Burkholderia cenocepacia–host cell contact controls the transcription activity of the trimeric autotransporter adhesin BCAM2418 gene

Abstract Cell‐to‐cell early contact between pathogens and their host cells is required for the establishment of many infections. Among various surface factors produced by bacteria that allow an organism to become established in a host, the class of adhesins is a primary determinant. Burkholderia cenocepacia adheres to the respiratory epithelium of cystic fibrosis patients and causes chronic inflammation and disease. Cell‐to‐cell contacts are promoted by various kinds of adhesins, including trimeric autotransporter adhesins (TAAs). We observed that among the 7 TAA genes found in the B. cenocepacia K56‐2 genome, two of them (BCAM2418 and BCAS0236) express higher levels of mRNA following physical contact with host cells. Further analysis revealed that the B. cenocepacia K56‐2 BCAM2418 gene shows an on–off switch after an initial colonization period, exhibits a strong expression dependent on the host cell type, and enhances its function on cell adhesion. Furthermore, our analysis revealed that adhesion to mucin‐coated surfaces dramatically increases the expression levels of BCAM2418. Abrogation of mucin O‐glycans turns BCAM2418 gene expression off and impairs bacterial adherence. Overall, our findings suggest that glycosylated extracellular components of host membrane might be a binding site for B. cenocepacia and a signal for the differential expression of the TAA gene BCAM2418.

TAAs form a large and diverse group of outer membrane proteins widely distributed in Gram-negative bacteria. They belong to a subfamily of autotransporter proteins and are secreted to the outer surface of the bacteria via the type Vc secretion system. These proteins have a typical trimeric surface modular architecture, composed of three identical monomers, with a C-terminal anchor and a variable extracellular set of fiber composed of stalk and globular-like head regions. While the membrane anchor domain is the defining feature of this class of proteins, highly conserved between all the TAA members, head and stalk organization is adaptable and vary among TAAs (Bassler, Hernandez, Hartmann, & Lupas, 2015;Cotter, Surana, & Geme 2005;Łyskowski, Leo, & Goldman, 2011).
In this work, we aimed to uncover the relevance of B. cenocepacia TAAs in the early stages of infection. In particular, our findings reveal the transcriptional alteration of BCAM2418 gene induced by the physical contact of the bacterium with bronchial epithelial cells.
Moreover, we found that overexpression of BCAM2418 gene contributes to the bacterial cell adhesion to host cells and is dependent on recognition of O-linked glycans from the host cell membranes.
Overall, this study not only defines the behavior of the TAAs during the step of bacterial adhesion but also provide insights aiming to determine potential targets for therapeutic proposals.

| Human cell lines and cell culture conditions
Two human bronchial epithelial cell lines were used: 16HBE14o-cell line, which is healthy lung cells expressing a functional CF transmembrane conductance regulator, and CFBE41o-cell line, which is homozygous for the delta F508 mutation corresponding to a CF airway. Both immortalized cell lines were kindly provided by Dr.
The four cell lines were incubated at 37ºC in a humidified atmosphere with 5% CO 2 . Bacterial lysis was achieved by enzymatic lysis with lysozyme and proteinase K (Qiagen). Total RNA was purified from a bacterial lysate using RNeasy mini kit (Qiagen), according to the manufacturer's protocol. To avoid contamination with genomic DNA, RNA was treated with RNase-free DNAse digestion kit (Qiagen) in a column during the purification process, for 1 hr at room temperature. To remove the DNA contamination from isolated RNA, overnight DNase (1 ml for 1.5 mg of RNA) treatment was performed at 37ºC followed by inactivation for 5 min at 65ºC. The total RNA concentration was estimated using a UV spectrophotometer (ND-1000 UV-Vis, NanoDrop Technologies).

| Quantitative assessment of TAAs genes expression upon host cell contact
For RT-PCR experiments, total RNA was converted to cDNA using TaqMan kit (Applied Biosystems) and then analyzed with Power SYBR Green master mix (Applied Biosystems), using primers to amplify TAA-encoding genes and SigA gene (used as an internal control; Table 1). All samples were analyzed in triplicate, and the amount of mRNA detected normalized to control SigA mRNA values.
Relative quantification of genes expression was calculated by using the ΔΔCT method (Livak & Schmittgen, 2001).

| Bacterial adhesion to epithelial cells
Adhesion experiments were carried out on 16HBE14o-, CFBE41o-, A549, and HeLa cell lines as described previously (Mil-Homens & Fialho, 2012), with some modifications. Cells were seeded in polystyrene microplates one day before infection at 1 × 10 6 cells/mL in a supplemented medium. The cells were infected with a multiplicity of infection (MOI) of 50:1. After infection, plates were centrifuged at 700 g for 5 min. The infected monolayers were incubated for a different time periods (15,30,120,180, and 300 min) at 37°C in an atmosphere containing 5% CO 2 . After incubation, each well was washed three times with PBS. For adhesion determination, the host cells were lysed by incubation with lysis buffer (10 mM EDTA, 0.25% Triton X-100) for 30 min at room temperature. The adhered bacteria were quantified by plating serial dilutions of the cell lysates. Results are expressed as a percentage of adhesion relatively to the initial bacterial dose applied. For further expression assays, the supernatant containing the nonadherent bacteria was recovered, and each well was washed three times with PBS. Cells and adherent bacteria were harvested by scratching and further resuspension in PBS. All the samples were centrifuged, suspended in RNAprotect Bacteria Reagent (Qiagen), and centrifuged again. The recovered pellets were storage at −80ºC until total RNA extraction and further gene expression assays.

| Extracellular digestion of surface-exposed host cell membrane components
16HBE14o-cells were seeded 1 day before infection at 1 × 10 6 cells/ mL. Before infection experiment, cellular monolayers were washed two times with HBS (HEPES buffer saline) and incubated with different digestive enzymes like: pronase E (Sigma-Aldrich; 3.12 μg/ml in HBS), trypsin (Gibco, ThermoFisher; 6.25 μg/ml in HBS,) for 1hr, O-glycosidase (New England BioLabs; 2,500 U/well), and PNGase F (New England BioLabs; 2,500 U/well) for 6 hr at 37°C CO 2 incubator. The supernatants with the resulting enzymatic-produced cellular fragments were recovered and inoculated with 5 × 10 7 CFU/mL of B. cenocepacia K56-2. The samples were incubated for 30 min at 37ºC. The concentrations of all the enzymes were optimized to guarantee the cellular and bacterial viability throughout the assay. HBS and serum-free MEM were used as controls.
The treated cellular monolayers were washed three times with HBS, and 1 ml of serum-free MEM was added to each well. The cells were infected with an MOI of 50:1. After infection, plates were centrifuged at 700 g for 5 min. The infected monolayers were incubated for 30 min at 37°C in an atmosphere containing 5% CO 2 . After incubation, the supernatant containing the nonadherent bacteria was recovered, and each well was washed three times with PBS. Cells and adherent bacteria were harvested by scratching and further resuspension in PBS. All the samples were centrifuged, suspended in RNAprotect Bacteria Reagent (Qiagen), and centrifuged again. The recovered pellets were storage at −80ºC until total RNA extraction and further gene expression assays.
The wells were washed twice with PBS. Approximately 5 × 10 7 CFU/mL was added to each coated well. The plates were incubated 15, 30 min, 2, 3, or 5 hr at 37ºC and washed 3 times with sterile PBS to remove unbound bacteria. The mucin-coated plates were also subjected to O-glycosidase treatment, as described previously. Adhesion to treated and untreated mucin-coated wells was performed during 2 hr at 37ºC.
For adhesion determination, the wells were treated with 0.5% (v/v) Triton X-100 solution to desorb the bound bacteria. Plates were incubated for 2 hr at room temperature under orbital agitation. One hundred microliters of the content of each well were removed, diluted in PBS, and plated on LB agar plates. Results are expressed as a percentage of adhesion relatively to the initial bacterial dose applied.
For further expression assays, the wells were scratched and the released bacteria suspended in PBS. The samples were centrifuged, suspended in RNAprotect Bacteria Reagent (Qiagen), and centrifuged again. The recovered pellets were storage at −80ºC until total RNA extraction and further gene expression assays.

| Statistical analysis
All experiments were performed in a minimum of three independent replicates. Relative comparisons were made between corrected values with ANOVA test for significance. A p value < .05 was considered statistically significant.

| B. cenocepacia K56-2 TAAs transcripts are produced at different levels after bacterial adhesion to bronchial epithelial cells
We have started this work by analyzing the expression profile of the 7 TAA genes in response to host cell contact (30 min). Thus, we mRNAs are the ones with higher and lower values of expression, respectively, spanning a difference of approximately 400× between them. BCAM2418 and BCAS0236 are the only genes that present significantly different levels of expression when compared to basal control (p values <.0001; Figure 1). Based on the increased expression of these two genes, we postulate a prominent role of their implicating proteins in the process of bacterial adhesion to host cells. Thus, we decided to pursue this study by characterizing the BCAM2418 gene, the one that has the higher shift in expression levels.

| BCAM2418 transcriptional levels after adhesion are reliant on the nature of host cells
To determine whether BCAM2418 expression after cellular contact was a direct response to a specific type of cell, we analyze BCAM2418 expression patterns after adhesion to a set of human cell lines. Figure 2a represented the results obtained for BCAM2418 expression after adhesion to HeLa, A549, 16HBE14o-, and CFBE41o-cell lines.

Gene Primer Sequence
SigA TA B L E 1 List of RT-PCR primers used in this study F I G U R E 1 Expression profile of TAA coding genes after adhesion to bronchial epithelial cells. Transcription levels of the 7 Burkholderia cenocepacia K56-2 TAA coding genes were obtained by qRT-PCR after 30 min of adhesion to 16HBE14o-cells. Results were normalized to the expression of the housekeeping SigA gene. Expression levels are represented as relative values in comparison to the expression levels in standard LB growth. All the results are from three independent experiments, and bars indicate SD. Expression of BCAM2418 and BCAS0236 is significantly higher when compared to standard LB growth. (****p < .0001)

| BCAM2418 gene expression profile during the early stages of infection
To follow the levels of BCAM2418 expression after adhesion to 16HBE14o-cell line (approximately 400-fold mRNA expression), we evaluate the transcription of this gene at early and later time points of cellular contact (from 15 min to 5 hr). The results represented in Figure 3 show a timeline pattern of BCAM2418 expression that reaches a peak after 30 min of adhesion. The same does not occur in the recovered nonadherent bacteria (not shown). The expression of this TAA gene seems to begin to increase soon after the first contact with the host cell (t = 15 min), starting to decrease after the adhesion process has taken place (t = 2 hr). Moreover, after 5 hr of cellular contact, the levels of BCAM2418 expression are lower than the ones obtained after the initial 15 min.

| Enzymatic treatment of the host cell surface before adhesion cause a decrease in BCAM2418 transcripts
To evaluate the requirement of a native cellular surface as a stimulus to induce BCAM2418 expression, we proceed with the enzymatic treatment of the host cell surface before the cellular adhesion. As shown in Figure 4a, the treatment with proteases (pronase E or trypsin) or O-and N-glycosidases before the adhesion event causes a significant reduction in the levels of

| Adhesion properties and BCAM2418 expression-based comparison of B. cenocepacia K56-2 to mucins and extracellular matrix proteins
Since one of the primary O-glycosylated-type protein of mucus is mucin, we tested the adherence capacity of B. cenocepacia K56-2 to mucins and the impact on BCAM2418 gene expression profile.
Two extracellular matrices (ECM) proteins, namely fibronectin and collagen type I, were used as controls. As shown in Figure 5a

F I G U R E 5
Burkholderia cenocepacia adherence to mucins and BCAM2418 transcript levels over different times of contact. (a) Adherence of B. cenocepacia K56-2 to BSA (control), collagen type I, fibronectin, and mucins during defining timelines-15, 30 min, 2, 3, and 5 hr. Results are expressed as the percentage of adhesion relatively to the initial bacterial load added to the cells. BSA was used as a negative control of the assay. All the results are from three independent experiments, and bars indicate SD. The binding capacity to mucins increases over time.
The adhesion percentage to mucins is significantly higher after 3h and 5h of contact (****p < .0001). (b) BCAM2418 transcript levels after adherence to mucins in a defined timeline. Transcription levels of B. cenocepacia K56-2 BCAM2418 coding gene were obtained by qRT-PCR after adhesion to mucin coatings during 15, 30 min, 2, 3, and 5 hr. Results were normalized to the expression of the housekeeping SigA gene. Expression levels are represented as relative values in comparison to the expression levels in standard LB growth. All the results are from three independent experiments, and bars indicate SD. Expression of BCAM2418 after 2, 3, and 5 hr of mucin adhesion is significantly higher when compared to standard LB growth. (**p < .01; ****p < .0001) BCAM2418 mutant, and hence, we were unable to perform the mutant phenotypic analysis and determine the impact of this gene product.
Previous studies have shown the specificity of Bcc species to interact with lung epithelial cells (McClean & Callaghan, 2009;Sajjan, Keshavjee, & Forstner, 2004). In this scenario, we aimed to disclosure the cellular component that is recognized by B. cenocepacia K56-2 and prompted an overexpression of the BCAM2418 gene. It is known that protein or glycoconjugates host receptors mediate binding of Bcc species to the airway host epithelium with bacterial surface components such as flagella, pili, and nonpilus adhesins (Drevinek & Mahenthiralingam, 2010;McClean & Callaghan, 2009;Urban, Goldberg, Forstner, & Sajjan, 2005). In this work, a 16HBE14o-cellular monolayer was enzymatically altered using either proteases (trypsin or pronase E) or glycosidases (O-or N-linked).
The cell surface shaving significantly reduces the BCAM2418 transcription in adherent B. cenocepacia K56-2 but does not affect the bacterial adhesion ( Figure 4). Moreover, the bacterial incubation with the resulting supernatant fraction seems to restore the signal- Helicobacter pylori where the mucins glycan-rich domains serve as receptors for infection (Gipson, Spurr-Michaud, Tisdale, & Menon, 2014;Huang et al., 2016;Li et al., 2019;Navabi et al., 2012;Vesterlund, Karp, Salminen, & Ouwehand, 2006). However, until recently, the structure elucidation of the mucin glycan moieties and their bacterial counterparts involved in the binding events remains difficult to determine. Taken together, our present findings suggest that membrane-tethered glycosylated mucins exposed on the surface of lung epithelial cells may represent a group of receptors mediating the primary association of B. cenocepacia K56-2 with host cells. We also hypothesize that the TAA BCAM2418, among other putative adhesive factors, operates to promote the initial contact of the bacteria in the infection process.
Comparative DNA sequence analysis of BCAM2418 genes (structural and promoter regions) of various B. cenocepacia isolates revealed that their lengths greatly vary according to the number of repetitive elements. These data support the hypothesis that the BCAM2418 gene may be subject to phase and antigenic variation during disease.
Phase variation permits the on/off switch in expression, while antigenic variation leads to the alteration in the amino acid sequence of extracellular regions of the protein to prevent recognition by the host immune system (Poole et al., 2013). Previous studies have revealed the alteration in TAAs expression in define infection-linked conditions and environments (Lu et al., 2013;Sheets & St Geme, 2011 and also, that this expansion and contraction of the Cha neck motifs could, in theory, balance the bacteria necessity to colonize with the need to disperse in the host or evade the immune system (Sheets, Grass, Miller, & St Geme, 2008;Sheets & St Geme, 2011). In silico analyses of BCAM2418 reveal the presence of an extensive number of amino acid repeats that vary in size but seems to maintain an SLST signature (Mil-Homens & Fialho, 2011). The overexpression of BCAM2418 might be a mechanism to induce variation in the number of these repeat motifs. This antigenic variation could, in turn, play a similar role to the one reported for Cha TAA. Nevertheless, we could not rule out the alteration in BCAM2418 expression as a result of an on/off switch. Furthermore, the serine and threonine enrichment of these repetitions may be related to putative O-linked glycosylation of BCAM2418 extracellular domains (Iwashkiw, Vozza, Kinsella, & Feldman, 2013;Zhou & Wu, 2009). O-linked glycosylation systems have been described in bacterial systems in the past years, particularly among pathogenic bacteria (Hanuszkiewicz et al., 2014;Vik et al., 2009). The glycosylation of bacterial proteins is usually related to surface and outer membrane proteins. Their sugar enrichment was shown to be linked to adhesive and invasive capacity and protective immunity (Lu, Li, & Shao, 2015;Szymanski & Wren, 2005). Thus, integrated in silico and experimental results could bring new insights regarding BCAM2418 importance as an essential virulence factor and a new key player in the earlier steps of B. cenocepacia infection.
In summary, we first profiled the expression of the 7 virulence-associated trimeric autotransporter adhesin (TAA) genes from B. cenocepacia K56-2 during the early stage of bacteria-host cell interaction. Among those, we found that BCAM2418 gene expression shows an on-off switch and a fine-tuned control in response to a time frame and a particular host cell environment. We also found that physical contact between the bacterium and the host cell is required to trigger the expression of the BCAM2418 TAA with the consequent increase of bacterial adhesion. Finally, we hypothesized that the glycosylated extracellular domain of transmembrane mucins might be cell surface receptors used by B. cenocepacia. Further research using mucin-based technologies will contribute to advance our understanding of the mechanisms underlying the early stages of the bacteria-host crosstalk.

ACK N OWLED G EM ENTS
The We also acknowledge John LiPuma from the University of Michigan, who kindly provided a Burkholderia strain.

CO N FLI C T O F I NTE R E S T
None declared.

E TH I C S S TATEM ENT
None required.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data are provided in full in the results section of this paper.