonmucoid P. aeruginosa is acquired from the environment. All ten of our non-mucoid environmental P. aeruginosa isolates were able to produce an enhanced dual biofilm when co-cultured with S. salivarius.. The ability of S. salivarius to create an enhanced biofilm with environmental strains suggests that S. salivarius could be utilizing P. aeruginosa Psl from these non-mucoid strains to colonize the lungs during early P. aeruginosa infection. Further studies are required to understand how S. salivarius colonization impacts early P. aeruginosa infection. While only one of the three non-mucoid CF isolates produced an enhanced dual biofilm, the two other isolates displayed colony morphology differing from that of high Psl-producing strains such as PAO1, suggesting that their biofilm matrix is not primarily comprised of Psl. When S. salivarius was co-cultured with mucoid CF isolates of P. aeruginosa, one isolate produced a significantly enhanced dual biofilm. These results are consistent with previous literature that demonstrates Psl contributes to biofilm formation in some mucoid CF P. aeruginosa isolates [43].
In summary, we report a novel mechanism by which S. salivarius and P. aeruginosa interact within a biofilm and in our in vivo model. Furthermore, we have previously shown that the oral commensal S. parasanguinis interacts with the P. aeruginosa exopolysaccharide alginate to promote streptococcal biofilm formation [20]. Collectively, these studies illustrate potential mechanisms by which oral commensal streptococci interact with P. aeruginosa to alter the composition of the CF airway microbial community. In conclusion, our data suggest a model in which oral streptococci exploit P. aeruginosa exopolysaccharides, resulting in enhanced commensal biofilm development. The novel interactions between S. salivarius and P. aeruginosa revealed in this study could have implications for CF airway microbial community development and warrant further study.

MATERIALS AND METHODS Bacterial strains, culture conditions, and reagents
Strains S. salivarius K12, P. aeruginosa PAO1, P. aeruginosa PAO1ΔpslA, P. aeruginosa PAO1pslA + , P. aeruginosa PA14, P. aeruginosa FRD1, and P. aeruginosa environmental, acute clinical, and CF isolates were used in this study (Table 1). S. salivarius was grown on Todd-Hewitt Broth (THB) agar (Becton Dickinson) and cultured statically at 37°C in 5% CO 2 in THB. P. aeruginosa was grown on Pseudomonas Isolation Agar (PIA; Becton Dickinson) and cultured in Luria broth (LB; Fisher) and incubated while shaking (250 rpm) at 37°C. PAO1pslA + and PAO1-GFP were selected for on PIA with 100 μg/mL carbenicillin (Sigma-Aldrich) and were cultured in LB with 100 μg/mL carbenicillin. DH10b (E. coli) was cultured while shaking in LB at 37°C. After transformation, DH10b was cultured in SOC medium (Fisher) while shaking at 37°C for an hour.

Biofilm formation assays
Overnight cultures of S. salivarius and P. aeruginosa were sub-cultured in THB and LB, respectively, and grown to exponential phase (OD 600 0.5-0.8). Sub-cultures were then inoculated into Tryptic Soy Broth (MP Biomedicals) with 0.5% yeast extract (Fisher) (TSBYE) containing 1% sucrose at a dilution of 1:1000 for S. salivarius and 1:100 for P. aeruginosa. The two strains were inoculated either separately for the single species biofilm or together for the dual species biofilm assays. 200 μL of each sample was added to a 96well plate (Nunc) in triplicate and incubated statically at 37°C in 5% CO 2 for 16 h. The biofilms were then stained with 0.1% crystal violet and dissolved in 30% acetic acid [44]. Absorbance was measured at 562 nm to quantify biofilm biomass using the Synergy HTX Multi-Mode Microplate Reader (BioTek).

Quantification of P. aeruginosa and S. salivarius in co-cultures
Cultures were grown in either TSBYE with 1% sucrose, or synthetic cystic fibrosis sputum (SCFM2). SCFM2 was made as previously described [23]. To quantify colony forming units of each species, serial dilutions in TSBYE of planktonic samples from a 6-h six-well biofilm assay were plated on THB agar square plates (Fisher) using the track dilution method [45]. Remaining planktonic cells were aspirated off, and adherent biofilm cells were then washed two times with phosphate-buffered saline (PBS), scraped and resuspended in 3 mL of TSBYE. The resulting suspension was serially diluted and plated.

Construction of the PAO1 pslA + complemented strain
The pslA gene was cloned by PCR amplifying~500 bp upstream and downstream of the coding region from the wildtype PAO1 strain using primer sequences described ( Table 2). The PCR product was cloned into the EcoRI and BamHI sites of the pBluescript K(+) shuttle vector (Addgene). The resulting plasmid, referred to as pBKSNS1, was converted to a mobilizable plasmid by incorporation of a moriT into the HindIII site [46] and transformed into competent E. coli strain DH10b using a standard transformation method [47]. The pBKSNS1 plasmid was introduced into PAO1ΔpslA through triparental mating [46]. Plasmid conjugation events were selected for using PIA with 100 μg/mL carbenicillin.

Psl purification
Purification was performed on PAO1pslA + cultures grown overnight at 37°C in six-well plates in TSBYE supplemented with 1% sucrose and 100 µg/mL carbenicillin. Cultures were pooled together, diluted 1:1 with 0.9% NaCl, and agitated with 0.01 M EDTA by centrifuging at 200 rpm for 30 min at 4°C to detach cell-associated Psl. Cultures were then centrifuged at 10,000 g for 15 min at 4°C to remove bacterial cells. The resulting supernatant was filtered with a 0.22 μm vacuum filter to remove excess cell debris. Exopolysaccharide was then precipitated with 1:1 volume of cold 100% ethanol for 1 h at −80°C, and resulting precipitate was centrifuged at 15,000 g for 15 min at 4°C. The pellet was resuspended in PBS containing 1 mM CaCl 2 and 10 mM MgCl 2 and was subsequently treated with DNase I (100 µg/mL), RNase A (100 µg/mL), and Proteinase K (100 µg/ mL) for 2 h at 37°C, then lyophilized [48].
pslA quantification Biofilms were cultured in TSBYE supplemented with 1% sucrose in six-well plates for 6 h in 5% CO 2 . Biofilms were washed with PBS, and adherent bacteria was resuspended. RNA was isolated using the Direct-zol RNA Mini Prep kit (Zymo Research). cDNA conversion was performed with the iScript cDNA Synthesis kit (Bio-Rad), on the CFX96 Real-Time PCR System (Bio-Rad). P. aeruginosa 16S rRNA was used to quantify total RNA present in samples. Primers specific to P. aeruginosa pslA and 16S rRNA are listed in Table 2. The delta-delta CT method was used to calculate fold change of gene expression.

SDS-PAGE and mass spectrometry protein analysis
S. salivarius and P. aeruginosa were cultured planktonically in TSBYE medium at 37°C in single and dual cultures until OD 600~1 .8 was reached.
Cultures were spun down, resuspended in tris-buffered saline, and lysed using the Bead Blaster 24 (Benchmark Scientific). Cell debris was centrifuged, and the resulting supernatant was mixed with 6× Laemmli  buffer and ran on an SDS Page gel. Overexpressed bands were excised from the gels and digested with trypsin. The digested peptide fragments were analyzed for protein identification by LC-MS/MS as described previously [49] and sent to the UAB Mass Spectrometry Core for identification. We repeated the SDS-PAGE experiment three times, the only reproducible change was this~50 kDa band. We also excised the bands from three biologic replicates and analyzed by mass spectrometry.

Immunofluorescence and confocal laser scanning microscopy
Bacterial strains were grown in either TSBYE supplemented with 1% sucrose or in SCFM2 in a sterile eight-well treated μSlide (Ibidi). Samples were incubated at 37°C under 5% CO 2 for 16 h. Biofilm samples were gently washed twice with PBS to eliminate planktonic bacteria and then incubated in PBS with 1% bovine serum albumin (BSA) for 10 min. Samples were then stained for 10 min with hexidium iodide (Thermo Fisher) at a 1:1000 dilution and a FITC-conjugated α-Psl antibody (Creative Biolabs) at a 1:100 dilution in PBS with 1% BSA. The samples were washed with PBS once more and then analyzed for fluorescence using the Nikon A1R Confocal Laser Scanning Microscope (Nikon Instruments Inc.) at the University of Alabama at Birmingham High Resolution Imaging Facility.

Drosophila melanogaster colonization assay
Drosophila melanogaster flies were maintained on Jazz-mix Drosophila food (Fisher). Three to seven-day-old flies were treated with antibiotics for 3 days (50 µg/mL vancomycin, 50 µg/mL erythromycin, and 50 µg/mL ampicillin). Flies were separated into vials in groups of ten and subsequently starved for 3 h prior to infection. S. salivarius and P. aeruginosa cultures were grown to A 600 2.0. For single species infection groups, 1.5 mL of the respective culture was centrifuged for 6 min at 6000 g and resuspended in 100 µl of 5% sucrose. For dual infection groups, 0.75 mL of each culture were combined, centrifuged, and resuspended in 5% sucrose. Resuspended cultures were then pipetted onto sterile 21 mm filter disks (Whatman) and placed into plastic vials containing 5 mL of 5% sucrose agar. After 24 h of infection, flies were briefly washed in 70% ethanol to remove outside contaminants and then washed with sterile PBS. Flies were crushed and resuspended in 500 µl PBS. The resulting homogenate was then serially diluted and plated on THB agar plates using the track dilution method to quantify bacterial colony-forming units inside flies.

Statistical analysis
All graphs represent sample means ± SD. The Shapiro-Wilk normality test was used to determine distribution of datasets. Statistical analysis of normally distributed data was performed using either Student's t test or one-way ANOVA with Tukey's multiple comparisons test. For experiments with two or more factors (for example, absorbance and treatment), a twoway ANOVA was used with either Tukey's or Šίdák's multiple comparisons test. Tests were performed using Graphpad Prism version 9 for Windows, La Jolla California USA, www.graphpad.com. Data were considered statistically significant if p < 0.05.