Temperature influences commensal-pathogen dynamics in a nasal epithelial cell co-culture model

ABSTRACT Chronic rhinosinusitis (CRS) is an inflammatory disease of the paranasal sinuses, and microbial dysbiosis associated with CRS is thought to be a key driver of host inflammation that contributes to disease progression. Staphylococcus aureus is a common upper respiratory tract (URT) pathobiont associated with higher carriage rates in CRS populations, where S. aureus-secreted toxins can be identified in CRS tissues. Although many genera of bacteria colonize the URT, few account for the majority of sequencing reads. These include S. aureus and several species belonging to the genus Corynebacterium, including Corynebacterium propinquum and Corynebacterium pseudodiphtheriticum, which are observed at high relative abundance in the healthy URT. Studies have examined bacterial interactions between major microbionts of the URT and S. aureus, but few have done so in the context of a healthy versus diseased URT environment. Here, we examine the role of temperature in commensal, pathogen, and epithelial dynamics using an air-liquid interface cell culture model mimicking the nasal epithelial environment. Healthy URT temperatures change from the nares to the nasopharynx and are increased during disease. Temperatures representative of the healthy URT increase persistence and aggregate formation of commensal C. propinquum and C. pseudodiphtheriticum, reduce S. aureus growth, and lower epithelial cytotoxicity compared to higher temperatures correlating with the diseased CRS sinus. Dual-species colonization revealed species-specific interactions between Corynebacterium species and S. aureus dependent on temperature. Our findings suggest URT mucosal temperature plays a significant role in mediating polymicrobial and host-bacterial interactions that may exacerbate microbial dysbiosis in chronic URT diseases. IMPORTANCE Chronic rhinosinusitis is a complex inflammatory disease with a significant healthcare burden. Although presence of S. aureus and microbial dysbiosis are considered mediators of inflammation in CRS, no studies have examined the influence of temperature on S. aureus interactions with the nasal epithelium and the dominant genus of the healthy URT, Corynebacterium. Interactions between Corynebacterium species and S. aureus have been documented in several studies, but none to date have examined how environmental changes in the URT may alter their interactions with the epithelium or each other. This study utilizes a polarized epithelial cell culture model at air-liquid interface to study the colonization and spatial dynamics of S. aureus and clinical isolates of Corynebacterium from people with CRS to characterize the role temperature has in single- and dual-species dynamics on the nasal epithelium.

to CRS pathology, and microbial dysbiosis in the sinuses and URT is considered a key contributor to chronic inflammation, furthering disease progression (4)(5)(6).The healthy URT microbiome is highly diverse, but most sequencing reads come from relatively few bacterial genera (7).Reduced microbial diversity is a hallmark of many airway diseases, yet for CRS, findings have been inconclusive, with some studies reporting reduced diversity in CRS populations and others increased diversity (4).One perspective is that CRS dysbiosis is not necessarily driven by a reduction in diversity but either a shift from commensal microbes to pathogenic ones or instigation of pathogenic behavior from pathobionts.Among the prevalent and abundant members of the URT microbiome, Staphylococcus aureus is commonly associated with CRS and has higher carriage rates in CRS patient populations (8)(9)(10).CRS inflammatory responses have been linked to S. aureus-specific toxins, and S. aureus toxins and antibodies to S. aureus-secre ted factors can be identified in CRS tissue samples (11)(12)(13)(14).Longitudinal carriage of S. aureus is associated with worsening CRS symptoms, and carriage is also linked to disease recurrence after endoscopic sinus surgery (15,16).Interestingly, antagonistic interactions with S. aureus have been documented for several prevalent and abundant members of the healthy URT microbiome (17)(18)(19)(20)(21)(22)(23)(24).Despite this, S. aureus can reach an average relative abundance nearing 48% of the sinonasal microbiome in the presence of these competing microbionts (25).No studies to date have investigated the environmental factors leading to S. aureus dominance over other microbionts within the context of CRS.
Differences in nasal mucosal temperature may influence the commensal, pathogen, and host dynamics in CRS.The nasal mucosa of healthy people has been reported to vary between 29°C and 32°C for the nares and nasal turbinates (26,27).However, the nasal cavity undergoes significant changes with the onset of CRS, including significant mucus production, mucosal swelling, possible polyp formation, and decreased epithelial barrier integrity (28).Consequently, nasal obstruction and decreased airflow are often seen in CRS patients, which can be alleviated through endoscopic sinus surgery (3,(29)(30)(31)(32).Additionally, inflammation and bacterial infection, key components of CRS pathology, have been shown to increase skin temperatures in post-surgery wounds and skin ulcers (33,34).Reduced airflow and inflammation are both thought to increase the tempera ture of the nasal mucosa (35).Supporting this theory, thermographic imaging of CRS patients has successfully been used as a strategy to identify which nasal passage or sinus cavity is affected due to significantly increased temperatures (36)(37)(38).This difference in temperature can have profound effects on bacterial growth and behavior.Indeed, S. aureus has been shown to induce significant changes in its transcriptome and proteome when grown in rich medium at temperatures ranging from 34°C to 40°C, with higher temperatures increasing its hemolytic capability (39).
In this study, we examine how temperature affects the interactions between nasal epithelial cells, S. aureus, and two CRS clinical isolates of Corynebacterium, the most abundant genus in the healthy URT (7), using an air-liquid interface (ALI) model of human nasal epithelial cells (HNECs).Corynebacterium species are significantly depleted in CRS, and several studies have found Corynebacterium is negatively correlated with S. aureus on skin and in the nares (7,25,(40)(41)(42).Additionally, some Corynebacterium species have been shown to reduce S. aureus virulence and potentially induce autoly sis, which has made Corynebacterium a candidate for probiotic usage (17,23,(43)(44)(45).Understanding how environmental changes affect the microbial constituents of the URT is important for elucidating the changes in the microbiome observed in disease, including the growth of S. aureus.We hypothesize that increasing temperature leads to a favorable environment for S. aureus to outcompete Corynebacterium on the URT epithelium.This study characterizes the role of temperature in commensal, pathogen, and nasal epithelium dynamics, shedding light on the complex dynamics of chronic URT disease pathology.

Lower temperature improves Corynebacterium pseudodiphtheriticum and Corynebacterium propinquum persistence but limits S. aureus growth in culture
To assess the impact of temperature on S. aureus and Corynebacterium, we first evaluated bacterial growth in rich liquid culture medium.For S. aureus, we evaluated the thor oughly studied methicillin-resistant strain USA300 LAC that is representative of difficultto-treat S. aureus strains encountered in CRS (46).We utilized two clinical isolates of Corynebacterium from CRS patients (16,47), Corynebacterium propinquum and Coryne bacterium pseudodiphtheriticum, due to their prevalence and high relative abundance in the URT (7).We also tested the Corynebacterium glutamicum strain ATCC 13032, a well-characterized and commonly used industrial strain, to serve as a comparison for the airway-adapted Corynebacterium clinical isolates' growth in vitro.Each of these strains was grown in 96-well microtiter plate with brain-heart infusion (BHI) broth for up to 48 hours at 37°C and 30°C (Fig. 1).Absorbance at 600 nm was measured at 4, 8, 12, 24, and 48 hours, and bacteria were harvested to determine viable colony-forming units (CFUs) at each timepoint.S. aureus absorbance values were significantly higher at 37°C at each timepoint measured after the initial reading, and CFU counts showed a modest reduction in CFUs at lower temperature (Fig. 1A).To test whether this trend is seen in other strains of S. aureus, including isolates from CRS patients, we examined the growth differences of USA100 and three CRS isolates of S. aureus at 37°C and 30°C (Fig. S1) which matched the trends seen in Fig. 1A.C. glutamicum CFUs were significantly lower at 30°C early during growth, but overall, both measurements for C. glutamicum showed little difference between growth at low and high temperatures over time (Fig. 1B).Unlike C. glutamicum, C. propinquum and C. pseudodiphtheriticum CRS isolates had significantly greater CFUs at 24-and 48-hour timepoints when grown at 30°C and significantly increased absorbance readings at 30°C throughout the assay, with the greatest differences observed at later timepoints (Fig. 1C and D).

Lower temperatures boost C. propinquum and C. pseudodiphtheriticum persistence on HNECs while limiting S. aureus growth
To examine whether trends observed in vitro in liquid culture also apply to bacterial growth in association with the nasal epithelium, we tested adherence and growth of S. aureus, C. propinquum, and C. pseudodiphtheriticum on polarized human nasal epithelial cells at air-liquid interface at 37°C and 30°C (Fig. 2).S. aureus adherence at 1 hour was unaffected by temperature; however, growth on HNECs after 6 hours was significantly lower at 30°C compared to the 37°C group (Fig. 2A).Additionally, we measured growth on HNECs for USA100 and three CRS isolates of S. aureus (Fig. S2) which matched the trend seen here.Temperature did not greatly affect C. propinquum and C. pseudodiphther iticum adherence or growth at 6 hours (Fig. 2B and C).However, at later timepoints, evaluated C. propinquum had significantly higher CFUs at 24 hours at 30°C (Fig. 2B), and C. pseudodiphtheriticum had a similar increase in growth at 30°C at both the 18-and 24-hour timepoints (Fig. 2C).
Interestingly, we observed both Corynebacterium clinical isolates had lower CFUs at 18 hours compared to 6 hours when grown at the higher 37°C temperature (Fig. 2B  and C).To ascertain whether these differences stemmed from altered aggregation on the surface of HNEC cultures, we utilized fluorescence microscopy to visualize Corynebac terium growth on HNECs.We modified the pJOE7706.1 plasmid backbone (48) to express the red fluorescent protein, tdTomato, under the control of an an isopropyl β-D-1thiogalactopyranoside (IPTG)-inducible promoter.We then modified a transformation protocol developed for C. glutamicum and electroporated pJOE7706.1-tdTomatointo C. propinquum and C. pseudodiphtheriticum to obtain fluorescent strains for microscopy.We inoculated HNECs with each isolate and grew them at 37°C and 30°C for 6, 18, and 24 hours, omitting the 1-hour timepoint due to negligible differences in CFUs.Corynebacte rium aggregate formation on HNECs appeared substantially increased at 30°C at the 18and 24-hour timepoints (Fig. 3A and B).Consistent with the HNEC CFU experiments (Fig. 2B and C), both Corynebacterium isolates had reduced aggregate size at later timepoints observed for the 37°C group (Fig. 3A and B).3D volume views of z-stacks acquired for the 24-hour timepoint showed both C. propinquum and C. pseudodiphtheriticum isolates grew on the apical surface of the epithelial layer (Fig. 3C and D).To assess whether S. aureus followed the same trends observed via CFU counts, we repeated colonization of HNECs at the 6-hour timepoint with GFP-expressing S. aureus (49).Coverage of the HNEC surface was largely reduced at 30°C (Fig. 3E) consistent with CFU data.Notably, while S. aureus aggregate formation and size appeared to be mildly reduced at 30°C, the majority of reduction in S. aureus appeared to be from an almost complete lack of bacteria not in large clusters or aggregates (Fig. 3E).Similar to Corynebacterium, S. aureus aggregates were found on the surface of the epithelial layer (Fig. 3E).Additionally, quantification of bacterial biomass on HNECs further supported the temperature-dependent trends in bacterial growth we observed (Fig. S3).

Temperature alters polymicrobial growth on HNECs
Considering how low and high temperatures affected growth of each species alone, we next asked how temperature may modulate interactions between S. aureus and Corynebacterium when grown in co-culture on HNECs, an environment that more closely represents the polymicrobial setting in the URT.We developed two models to investi gate co-inoculation and sequential inoculation with S. aureus and each Corynebacterium CRS isolate (Fig. S4) and characterized the effects of incubation at 37°C or 30°C on dual-species colonization of HNECs.S. aureus CFUs remained significantly decreased at 30°C when grown with either C. propinquum or C. pseudodiphtheriticum using the co-inoculation scheme (Fig. 4A).S. aureus growth in co-culture did not differ substantially from S. aureus grown alone on HNECs, as determined by measuring the fold-change in CFUs from dual-species culture normalized to CFUs from single-species culture on HNECs (Fig. 4C).Sequential inoculation did not result in significant changes in S. aureus CFUs at lower temperatures or with C. pseudodiphtheriticum at higher temperatures (Fig. 4A).However, S. aureus had a significant reduction in CFUs when grown at 37°C with C. propinquum using the sequential model of infection (Fig. 4A).In contrast to S. aureus, growth of both Corynebacterium isolates benefitted from sequential inoculation at 37°C, with C. propinquum having the largest increase from single-species culture (Fig. 4C).Despite an increase in CFUs compared to single-species culture at 37°C (Fig. 4C), C. pseudodiphtheriticum had reduced colonization in the presence of S. aureus during sequential colonization of HNECs when compared to growth at 30°C (Fig. 4B).
We next examined growth following dual-species sequential inoculation using fluorescence microscopy and biomass quantification (Fig. 5; Fig. S3).Strikingly, sequential inoculation with S. aureus improved growth of C. propinquum on HNECs at 37°C at the 24-hour timepoint (Fig. 5A) supporting the increase in CFUs seen in Fig. 4. In contrast, C. pseudodiphtheriticum grew similarly in the presence of S. aureus as observed in singlespecies assays (Fig. 3B), although aggregates appeared smaller at 30°C in sequential inoculation with S. aureus than when grown alone (Fig. 5B).S. aureus growth on HNECs appeared substantially reduced when grown with C. propinquum at either temperature and with C. pseudodiphtheriticum at 30°C (Fig. 5).

Bacterial-induced cytotoxicity toward HNECs is reduced at lower tempera tures
Having established single-and dual-species assays to evaluate growth of Corynebac terium and S. aureus on HNECs, we next examined the role temperature may have in modulating cytotoxicity of each species toward the epithelium.Basolateral media were collected from single-and dual-species colonized HNECs from sequential model of inoculation with fresh cell culture medium added 6 hours prior to harvest.Using a lactate dehydrogenase release assay to test the basolateral media, we determined epithelial cytotoxicity for uninfected controls and colonized groups.S. aureus cytotoxic ity was not found to be significantly altered by temperature (Fig. 6).C. propinquum showed significantly increased cytotoxicity at 37°C compared to 30°C alone and in dual culture with S. aureus (Fig. 6).We also used a two-way ANOVA to examine the effects of colonization status (untreated controls, single-and dual-infected groups) and temperature on cytotoxicity.The main effect of colonization status on cytotoxicity was significant (P = 0.0149).Additionally, the main effect of temperature was also statistically significant (P = 0.001).However, the combined interaction between colonization status and temperature was not found to be significant (P = 0.56).

DISCUSSION
In this study, we sought to characterize the role temperature may play in modulating interactions between bacterial constituents of the URT in the presence of the nasal epithelium.The temperature of the healthy nasal epithelium is approximately 30°C, and this temperature increases during chronic inflammatory URT disease (26,27,30,35,37,50) .Given the microbial dysbiosis commonly observed in CRS populations (51), we hypothesized that increased temperatures in the diseased URT may benefit one of the major CRS pathogens, S. aureus (9,(12)(13)(14)16), allowing for it to outcompete commensal bacteria.Here, we evaluated how temperature affects a well-studied MRSA strain alongside two species of Corynebacterium collected from people with CRS (16) using a human epithelial cell co-culture model.
By first testing single species of growth in rich culture medium and on HNECs, we found the clinical isolates of Corynebacterium had reduced persistence at a higher temperature of 37°C compared to 30°C (Fig. 1C and D; Fig. 2B and C).In contrast, S. aureus growth benefited at a higher temperature, which supported our initial hypothe sis.Fluorescence microscopy supported these findings, showing substantial increases in Corynebacterium growth and reduction in S. aureus growth on nasal epithelial cells incubated at the lower temperature of 30°C (Fig. 3).Considering these findings in the context of colonization and pathogenicity, S. aureus is a notorious pathogen known to infect numerous body sites where the temperature is at or above 37°C, leading to osteomyelitis, sepsis, soft tissue infections, and pneumonia (10).S. aureus transcrip tional changes between 34°C and 40°C have been investigated (39) in rich medium, and significant increases in hemolytic activity at higher temperatures were observed, indicative of increased pathogenic potential at higher temperatures.Here, our study investigated the effects of temperature on S. aureus behavior when in the presence of epithelial cells, with the nutritional environment originating solely from epithelial sources, at temperatures representative of the healthy nasal mucosa and the impacted sinus in CRS disease.Although increased protease expression has been observed at lower temperatures in BHI (39), and protease activity is implicated in biofilm dispersal (52), S. aureus was mostly found in large aggregates on the surface of epithelial cells at lower temperatures (Fig. 3E).However, the nature of S. aureus biofilm formation may vary in different environments, relying on other secreted factors for dispersal such as nuclease and phenol-soluble modulins (PSMs) (53,54), with PSMs known to be more highly expressed at higher temperatures (39).
For Corynebacterium, the effects of temperature on growth and metabolism of C. glutamicum have been previously studied to optimize the growth of this environmental strain used in industry for production of amino acids, and lower temperatures were found to benefit its growth over longer time periods (55)(56)(57).Our data support this observation for CRS clinical isolates of Corynebacterium, with growth and persistence of C. propinquum and C. pseudodiphtheriticum both found to be higher at the lower temperature of 30°C compared to 37°C in rich medium and on epithelial cells (Fig. 1C and D; Fig. 2B and C).Interestingly, in vitro growth curves showed that although a lower temperature benefitted persistence of both Corynebacterium species at later timepoints, culture at 37°C increased growth of C. propinquum and C. pseudodiphtheriticum within the first 16 hours (Fig. 1C and D).Since there have been few studies investigating physiology and genetic regulation in nasal-colonizing species of corynebacteria, if these changes in growth may be due to nutrient depletion, bacterial cell lysis or activation of dispersal or anti-aggregation mechanisms at higher temperatures is unknown.No studies to our knowledge have investigated the effect of lower temperatures representa tive of the healthy URT mucosal surface on commensal URT Corynebacterium species and human epithelial cells.Additionally, although research on Corynebacterium diphtheriae identified adhesins important for colonizing the epithelium (58), there has been no characterization of biofilm or aggregate formation of this species or other corynebacteria on nasal epithelial surfaces over time, as we report here.Some Corynebacterium species are known to form aggregates (59), possibly due to the hydrophobic nature of their outer mycolic acid membrane, but it is unknown whether this occurs for commensal Corynebacterium in the URT.Our results suggest that Corynebacterium can develop large biofilm-like structures in the nutritional milieu of the URT, which may impact interactions between Corynebacterium and other nasal microbiome constituents.
When examining dual-species colonization of HNECs, viable colony-forming units compared to single-species controls were not markedly different for S. aureus, whereas C. propinquum had a substantial increase at 37°C when cultured in the sequential model of inoculation with S. aureus (Fig. 4).Interestingly, S. aureus co-inoculated with C. propinquum at 37°C achieved higher CFU than when inoculated sequentially after C. propinquum, suggesting C. propinquum before colonization may limit S. aureus growth or adherence in association with the nasal epithelium (Fig. 4A).In contrast, C. pseudo diphtheriticum largely maintained a temperature-sensitive trend in the presence of S. aureus, continuing to benefit at the lower temperature of 30°C.Fluorescence microscopy supported these data, confirming increased C. propinquum aggregates associated with HNECs at 37°C in the presence of S. aureus (Fig. 4).It is possible the increase of C. propinquum when grown together with S. aureus at higher temperatures is due to changes in available nutrients in the airway surface liquid due to metabolites produced either by S. aureus or from epithelial cells, which could result either from altered secretion due to the presence of staphylococcal proteins or increased S. aureus-induced cytotoxic ity.Of note, previous studies have found metabolic interactions between URT micro bionts, and many of these interspecies interactions have been documented showing synergism, which may explain the interactions seen for C. propinquum in the presence of S. aureus (60,61).The differences observed for S. aureus interactions with C. propinquum compared to C. pseudodiphtheriticum underscore that we know relatively little about species of commensal Corynebacterium, and further studies are warranted to uncover diverse mechanisms by which members of this genera interact with microbial constitu ents in the URT.
After focusing on how URT-colonizing bacteria responded to growth at different temperatures, we next asked how the nasal epithelium would adjust to these tempera tures and associated changes in bacterial growth by examining cytotoxicity.Intriguingly, C. propinquum co-culture induced higher levels of cytotoxicity than C. pseudodiphther iticum, with significantly higher measurements observed at 37°C (Fig. 6), the tempera ture at which colonization for C. propinquum was lowest according to both CFUs and microscopy (Fig. 2B and 3A).This suggests that temperature may play a significant role in the host response to bacterial colonization, as the uninfected controls showed no noticeable temperature-dependent difference in cytotoxicity.Temperature is likely a key factor affecting bacterial behavior in the URT, with increased temperature potentially acting as an environmental switch that induces a more pathogenic phenotype even in species that are normally considered to be commensals such as Corynebacterium.Overall, these data indicate a role for temperature in both colonization and bacterial behavior of C. propinquum and C. pseudodiphtheriticum that is species specific, as both isolates had similar growth at lower temperatures yet significant differences in epithe lial cytotoxicity.Considering this in the context of CRS, it is possible that the native microbiome constituents in the URT could have a vastly different impact on inflammation depending on the microenvironment of the mucosal epithelium.Elucidating factors involved in species-specific epithelial cytotoxicity of Corynebacterium species, using isolates collected from the healthy and diseased URT, is a topic that warrants further study.
Additionally, temperature can affect other aspects of airway physiology, such as mucociliary clearance (MCC), hydration of the periciliary layer (PCL), and production of mucins, that may influence microbial colonization.Rates of MCC and hydration of the PCL have been found to be diminished at lower temperatures (62), which could impact adherence and colonization, especially by non-motile organisms such as staphylococci and corynebacteria.Accumulation of high concentrations of mucins in diseases like cystic fibrosis (CF) and chronic obstructive pulmonary disease (COPD) is known to disrupt airway mucus homeostasis, dampening MCC and promoting chronic infections (63).Colder temperatures have been shown to induce mucin secretion by bronchial epithelial cells from COPD patients (64).Mucus accumulation can also lead to generation of steep oxygen gradients in CF mucus (65), and local hypoxia can support the existence of anaerobic microbes in the mucus layer in CF airways and CRS sinuses (66).S. aureus and Corynebacterium spp.are facultative anaerobes, and previous studies demonstrated that growth under low oxygen conditions greatly affects the S. aureus transcriptome and promotes metabolic adaptation through induction of pathways for fermentation of carbohydrates (67,68).Prior work on C. glutamicum confirmed the nar gene cluster facilitates anaerobic growth when nitrate is present as a terminal electron acceptor (69), and the oral microbe Corynebacterium durum was observed to produce free fatty acids when grown in low-oxygen static culture conditions that affected chain length of Streptococcus sanguinis in a strain-dependent manner (70).How metabolism and physiology are affected by low oxygen in URT-coloniz ing species of Corynebacterium have not been investigated to our knowledge.Mucindegrading anaerobes enriched from CRS samples were previously shown to modulate the expression of S. aureus metabolic and virulence-associated genes, suggesting an additional avenue by which changes in the URT environment may further influence microbial community composition and potentiate bacterial virulence (71).Future studies investigating temperature-induced changes in mucin production and PCL hydration in the healthy and diseased sinuses and nares could shed light on how these aspects of the host response affect bacterial colonization and chronic infection.
The work presented here addresses gaps in knowledge regarding how S. aureus interacts with Corynebacterium when colonizing the nasal epithelium and the role of temperature in regulating interspecies interactions in the URT, yet there are limitations to this study.The clinical isolates of Corynebacterium we evaluated were from subjects with CRS, and these strains could have acquired adaptations to the chronically dis eased URT due to selective pressures encountered in the CRS sinuses that lead them to interact differently with S. aureus.Future studies would benefit from investigating Corynebacterium species that may be more representative of the healthy URT.This work investigates interactions between species from two prevalent genera that colonize the URT; however, the respiratory microbiome comprises numerous microbes that exist in a complex polymicrobial environment (7,72).Future studies examining how S. aureus and Corynebacterium interactions are affected by the presence of additional commen sal URT colonizers or another pathobiont species in a triple-species or multi-member polymicrobial model would extend this work and better represent the diverse URT microbial community.The HNEC ALI culture model used here provides a controlled environment for studying bacterial interactions with the epithelium and the nutritional environment in the URT; however, this model lacks significant components of the innate and adaptive immune system, as well as other cell types present in the URT in addition to the epithelium.Considering how temperature affects inflammatory responses in in vivo models of CRS could uncover additional interactions that occur between S. aureus and Corynebacterium to influence disease progression or highlight potential avenues for more effective therapy to address chronic infections.

Bacterial strains and growth conditions
Bacterial strains and plasmids used in this work are summarized in Table 1. S. aureus and Corynebacterium strains were cultured overnight at 37°C with shaking in brain-heart infusion broth (BD Biosciences) unless otherwise noted.Cultures were inoculated using colonies grown on BHI with 1.5% agar (BD Biosciences) at 37°C.E. coli cultures were grown in Luria broth shaking at 37°C.The S. aureus USA300 LAC 13c clone used for all experiments was a gift from Tammy Kielian (73).The E. coli strain GM2163 was a gift from Chuck Turnbough.Corynebacterium propinquum and Corynebacterium pseudodiphtheriti cum strains were isolated from the sinuses of subjects with CF and chronic rhinosinu sitis by culturing from sinonasal swabs collected during endoscopic sinus procedures that were clinically indicated for the management of patients' chronic sinus disease (University of Pittsburgh IRB REN16110185) (16).Species identification was confirmed by PCR and sequencing of the 16s rRNA gene using primers 63f (5′-CAG GCC TAA CAC ATG CAA GTC-3′) and 1387r (5′-GGG CGG AGT GTA CAA GGC-3′) (74).Plasmids in S. aureus were maintained in 25 µg/mL of chloramphenicol (Sigma).Plasmids in Corynebacterium strains were maintained in 50 µg/mL of kanamycin (ThermoScientific).Plasmids in E. coli strains were maintained in 100 µg/mL of ampicillin (Fisher BioReagents).

Recombinant DNA and genetic techniques
Plasmid DNA was prepared from E. coli DH5-alpha using the NEB Monarch Plasmid Purification Miniprep Kit and then electroporated into S. aureus strain RN4220 as previously described (75).DNA was then moved from RN4220 into S. aureus LAC 13c through transduction with bacteriophage alpha-80 as previously described (80).For transformation into Corynebacterium strains, plasmid DNA was prepared from E. coli strain GM2163 (dam−/dcm−) (77).Restriction enzymes, polymerases, and enzymes for DNA modification purchased from New England Biolabs (Beverly, MA) were used according to manufacturer's instructions.Oligonucleotides were synthesized by Integrated DNA Technologies (Coralville, IA).Non-radioactive sequencing was performed at the University of Alabama at Birmingham Heflin Center for Genomic Sciences Core Sequencing Facility.

Preparation of electrocompetent Corynebacterium
Methods for preparing electrocompetent Corynebacterium were based on protocols described by Eggeling and Bott in the Handbook of Corynebacterium glutamicum (81).Electrocompetent Corynebacterium was made by starting a 5-mL overnight culture in BHI sorbitol (BHIS) medium, composed of BHI + 9.1% sorbitol (Fisher BioReagents), and culturing overnight at 30°C with shaking 200 rpm.Next, 100 mL of BHIS was inoculated with 2-5 mL of the overnight culture to achieve a starting OD 600 of 0.2, followed by 3-5 hours of growth at 30°C with shaking.Once an OD 600 of 2.0 was reached, bacteria were centrifuged at 3,400 RCF at 4°C.Cells were washed 3× with chilled 1 mM Tris-HCl and 10% glycerol (Fisherbrand).Bacteria were then resuspended in chilled 10% glycerol and aliquoted for storage at −80°C.

Transformation of Corynebacterium
Electroporation of Corynebacterium was accomplished by thawing aliquots of electro competent Corynebacterium on ice for 15 minutes prior to incubation with 200 ng of plasmid prepared from E. coli strain GM2163 for 15 minutes on ice.The mixture was transferred to a 2-mm gap electroporation cuvette (Fisherbrand) and electroporated using a BioRad Gene Pulser set at 2.5 kV, 25 µF, and 200 Ω.After electroporation, 1 mL of BHIS was added followed by a 6-minute incubation in a 46°C water bath.Bacteria were then incubated at 30°C with shaking for 1 hour, followed by plating 100 µL of bacterial suspension on BHIS agar plates containing 50 µg/mL kanamycin.

Cell lines and growth conditions
The human nasal epithelial cell line RPMI2650 (ATCC CCL-30) was maintained in complete Eagle's minimum essential medium (Corning) containing L-glutamine (Gibco), Plasmocin (InvivoGen), Penicillin-Streptomycin (Gibco), and 10% fetal bovine serum (US qualified, Gibco) at 37°C and 5% CO 2 unless otherwise noted.Adherent cells were washed once with Dulbecco's phosphate-buffered saline (DPBS) prior to trypsinization with 0.25% trypsin + EDTA (Corning).Cells were collected by centrifugation at 1,400 RCF for 3 minutes at 4°C and resuspended in media.Transwell inserts (Greiner Bio-One) were coated with Vitrogen plating medium (VPM), consisting of MEM without glutamine or phenol red (Gibco), 10 µg/mL fibronectin (Corning), 100 µg/mL of bovine serum albumin (Life Technologies), and 30 µg/mL of PureCol bovine collagen (Advanced Biomatrix), added to the apical side of the insert, and crosslinked under UV light for a minimum of 45 minutes.VPM was then removed, and 2.5 × 10 5 cells were added to the apical side of the insert.Apical and basolateral media were replaced every other day until 7 days after seeding, when apical media were removed for transition to ALI.After maintaining at ALI for seven additional days, both apical and basolateral sides of the insert were washed with DPBS, and media without Plasmocin or Penicillin-Streptomycin were added to the basolateral side only for infection assays.

Co-culture HNEC assays
Overnight bacterial cultures were centrifuged at 3,000 RCF and washed once with sterile phosphate-buffered saline (Fisher).Bacteria were resuspended in MEM without glutamine or phenol red (Gibco) at an OD 600 of 0.5.For inoculation of HNECs, S. aureus was further diluted in MEM to a final OD 600 of 0.02, and Corynebacterium isolates were diluted to a final OD 600 of 0.2.Then, 50 µL of bacterial suspension was added to the apical surface of the ALI HNEC cultures for 1 hour with incubation at 37°C or 30°C with 5% CO 2 .After 1 hour, apical media were removed, and co-cultures were further incubated until 6, 18, or 24 hours of post-inoculation were reached.Co-cultures were then washed once apically with MEM to remove non-adherent bacteria.Basolateral media were removed and aliquoted for cytotoxicity assays.A solution of MEM with 0.1% Triton X-100 (BioRad) was added to the transwell insert, followed by 15 minutes of orbital shaking at 250 rpm.Cells were scraped from the filter, and the cell suspension was transferred into 950 µL of MEM with 0.1% Triton X-100 for a total of 1 mL.Tubes were gently vortexed for 3 minutes, followed by serial dilution plating to determine CFUs.For sequential inoculation of HNECs, Corynebacterium strains were inoculated as described above and incubated for 18 hours, followed by S. aureus inoculation and incubation for an additional 6 hours.Differential plating to obtain S. aureus and Corynebacterium CFU counts was accomplished by plating dual-species groups on mannitol salt agar (Oxoid) to enumerate S. aureus and BHI agar plates with 100 µg/mL fosfomycin (TCI) to enumerate Corynebacterium.

Fluorescence imaging of co-culture biofilms
Biofilm assays were performed as described above, with the addition of 0.1 mM IPTG to MEM used for inoculation, using bacterial strains expressing GFP (S. aureus with pCM29) or tdTomato (Corynebacterium with pJOE7706.1-tdTomato).After each timepoint, basolateral media were removed, and the transwell insert was washed with Dulbecco's phosphate-buffered saline and fixed overnight at 4°C with cold 4% parafor maldehyde (PFA; Electron Microscopy Sciences) diluted in DPBS.After fixation, PFA was removed, and samples were washed with DPBS followed by permeabilization with DPBS with 0.1% Triton X-100 for 15 minutes at room temperature.Samples then washed with DPBS followed by staining with a Hoechst (Invitrogen) diluted in DPBS for 15 minutes while shaking at room temperature.Samples were washed with DPBS, and filter inserts were cut out and mounted onto slides with ProLong Gold antifade reagent (Invitrogen).
Microscopy was performed on a Nikon A1R confocal microscope with a Plan Apo VC 60× Oil DIC N2 lens to generate images for publication.For quantification, samples were imaged on a Nikon Eclipse Ti2 widefield microscope.Image analysis was carried out using the Nikon NIS-Elements AR software package (version 5.42.02Build 1801).For each of three to six independent biological replicates, three separate fields of view were imaged for quantification.Volume measurements were obtained for each image stack after automatic thresholding (Original Otsu method) was performed in NIS-Elements AR.

Cytotoxicity measurements
Basolateral media were collected from co-culture assays and stored at −80°C.For positive controls, 10 µL of 10× lysis solution from the CytoTox 96 Non-Radioactive Cytotoxicity Assay (Promega) was added to 90 µL of basolateral media and incubated on the apical side of ALI cultures for 45 minutes before collecting.Basolateral media samples were thawed on ice prior to using the CytoTox 96 Assay Kit, according to the manufacturer's instructions.Cytotoxicity was calculated as a percentage value using cell culture medium as the negative control and fully lysed cells as the positive control.In uninfected cells, cytotoxicity is displayed on the graph as the experimental negative control.

Statistical analyses
Statistical analyses were performed with Graph Pad Prism version 10.1.0(264) software (GraphPad by Dotmatics).Two-way ANOVA with pre-determined multiple comparisons was used to evaluate the statistical significance between temperature and either time or bacterial group.Non-parametric unpaired two-tailed t-tests were used to evaluate the significance between colony-forming units for growth curves in rich medium.The P value of 0.05 was used as the cutoff for statistical significance, and P values below 0.10 are indicated on the graph with the specific value shown.

FIG 3
FIG 3 Lower temperatures support large aggregate formation of Corynebacterium on human nasal epithelial cells and reduce S. aureus colonization.The 2D fluorescence microscopy of tdTomato-expressing C. propinquum (A) and C. pseudodiphtheriticum (B) on polarized air-liquid interface HNECs stained with Hoechst grown at 37°C or 30°C for 6, 18, and 24 hours.3D and side views of the 24-hour images for C. propinquum (C) and C. pseudodiphtheriticum (D) on Hoechst-stained HNECs.2D and 3D fluorescence microscopy images of GFP-expressing S. aureus (E) grown on HNECs for 6 hours.Images shown are representative of three to six independent biological replicates.Scale bar = 50µm.

FIG 5
FIG 5 Corynebacterium has species-specific interactions with S. aureus at higher temperatures.Dual-species culture on human nasal epithelial cells at air-liquid interface with GFP-expressing (green) S. aureus and tdTomato-expressing (red) C. propinquum (A) or C. pseudodiphtheriticum (B) grown using the sequential model of inoculation at either 37°C or 30°C.Images shown are representative of three to six independent biological replicates.Scale = 50µm.

FIG 6
FIG 6 Lower temperatures reduce human nasal epithelial cell cytotoxicity independent of S. aureus and Corynebacterium.Lactate dehydrogenase release assay on the basolateral medium of HNECs colonized with S. aureus, C. propinquum, C. pseudodiphtheriticum, or dual species using the sequential model of inoculation.HNECs were incubated with bacteria at either 37°C (red) or 30°C (blue).n = 4 biological replicates.Two-way ANOVA.*P < 0.05.

TABLE 1
Bacterial strains and plasmids