To determine the effects of M. catarrhalis on airway epithelial cells, we infected cultures of airway epithelial cells differentiated at air-liquid interface (ALI) with M. catarrhalis strain MC14 and monitored bacterial abundance and cellular responses. Over a 5-day incubation period, the number of cell-associated bacteria decreased by ~3 log units ( Figure 1A ). We did not observe any adverse effects of M. catarrhalis on bronchial epithelial cell morphology or ciliary motion throughout infection. Trans-epithelial resistance (TEER), a measure of epithelial barrier function, did not change at 24 h and tended to increase by 48 h post-infection ( Figure 1B ). This is in contrast to the effect of another common respiratory tract pathogen, S. aureus, which caused a decrease in TEER indicating damage to the epithelial barrier.

To determine the effects of RV infection on M. catarrhalis adherence and survival, we inoculated the apical surface of differentiated bronchial epithelium with RV-A16, followed 2 hours later by infection with a clinical isolate of M. catarrhalis (strain MC14). Repeat experiments in cells from four epithelial cell donors demonstrated that RV-A16 infection significantly increased cell-associated M. catarrhalis (geometric means for MC = 2.2×10⁴, MC+RV = 4.9×10⁵; P = 0.003) ( Figure 2A ). Notably, M. catarrhalis did not influence RV replication ( Figure 2B ). RV infection caused significant cytotoxicity, as measured by LDH release (mean for control = 0.08, MC = 0.11, RV = 0.67, MC+RV = 0.64; control vs RV or MC+RV, P < 0.0001) ( Figure 2C ) that was not altered by M. catarrhalis. Both RV replication (ρ = 0.86, P = 0.0004) ( Figure 2D ) and cytotoxicity (ρ = 0.72, P = 0.007) ( Figure 2E ) were positively correlated with cell-associated M. catarrhalis. These findings suggest that RV replication promotes M. catarrhalis attachment or persistence. It is of note that of the four BEC donors that we analyzed, two of them (denoted by circle and squares), had greater bacterial burden with RV co-infection than the other donors. These donors also had more RV replication and RV-induced cytotoxicity. These observations suggest that some individuals may be more susceptible to M. catarrhalis colonization during RV infection.

We next tested whether RV and/or M. catarrhalis infection influenced epithelial gene expression. Using principal component analysis (PCA), PC1 was related to RV infection and PC5 was related to M. catarrhalis infection ( Figure 3A ). RV induced a robust transcriptional response (5143 mRNAs), while M. catarrhalis alone upregulated only 15 mRNAs. There were no significant differences in gene expression between cells infected with RV vs. RV+MC ( Figure 3B ). Gene Ontology (GO) enrichment analysis of significantly expressed genes revealed that RV induced a number of innate immune and antiviral pathways. Figure 3C shows the top 10 enriched pathways for each treatment. M. catarrhalis downregulated the expression of 15 genes that are associated with chemotaxis and cell proliferation. Among them were the transcription factors early growth response (EGR) 1, 2, and 3, which are involved in mitogenesis and fibrinogenesis suggesting that M. catarrhalis could inhibit epithelial regeneration.

We next tested whether M. catarrhalis altered the epithelial immune responses to RV. RV-A16 infection alone significantly increased secretion of chemokines (CXCL5, CXCL10, eotaxin-3, and G-CSF) and proinflammatory cytokines (IL-1β, IL-6, and TNF-α). RV-A16 infection also led to the secretion of TNF-related apoptosis-inducing ligand (TRAIL) which activates cell death by binding to the TRAIL receptor (Girkin et al., 2017). Figure 3D is a representation of the cytokines that were measured. As expected, RV-A16 infection also led to the secretion of the anti-viral cytokine IL-28A (IFN-λ2). Moraxella catarrhalis alone did not significantly induce cytokine secretion from the epithelium and did not alter RV-induced cytokine secretion.

RV-A and RV-C cause more severe respiratory illness and are more often associated with increased detection of M. catarrhalis in vivo compared to RV-B (Bashir et al., 2018). To evaluate the differential effect of RV species on replication and M. catarrhalis cell adhesion and cell association, we infected fully differentiated ALI cultures from a single bronchial epithelial cell donor with two types of each RV species (A16, A35, B52, B72, C2, and C15). Compared to RV uninfected cells, both RV-C types significantly increased cell-associated M. catarrhalis while RV-B types did not (P = 0.0003 vs RV-C2; P = 0.03 vs RV-C15) ( Figure 4A ). M. catarrhalis did not affect RV replication ( Figure 4B ). Compared to RV-B types, RV-A and C types also caused more cytotoxicity as indicated by the increase in LDH levels ( Figure 4C ; one-way ANOVA; A16 and A36 P < 0.0001, C2 P = 0.002, C15 P = 0.02 vs B72). To examine strain-specific effects on cell association, we next tested four different strains of M. catarrhalis (three clinical isolates and laboratory strain 035E). RV-C15 infection induced increased adhesion of all of the bacterial strains to AECs. None of the strains caused significant epithelial cell death or augmented RV-driven cell death ( Figure 4D ).

To identify the factors that contribute most to M. catarrhalis cell association, we used a mixed-effects model that included RV species, RV replication, epithelial cell donor, and cytotoxicity. In this model M. catarrhalis cell association was significantly related to RV species (C > A > B, P = 0.003) and RV replication (P = 0.05).

As we were able to show that RV-C species had the greatest effect on M. catarrhalis cell association and epithelial outcomes such as cytotoxicity, we used RV-C15 as a representative to look at the timeline of infection. To determine how soon M. catarrhalis binds to RV-infected cells, we infected ALI cultures with RV-C15 for 16 h (approximately one replication cycle) and then added M. catarrhalis for 2-36 h. Moraxella catarrhalis had increased cell adhesion starting 2 h post-infection and persisting for at least 36 h post-infection ( Figure 5A ). As M. catarrhalis showed maximal adhesion at 8 h after infection, we used this time point to evaluate conditions affecting bacterial adhesion in subsequent experiments.

We next used immunohistochemistry to visualize M. catarrhalis-cellular interactions. Cells were infected with RV-C15 for 16 h and then incubated with M. catarrhalis for 8 h. Without RV infection, few bacteria adhered to the epithelium. In contrast, M. catarrhalis attached to the apical surface of RV-infected cells, particularly cells that stained positive for RV and that were extruded from the infected epithelium and are dying ( Figures 5B, C ). While M. catarrhalis has been reported to internalize into lung epithelial cell lines and primary cells by macropinocytosis (Slevogt et al., 2007), we did not observe internalized bacteria in RV-infected or uninfected cells. Analysis of the immunofluorescent signal intensities showed a significantly higher signal for UspA in the presence of RV-C15 co-infection (P < 0.0001) ( Figure 5D ).

M. catarrhalis contains several OMPs that are important for adhesion. These include lipooligosaccharide (Spaniol et al., 2008), UspA1 (Lafontaine et al., 2000; Spaniol et al., 2008), UspA2 (Lafontaine et al., 2000), and Hag/MID (Bullard et al., 2005). Human airway epithelial cells express CEACAM1, CEACAM5, and CEACAM6 (Klaile et al., 2013) and UspA1 has been shown to bind to the N-terminus of CEACAM1 (Hill and Virji, 2003). M. catarrhalis adhesion can be prevented by using monoclonal antibodies that bind to the N-terminus of CEACAM1 (Hill and Virji, 2003; Green et al., 2011). RV-A16 infection significantly increased the expression of CEACAM in the epithelial cell donors with increased M. catarrhalis cell association ( Figure 5A ). Infection with RV-A and RV-C induced the expression of CEACAM1 by ~30-fold, while RV-B and M. catarrhalis did not have an effect on CEACAM1 expression ( Figure 5B ). Blocking CEACAM with mouse monoclonal antibodies D14HD11 that bind to the N-terminus of all CEACAMs, that have previously been used to inhibit Neisseria meningitidis adhesion (Green et al., 2011), did not inhibit M. catarrhalis cell adhesion ( Figure 6D ) despite being significantly upregulated beginning at 18 hours post-infection with RV-C15 ( Figure 5C ). We also used rat monoclonal antibodies YTH71.3 to CEACAM that have previously been used to inhibit Moraxella catarrhalis (Hill and Virji, 2003), but did not see any effect (data not shown). Moraxella catarrhalis did not influence CEACAM1 expression ( Figures 5B, C ). This is in contrast to a previous study that showed induction of CEACAM1 by M. catarrhalis (Klaile et al., 2013). In addition, fluorescent microscopy demonstrated that bacterial adhesion did not colocalize with CEACAM staining of RV-infected epithelial cells (Pearson’s correlation coefficient range = -0.09 to 0.14; P value range = 0.1 to 0.8) ( Figures S1A, B ). In contrast, blocking UspA1 and A2 with mouse monoclonal antibodies significantly reduced RV-induced M. catarrhalis adherence to airway epithelium at 24 h ( Figure 6E ). Interestingly, UspA1 and UspA2 showed minimal induction even at 36 h post-infection with RV-C15 infection ( Figures S2A, B ), showing that the increased adhesion of M. catarrhalis was not due to the increased expression of the outer membrane adhesion protein UspA.

Moraxella catarrhalis strain MC14 was isolated from a clinical specimen by the University of Wisconsin Clinical Pathology Laboratory. In addition, M. catarrhalis strains RSM43 and RSM163 were isolated from the nasal secretions of children with acute respiratory illnesses (Kloepfer et al., 2014), and strain 035E was originally isolated from the middle ear fluid of a patient with otitis media (Unhanand et al., 1992). For epithelial inoculation, we cultured M. catarrhalis from a frozen glycerol stock onto chocolate agar plates that were incubated at 37 °C in a 5% CO₂ environment for 48 h. Bacteria were re-streaked onto fresh chocolate agar plates and cultured for a further 24 h.

RV strains were cloned from clinical isolates and grown as described previously (Nakagome et al., 2014). The types used in the study were RV-A16, RV-A36, RV-B52, RV-B72, RV-C2, and RV-C15 (Bochkov et al., 2011; Lee et al., 2012; Nakagome et al., 2014).

Moraxella catarrhalis was cultured overnight in 3 ml brain-heart-infusion (BHI) broth at 37 °C. The cultures were centrifuged at 21,130 × g for 5 min at ambient temperature and washed in 1 ml of Buffer 1 (150 mM NaCl, 10 mM EDTA, 20 mM Tris-HCl, pH 8). The cells were resuspended in 1 mL of fresh Buffer 1 with 10 µL of 100 mg/mL RNase A and 163 µl of 10% [wt/vol] SDS. The samples were incubated at 37 °C for 90 min. Subsequently, 20 µl of 20 mg/ml proteinase K were added and the samples were incubated at 65 °C for 20 min. After an additional incubation at 37 °C for 16 h to ensure complete lysis, genomic DNA was purified using standard phenol-chloroform extraction and precipitated with isopropanol. Genomic libraries for Illumina MiSeq 2×150-bp paired-end sequencing were prepared and sequenced by the University of Wisconsin-Madison Biotechnology Center. The raw reads were corrected using fastp 0.20.0 (Chen et al., 2018) and draft genomes were generated using SPAdes v3.11.0 with default parameters (Prjibelski et al., 2020).

Human bronchial and tracheal epithelial cells were obtained from residual tissue from lungs destined for transplantation in collaboration with the University of Wisconsin - Health Lung Transplant Program. The protocol was reviewed by the University of Wisconsin Institutional Review Board and was deemed “not human subjects research.” Cryopreserved aliquots of cells were thawed and expanded as monolayers in PneumaCult-Ex Plus Medium (StemCell Technologies) supplemented with 0.1% [vol/vol] gentamicin (Sigma) and 0.1% [vol/vol] fluconazole (Novaplus). Once the cells reached 80% confluence, they were transferred to 12-well plates with Transwell semi-permeable inserts (Corning 3460) and were allowed to differentiate in PneumaCult-ALI medium (StemCell Technologies) supplemented with 0.1% [vol/vol] gentamicin (Sigma) and 0.1% [vol/vol] fluconazole (Novaplus) at the air-liquid interface for at least 21 days when ciliary motion was observed. The culture medium was changed to an antibiotic-free medium containing 0.05% [vol/vol] hydrocortisone 48 hours prior to RV infection. All ALI cultures were used between 28 – 35 days of air lifting.

The apical surface of the epithelium was washed with pre-warmed phosphate-buffered saline (PBS) supplemented with 100 mg/l of Ca²⁺ and Mg²⁺and then infected with 10⁷ PFU of RV (in 50 μl of antibiotic-free culture medium) at a multiplicity of infection (MOI) of 10. The cells were incubated at 34 °C for 2 h, after which the apical surface was washed (3×) with pre-warmed PBS (containing Ca²⁺ and Mg²⁺). For infection of differentiated cultures, 3-4 colonies of M. catarrhalis were picked from the growth plates and used to make a bacterial suspension in pre-warmed PBS at a concentration of 2 McFarland units (~6×10⁸ CFU/ml). The apical surface was then infected with 50 μl of the M. catarrhalis suspension (~3×10⁷ CFU) and incubated at 37 °C for 48 h.

For time-course experiments in Figure 1A , differentiated respiratory epithelium was infected with 10⁷ PFU of RV-A16 at 34 °C for 2 h and then with 3×10⁷ CFU of M. catarrhalis. At specified timepoints after M. catarrhalis infection, the apical surface was washed with 0.5 ml of pre-warmed PBS (with Ca²⁺ and Mg²⁺) and cell lysates were collected in 350 µl RLT Plus Buffer from AllPrep DNA/RNA Mini Kit (Qiagen) containing 0.5% [vol/vol] Reagent DX (Qiagen) for quantitative PCR. To quantify cell associated live bacteria in Figure 5A , ALI cultures were infected with 10⁷ PFU of RV-C15 at 34 °C for 2 h and then for at 37 °C for a further 16 h before infecting with 3×10⁷ CFU of M. catarrhalis for specified durations. The apical surface washed with 0.5 ml of pre-warmed PBS (with Ca²⁺ and Mg²⁺). Cell associated bacteria were quantified by adding 100 μl of 1% [vol/vol] saponin on to the apical epithelial surface at RT for 15 min, washing with 0.5 ml of warmed PBS (with Ca²⁺ and Mg²⁺) and plating the serially diluted washes on chocolate agar. The colony counts were done 48 h after plating.

For trans-epithelial electrical resistance (TEER) measurements, the apical surface of ALI cultures were infected with 50 μl of the M. catarrhalis suspension (~3×10⁷ CFU) that was prepared as described earlier and incubated at 37 °C for 24 or 48 h. As a control, a clinical isolate of Staphylococcus aureus was cultured from a frozen glycerol stock onto blood agar plates that were incubated at 37 °C in a 5% CO₂ environment for 48 h. Bacteria were re-streaked onto fresh blood agar plates and cultured for a further 24 h. For infection of ALI cultures, a colony of S. aureus was picked from the growth plates and used to make a bacterial suspension in pre-warmed PBS at a concentration of 2 McFarland units (~6×10⁸ CFU/ml). The apical surface of the ALI cultures were then infected with 50 μl of the suspension (~3×10⁷ CFU) and incubated at 37 °C for 24 or 48 h.

The apical surface of the epithelium was washed with pre-warmed phosphate-buffered saline (PBS) supplemented with 100 mg/l of Ca²⁺ and Mg²⁺and then infected with 10⁷ PFU of RV (in 50 μl of antibiotic-free culture medium) at an MOI of 10. The cells were incubated at 34 °C for 2 h, after which the apical surface was washed (3×) with pre-warmed PBS (containing Ca²⁺ and Mg²⁺). For infection of differentiated cultures, 3-4 colonies of M. catarrhalis were picked from the growth plates and used to make a bacterial suspension in pre-warmed PBS at a concentration of 2 McFarland units (~6×10⁸ CFU/ml). The suspension was diluted 100-fold to obtain a suspension with a concentration of 6×10⁶ CFU/ml. The apical surface was then infected with 50 μl of the 6×10⁶ CFU/ml M. catarrhalis suspension (~3×10⁵ CFU) and incubated at 37 °C for 48 h. Following the incubation period, the apical surface was washed with 0.5 ml of pre-warmed PBS and 100 µl of 100 µg/ml of gentamicin was added to the apical surface and incubated at 37 °C for 1 hour. The apical surface was washed with 0.5 ml of pre-warmed PBS and 100 µl of 1% saponin was added to the apical surface for 15 minutes at room temperature. The apical surface was washed again with 0.5 ml of pre-warmed PBS and serial dilutions of the washes were plated to determine live intracellular CFU counts.

Culture medium in outer wells of the Transwell plate were removed and replaced with 1 ml of warmed DMEM/F12 medium (Gibco) and 0.5 ml of DMEM/F12 in the insert (on apical surface of cells). The cells and medium were allowed to incubate at RT for 15 min. The TEER was measured using an epithelial Voltohmmeter (EVOM2; World Precision Instruments, USA).

Nucleic acids from the epithelium (± bacteria) were harvested from the inserts using 350 µl RLT Plus Buffer from AllPrep DNA/RNA Mini Kit (Qiagen) containing 0.5% [vol/vol] Reagent DX (Qiagen). The lysate was transferred to PowerBead tubes (Qiagen) for bead beating (5 min, 50 oscillations/s, Qiagen TissueLyser LT). RNA and DNA were extracted as separate fractions (AllPrep^® DNA/RNA Mini Kit). Moraxella catarrhalis copB and RV RNA levels were quantified by real-time PCR using specific primers ( Table 1 ) as previously described (Greiner et al., 2003; Bochkov et al., 2011). The CFUe measurements were calibrated by using a standard curve. To generate the standard curve, M. catarrhalis strain MC14 was cultured overnight in tryptic soy broth. The suspension was serially diluted and plated to obtain CFU counts. Two milliliters of the original suspension were centrifuged at 13,000 rpm for 30 minutes. The resulting bacterial pellet was lysed by adding warmed (55 °C) Solution PM1 followed by DNA extraction (AllPrep^® Power Viral Kit, Qiagen). The extracted DNA was serially diluted to make a standard curve corresponding to 10⁶ to 10² CFU.

For the quantification of mRNA expression, cDNA was prepared from total RNA extracted from ALI culture lysates using TaqMan™ Reverse Transcription Reagents (Applied Biosystems). CEACAM1 mRNA expression was measured using specific primers (PrimerBank ID 329112546c1). UspA1 and UspA2 expression were quantified using specific primers as described previously (Lafontaine et al., 2001) (Wang and Seed, 2003). Gene expression was quantified using Power SYBR Green PCR Master Mix (Applied Biosystems) on an Applied Biosystems™ 7500 Real-time PCR system and normalized against the expression of the stable housekeeping gene PPIA (PrimerBank ID 114520617c1). The expression was normalized against copB. All primers used in the study are given in Table 1 .

RNA was converted to cDNA using the SMART-Seq v4 Ultra Low Input RNA kit for sequencing (Takara Bio cat. 634898). Standard library preparation was completed using the Illumina Nextera XT DNA Library Preparation Kit and library quality and concentration were assessed using an Agilent 2100 bioanalyzer. Indexed samples were pooled and sequenced on the Illumina Hi-Seq 4000 with 100 base-pair, paired-end sequencing to a minimum of 10 million mapped reads per sample. RNA-seq reads were mapped to genome assembly GRCh37(hg19) reference sequence using STAR (version 2.6.1) (Dobin et al., 2013). Read counts were adjusted to counts per million and normalized using TMM normalization (Robinson and Oshlack, 2010). PCA was conducted in R (version 3.6.2) using prcomp. voom was used to adjust for significant technical covariates (Law et al., 2014). Genes significantly associated with RV or M. catarrhalis treatment were identified using Limma linear mixed modeling in R (Limma R package version 3.5 https://bioconductor.org/packages/release/bioc/html/limma.html). An FDR-adjusted p value of 0.05 was used as a significance threshold, using the method of Benjamini and Hochberg (Hochberg and Benjamini, 1990).

Cellular cytotoxicity was estimated by measuring extracellular lactate dehydrogenase using a CytoTox 96^® Non-Radioactive Cytotoxicity Assay, (Promega) following manufacturer’s instructions.

The cytokine levels in the culture medium in the basal compartment of the ALI culture system were measured by multiplex ELISA using a MILLIPLEX MAP 10-plex Human Cytokine/TH17 Mag Kit (Millipore Sigma). Epithelial cell gene expression was assessed by using 2×150 bp paired-end NovaSeq platform at the University of Wisconsin – Madison Biotechnology Center’s Gene Expression Center Core Facility (Madison, WI) (Research Resource Identifier – RRID : SCR_017757) for RNA library preparation and the DNA Sequencing Facility (RRID : SCR_017759) for sequencing.

Gene enrichment analysis of differentially expressed genes was done using the ShinyGO 0.76 online software (Ge et al., 2020). The pathways were sorted by -log10 (FDR) values with an FDR cutoff of 0.05.

To block UspA1 and A2, we incubated 50 μl of a 2 McFarland unit M. catarrhalis suspension (~3×10⁷ CFU) with an equal volume of undiluted culture supernatant from the mouse hybridoma 17C7 (ATCC HB-11093) (Helminen et al., 1994; Aebi et al., 1997), (37 °C, 1 h). The mixture was then added to the apical surface of differentiated airway epithelial cells, with or without RV infection. Moraxella catarrhalis was quantified by qPCR at 8 and 24 h post-infection.

To block CEACAM, airway epithelial cells were infected with RV-C15 for 16 h. After washing with 0.5 ml of pre-warmed PBS, mouse anti-pan CEACAM antibody D14HD11 (ab4567, Abcam; 100 μl of 100 ng/ml) or of rat monoclonal anti-CEACAM antibody YTH71.3 (Santa Cruz Biotechnology; 100ul of 100 ng/ml) was added to the apical surface and incubated at 37 °C for 2 h. We washed the cells three times and then incubated them with 50 μl of a 2 McFarland unit M. catarrhalis (~3×10⁷ CFU, 37 °C, 8 h or 24 h).

Mature ALI cultures were fixed in 10% [vol/vol] neutral-buffered formalin, embedded in paraffin wax and sectioned. The sections were deparaffinized by heating (60 °C, 20 min) and then transferring to 3 changes of xylenes. The sections were rehydrated using graded ethanol solutions (100%, 95%, 80%, 70% and 50% [vol/vol]) and deionized water. Antigen retrieval was performed by heating at 80 °C in a water bath for 2.5 h. Sections were permeabilized (1% normal goat serum with 0.4% [vol/vol] Triton X-100, 5 min), blocked (PBS with 0.1%[vol/vol] Tween 20 and 5% [vol/vol] goat serum), and then incubated (overnight, 4 °C) with antibodies to RV-C15 VP1 (1:200) or mouse anti-pan CEACAM antibody D14HD11 (ab4567, Abcam) (1:2000) diluted in 1% [vol/vol] normal goat serum containing 0.1% [vol/vol] Tween 20. After washing (PBS with 0.1% [vol/vol] Tween 20, 5 min), sections were incubated with the secondary antibody (Alexa Fluor^® 488 goat anti-mouse), blocked for 1 h with 5% mouse serum in PBS, and incubated with the primary antibody to M. catarrhalis UspA (Helminen et al., 1994; Aebi et al., 1997)(undiluted culture supernatant from the mouse hybridoma 17C7 - ATCC HB-11093, overnight, 4 °C). After washing three times, the second secondary antibody (Alexa Fluor^® 555 goat anti-mouse) was added (1 h, ambient temperature). The sections were counterstained with 4’,diamidine-2’-phenylindole dihydrochloride at 1:1000 for 15 min before mounting with FluorSave Reagent (EMD Millipore). Images were acquired on a Nikon A1 R confocal microscope (20× objective) and on a Nikon AX R point scanning confocal system equipped with 4 high-sensitivity GaAsP detectors using a Plan Apo lambda S 100× silicone immersion objective (NA 1.35) at 2048 x 2048 pixel density and 4x scan zoom. The 100× images were processed with Denoise.ai after acquisition. Image analysis was done using RGB Profile Plot ( Figure 5D ) and Colocalization Finder (Figure S1B) plugins on ImageJ v1.53t.

Statistical analyses were performed using GraphPad Prism v.9.3.1 (GraphPad Software, Inc). Student’s ordinary t-test or the Mann-Whitney test were used to compare two groups. When more than two groups were compared, ordinary one-way ANOVA was used for the analysis. Correlation coefficients were calculated using Spearman’s rho statistic. To identify factors that significantly predict M. catarrhalis cell association, linear regression modeling with random slope and random intercept using SAS procedure Mixed was conducted. The likelihood ratio test (LRT) was used to for model selection. A p-value < 0.05 was considered statistically significant. SAS software (v.9.4, SAS Institute, Cary, NC) was used to develop the mixed-effects model.