Role of Fibronectin in the Adhesion of Acinetobacter baumannii to Host Cells

Adhesion to host cells is an initial and important step in Acinetobacter baumannii pathogenesis. However, there is relatively little information on the mechanisms by which A. baumannii binds to and interacts with host cells. Adherence to extracellular matrix proteins, such as fibronectin, affords pathogens with a mechanism to invade epithelial cells. Here, we found that A. baumannii adheres more avidly to immobilized fibronectin than to control protein. Free fibronectin used as a competitor resulted in dose-dependent decreased binding of A. baumannii to fibronectin. Three outer membrane preparations (OMPs) were identified as fibronectin binding proteins (FBPs): OMPA, TonB-dependent copper receptor, and 34 kDa OMP. Moreover, we demonstrated that fibronectin inhibition and neutralization by specific antibody prevented significantly the adhesion of A. baumannii to human lung epithelial cells (A549 cells). Similarly, A. baumannii OMPA neutralization by specific antibody decreased significantly the adhesion of A. baumannii to A549 cells. These data indicate that FBPs are key adhesins that mediate binding of A. baumannii to human lung epithelial cells through interaction with fibronectin on the surface of these host cells.


Introduction
The ability of bacteria to interact with eukaryotic cells, leading to their own internalization, seems to be a critical event in the pathogenesis process of several microorganisms [1]. Invasive bacteria reach a compartment in which they are protected against host clearance mechanisms, can replicate and prepare themselves to gain access to tissues and circulatory system.
Acinetobacter baumannii have long been considered as nosocomial pathogen with low virulence. However, several recent studies have shown that this microorganism is more virulent than expected [2]. Interaction between A. baumannii and the host epithelial cells is important in determining the outcome of infections. Various studies have shown that A. baumannii are able to adhere to and invade human epithelial cells; and induce epithelial cells death [3][4][5][6]. However, there is relatively little information on the mechanisms by which A. baumannii bind to and interact with host cells.
Since the initial reports on A. baumannii adherence and invasion, attempts have been made to elucidate the mechanisms by which A. baumannii promote adherence and invasion in host cells. Acinetobacter sp and A. baumannii used pili, fimbrial-like structures and outer membrane protein A (OmpA) to facilitate its adhesion and invasion in host cells [3,5,7]. However, the host cells surface factors that mediate adherence of A. baumannii are largely uncharacterized. Potential host cell receptors for A. baumannii adhesion can include extracellular matrix (ECM) proteins, such as integrin and fibronectin; which have been used as bridging molecules to achieve the attachment and the invasion of host cells by pathogens like Pseudomonas aeruginosa, Escherichia coli, Streptoccocus pyogenes and Staphylococcus aureus [8][9][10][11].
In support of this, bacterial fibronectin-binding proteins (FBPs) were consistently identified as key players in the process of host cells adhesion to and internalization of different pathogens [12]. FBPs also mediate the attachment of bacteria to implanted material, which becomes coated with host proteins [13], indicating a possible role in infections associated with catheters and prosthetic devices. Most of them are members of large gram negative bacteria, and more thoroughly described in E. coli with most extensive attention [12]. At least 10 different proteins from E. coli bind to fibronectin leading to internalization of E. coli by human host cells including epithelial cells [12].
In the case of A. baumannii, only one study has reported in vitro that A. baumannii adhesive property involves fibronectin [14]. However, the characterization of the role played by this protein is still limited and the FBPs mediating the binding between fibronectin and A. baumannii need to be determined. The present study, therefore, aimed to examinate the binding of fibronectin to A. baumannii and the identification of the FBPs involved in this process.

Interaction of A. baumannii with immobilized fibronectin
We showed that all A. baumannii strains studied here adhered more to fibronectin pre-coated wells than to BSA precoated wells.
The adhesion of A. baumannii to immobilized fibronectin was significantly greater for 77 and ATCC 19606 strains than for 113-16 strain (Fig. 1A). Moreover, soluble fibronectin (from 10 to 1,000 mg/mL) used as a competitor was able to almost completely inhibit all three A. baumannii strains binding with an elective plasmatic fibronectin concentration inhibiting bacterial adherence at 50% (IC 50 ) of <200, 500 and 10 mg/mL, respectively (Fig. 1B). In contrast, incubation of all three A. baumannii strains with BSA (1,000 mg/mL) did not inhibit significantly the binding of A. baumannii ATCC 19606, 77 and 113-16 strains to immobilized fibronectin. From these data, we suggest that A. baumannii has specific ligands for fibronectin.

Identification of A. baumannii FBPs
After growing the 3 strains of A. baumannii for 4 h at 37uC, the enriched OMPs electrophoretic profiles were compared. We observed a significant loss of a 29 kDa OMP in the 113-16 strain when compared with the ATCC 19606 and 77 strains. Western blotting analysis performed following 10% SDS-PAGE of OMPs of each A. baumannii strain (ATCC 19606, 77 and 113-16) and their transfer to nitrocellulose membranes revealed that incubation of this nitrocellulose membrane with fibronectin (10 mg/mL, 1 h) formed three stables complexes with fibronectin for each strain ( Fig. 2A). The three bands have apparent molecular masses of 80, 36 and 32 kDa ( Fig. 2A). To identify these proteins which bind the fibronectin, we excised from SDS-PAGE gel the bands representing fibronectin-binding proteins (FBPs) and subjected them to MS-MS/MS analysis. Data obtained from peptide mass fingerprinting were matched against the NCBI database (http://www.ncbi.nlm. nih.gov). The amino acid sequences identification revealed amino acid identity of 48, 51 and 60% with A. baumannii TonB-dependent copper receptor, A. baumannii OMPA and A. baumannii 34 kDa OMP, respectively (Table 1, Data S1, S2, and S3). For each protein, Mascot probability based mowse score with protein score greater than 84 are significant (p,0.05). To confirm the binding of OMPA to fibronectin, we performed the same experiments with the recombinant OMPA (rOMPA) produced by Escherichia coli. Western blot analysis showed that rOMPA forms a stable complex with fibronectin (Fig. 2B).

Involvement of fibronectin in the adherence of A. baumannii to human lung epithelial cells
To evaluate the involvement of fibronectin in the A. baumannii adherence to A549 cells, we studied the effect of fibronectin inhibitor, RGD, and anti-human fibronectin antibody on A. baumannii ATCC 19606, 77 or 113-16 strain adherence to A549 cells for 2 h. We showed that pretreatment of A549 cells with 0.5 and 1 mg/mL of RGD reduced significantly the adherence of ATCC 19606, 77 and 113-16 strains to A549 cells to 72.0565.39% and 2666%, 61.74613.59% and 45.8861.43%, and 41.4960.11% and 34.7467.68%, respectively (Fig. 3A). Pretreatment of A549 cells with RGD did not affect the A549 cells viability (data not shown). Similarly, pretreatment of A549 cells with 28 mg/mL (1:25) anti-human fibronectin antibody reduced significantly the adherence of ATCC 19606, 77 and 113-16 strains to A549 cells to 65.9760.69%, 57.3360.29%, and 23.3362.93%, respectively. In contrast, pretreatment of A549 cells with unspecific antibody to fibronectin, a mouse IgG, or with rat anti-human E-cadherin antibody, antibody against an irrelevant epithelial surface protein cells, did not reduce significantly the adherence of ATCC 19606, 77 and 113-16 strains to A549 cells (Fig. 3B). The results obtained above with adherence assays were confirmed by fluorescence microscopy experiments, with differential fluorescence labeling used to distinguish between A. baumannii and fibronectin of A549 cells. After 2 h incubation of A549 cells with A. baumannii ATCC 19606, 77 or 113-16 strain, bacterial cells (green fluorescence) were colocalized with the fibronectin (red fluorescence) in the cell membrane, indicating that they were attached to A549 cells (Fig. 3C). Interestingly, the interaction between A. baumannii and fibronectin was more evident after 24 h of A549 cells incubation with A. baumannii strains. We observed that bacterial cells (green fluorescence) were colocalized with the fibronectin (red fluorescence) in the cell membrane and inside the cell (Fig. 3D). Altogether, we found that fibronectin plays an important role in the interaction between A. baumannii and human lung epithelial cells.   (Fig. 4). In contrast, pretreatment of A. baumannii strains with unspecific antibody to fibronectin, a mouse IgG, or with rabbit anti-CarO of A. baumannii antibody antibody against CarO which is an outer membrane protein involved in the carbapenem permeability, did not reduce significantly the adherence of ATCC 19606, 77 and 113-16 strains to A549 cells (Fig. 4). From these data, we found that OMPA which is a FBP mediated the interaction between A. baumannii and human lung epithelial cells.

Discussion
A. baumannii is an opportunistic pathogen that has emerged as a problematic organism, causing severe infections in human hosts. This is compounded by the fact that clinical isolates of this organism are often resistant to multiple types of antimicrobial therapies. To better understand the mechanisms involved in pathogenicity with the long-term goal of finding targets for novel therapies of the future, the work presented in this manuscript focuses on the factor involved in the interaction of A. baumannii to host cells.
A key virulence strategy of Gram-negative pathogens is their ability to interact with proteins of the human serum or ECM. These interactions are associated with bacterial immune evasion or adherence to and invasion into host cells [15][16][17]. In this study, we demonstrated that A. baumannii present specific ligands for fibronectin confirming that fibronectin is a potential host for A. baumannii. The fibronectin binding represents a preferential receptor for different bacteria [18,19]. In addition, we showed that A. baumannii binding to immobilized fibronectin is mediated by the OMPs. The blocking of the immobilized fibronectin using OMPs extracts, abolished attachment of A. baumannii to immobilized fibronectin (data not shown). In agreement with this data, previous studies demonstrate the role played by OMPs in the bacterial binding to fibronectin [20][21][22][23]. Hence, A. baumannii seems to possess the FBPs among their OMPs.
The identification of three FBPs, OMPA, TonB-dependent copper receptor, and 34 kDa OMP, highlights the role of these proteins in the interaction between A. baumannii and host cells. The expression of these three FBPs was not affected by the acquired resistance by A. baumannii. Although there is no experimental evidence up to now, the interaction and the invasion of A. baumannii contribute towards not only systemic infection but also protect the bacteria from antibiotics leading their persistence in the tissue [24,25].
Interestingly, we identified TonB-dependent copper receptor as a FBP. TonB protein works as an energy transducer coupling cytoplasmic membrane proton motive force to active transport of iron-siderophores and vitamin B12 across the outer membrane of E. coli and presumably of other gram-negative bacteria. The  function of TonB is apparently to transduce energy to the outer membrane receptors, perhaps by releasing otherwise tightly bound ligands into the periplasmic space [26,27]. Apart from its capacity to transport molecules, in Pasteurella multocida and Bacteroides fragilis, TonB is able to recognize and bind to fibronectin [22,28]. It also has been suggested that TonB may allow some species to adhere to and colonize different niches in the host [28]. Considering that A. baumannii OMPA is involved in the interaction between A. baumannii and human epithelial cells, in its translocation to mitochondria and cell nucleus, and in its induction of the human epithelial cell death [29][30][31][32], it is likely that OMPA binding to fibronectin play an important role in this phenomena. In agreement with our data, P. multocida have used OMPA to bind to host cells and to interact with fibronectin [33]. In the case of the 34 kDa OMP, we suggest that this FBP is the same protein previously described by different authors as being from 31 to 36 kDa or 33 to 36 kDa [34][35][36]. Similarly to OMPA, it was reported that OMP 33-36 kDa induces the cell death of human cervical epithelial cells (unpublished data). We suggest that this cell death induced by OMP 33-36 kDa would be a consequence of the interaction between OMP 33-36 kDa and host cells. Adherence-inhibition assays conducted in the presence of fibronectin inhibitor or anti-human fibronectin antibody suggests that host fibronectin may be a specific receptor for A. baumannii. In fiboronectin adherence-inhibition assays, the presence of endogenous fibronectin on the surface of human lung epithelial cells may have facilitated the adherence of A. baumannii strains to human lung epithelial cells. Presumably, pretreatment with fibronectin inhibitor resulted in its binding to fibronectin on the surface of A549 cells, thus preventing interaction between FBPs and the endogenous cellular fibronectin on the surface of A549 cells. Inhibition resulting from pretreatment of A549 cells with rabbit anti-human fibronectin was more significant for 113-16 strain than ATCC 19606 and 77 strains suggesting the complexity of adherence for susceptible strains and the involvement of multiple factors including other ECM molecules in the bacterium adherence to host cells [33]. Similar results were reported for P. multocida [28]. Several studies have shown bacterial adherence to host fibronectin [20,37,38]. Bacteria, through their OMPs bind to fibronectin [12]. Our data clearly show that OMPA bind to fibronectin (Fig. 3). Pretreatment of A. baumannii with mouse anti-A. baumannii OMPA prior incubation with A549 cells strongly reduced the adherence of A. baumannii to A549 cells. Similar results have been obtained with P. multocida OMPA [33].
In summary, all data reported here indicate that A. baumannii binding to fibronectin was mediated by the OMPs. The OMPs were identified as FBPs and were involved in the interaction between A. baumannii and fibronectin. Our results encourage us to continue our studies in an attempt to understand the involvement of A. baumannii adherence to host cells in the pathogenesis of this pathogen.

Human cell culture and infection
Type II pneumocyte cell line A549 derived from a human lung carcinoma (LGC Standards, United Kingdom) were grown in DMEM medium supplemented with 10% heat-inactivated foetal bovine serum (FBS), vancomycin (50 mg/mL), gentamicin (20 mg/mL), amphoterecin B (0.25 mg/mL) (Invitrogen, Spain) and 1% HEPES (Gibco, Spain) in a humidified incubator, 5% CO 2 at 37uC. A549 cells were routinely passaged every 3-4 days. The cells were seeded 24 h in 24 well plates prior to infection with A. baumannii strains at a multiplicity of infection (MOI) of 100. Immediately before infection, A549 cells were washed three times with prewarmed phosphate-buffered saline (PBS) and further incubated in DMEM without FBS and antibiotics.

Coating fibronectin on wells
Coating fibronectin on wells were performed in sterile 96 well plates as described previously [23]. Briefly, wells were coated overnight at 4uC with 1.25 mg of plasmatic fibronectin in PBS (125 mL of fibronectin at 10 mg/mL per well) or with 20 mg/mL of BSA. Then wells were washed four times with 125 mL of 1% (w/v) BSA in PBS, and blocked for 1 h at room temperature with 125 mL of 1% BSA in PBS. Just before adding bacteria, wells were washed six times with PBS.

Outer membrane preparations and rOMPA production
Outer membrane fraction from A. baumannii ATCC 19606, 77 or 113-16 strain were isolated as described previously [39]. Briefly, 1 L of MHB was inoculated with 5 mL of an overnight culture. Bacterial cells were pelleted by centrifugation at 5,0006 g for 10 min, resuspended in 10 mL of 10 mM phosphate buffer, pH 7.2, and lysed on ice by sonication for a total of 5-7 min in 30 s intervals. Unlysed cells were pelleted by centrifugation at 5,0006 g for 5 min, and the supernatant was centrifuged at 20,0006 g for 45 min at 4uC to pellet cell envelopes. Inner membranes from the resulting pellet were solubilized for 30 min at room temperature with 5 mL of 2% N-laurylsarcosinate in 10 mM phosphate buffer, pH 7.2. After solubilization, the insoluble outer membrane fraction was pelleted by centrifugation at 20,0006 g for 45 min at 4uC. The outer membrane fraction was washed by resuspension in 2 mL 62.5 mM Tris-HCl, pH 6.8 and centrifuged at 20,0006 g for 45 min at 4uC. The resulting pellet was stored at 280uC until use. rOMPA was produced from E. coli as described previously by our group [40].

Fibronectin-binding assays
Fibronectin-binding assays were performed as described previously [23], with some modifications. Briefly, A. baumannii ATCC 19606, 77 or 113-16 strain grown overnight at 37uC in MHB were resuspended in PBS and collected by centrifugation at 5,0006 g for 10 min. Bacteria were washed twice in sterile PBS and resuspended in the same sterile buffer. Two assays were performed: i) 50 mL of bacterial suspension were mixed with 50 mL of PBS, added to BSA or fibronectin coated wells and incubated 3 h at room temperature, and ii) 50 mL of bacterial suspension were mixed with increasing amounts of human plasmatic fibronectin (0, 10, 100 and 1,000 mg/mL), added to fibronectin coated wells and incubated 3 h at room temperature for bacterial adsorption. After rinsing thrice with sterile PBS, 50 mL of bacterial suspension were added for 3 h at room temperature for bacterial adsorption. Non-adhered bacteria were discarded and wells were washed six times with sterile PBS to remove unbound bacteria. Adherent bacteria were then collected with 125 mL of sterile PBS containing 0.5% Triton X-100. Diluted lysates were plated onto sheep blood agar (Becton Dickinson Microbiology Systems, USA) and incubated at 37uC for 24 h for enumeration of developed colonies and then the determination of the number of bacteria that attached to fibronectin. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting assay After the determination of the OMPs and rOMPA amounts by using the bicinchoninic assay assay (Thermo Scientific, Spain), three or six mg of proteins were mixed with an equal volume of 26 Laemmli buffer, denaturated by heating the mixture for 5 min at 95uC, and then resolved by 10% SDS-PAGE. The gels were stained with SimplyBlue TM SafeStain following the manufacturer's instructions (Invitrogen, Spain).

Western blotting
To investigate the possible involvement of OMPs and OMPA in the A. baumannii interactions with fibronectin, OMPs extracts and rOMPA were used for western blotting. Western blotting analysis was performed following the SDS-PAGE as indicated above. The separated proteins were transferred using nitrocellulose membranes (Amersham Bioscience, Spain), and the membranes were blocked for 1 h with PBS and 0.1% (v/v) Tween 20 (PBST buffer) containing 5% (w/v) milk. The nitrocellulose membranes were then incubated with human fibronectin (10 mg/mL) for 1 h, washed twice with PBST buffer, and incubated overnight at 4uC with primary antibody: rabbit anti-human fibronectin (1:2,500 dilution) diluted in PBST buffer containing 5% milk. After washing with PBST buffer, the membranes were incubated for 1 h at room temperature with a horseradish peroxidase-conjugated goat anti-rabbit IgG antibody (Amersham Bioscience, Spain) (dilution 1:10,000) diluted in PBST buffer containing 5% milk. Subsequently, immunoreactive proteins were visualized using the enhanced chemiluminescence protocol (Super ECL, (Amersham Bioscience, Spain).
For peptide mass fingerprinting, SimplyBlue TM SafeStainstained bands representing FnBPs were excised from SDS-PAGE gel and sent to the Centro de Investigación Príncipe Felipe de Valencia for MALDI-TOF-TOF (MS-MS/MS) analysis. Data obtained from peptide mass fingerprinting were matched against the NCBI database (http://www.ncbi.nlm.nih.gov) using the Mascot program.
Immunofluorescence A549 cells plated on coverslips in 24 well plates for 24 h and infected with A. baumannii ATCC 19606, 77 or 113-16 strain at 37uC for 2 and 24 h were removed and washed five times with cold PBS. A549 cells were fixed in methanol for 8 min at 220uC, permeabilized with 0.5% Triton X-100 and blocked with 20% porc serum in PBS. Primary antibodies: mouse anti-A. baumannii OMPs [41] and rabbit anti-human fibronectin (Sigma, Spain) were used at dilution of 1:50 and 1:400 in PBS containing 1% BSA for 2 h, respectively. After washing with PBS, the coverslips were incubated with the secondary antibodies: Alexa488-conjugated goat anti-mouse IgG, and Alexa594-conjugated goat anti-rabbit IgG (Invitrogen, Spain) at dilution of 1:100 and 1:800 in PBS containing 1% BSA for 1 h, respectively. The fixed coverslips were incubated for 10 min at room temperature with DAPI (Appli-chem, Germany) (0.5 mg/mL), washed with PBS, mounted in fluorescence mounting medium (Dako Cytomation, Spain), and visualized using a Leica fluorescence microscope (DM-6000; Leica Microsystems Wetzlar GmbH, Germany).

Adhesion assay
i) A549 cells were pretreated with RGD (0.5 and 1 mg/mL), a fibronectin inhibitor, for 30 min, rabbit anti-human fibronectin (28 mg/mL) for 60 min, mouse IgG (1:25) for 60 min, or with rat anti-human E-cadherin (1:25) for 60 min and infected with 10 8 cfu/mL of A. baumannii ATCC 19606, 77 or 113-16 strain for 2 h with 5% CO 2 at 37uC. ii) A549 cells were infected for 2 h in 5% CO 2 at 37uC with 10 8 cfu/mL of A. baumannii ATCC 19606, 77 or 113-16 strain previously incubated for 1 h at room temperature with mouse anti-A. baumannii OMPA (diluted in PBS at 1:1,000 and 1:250), mouse IgG (1:25) for 60 min, or with rabbit anti-A. baumannii CarO (1:25) kindly gift by AM Viale [42]. Subsequently, infected A549 cells were washed five times with prewarmed PBS and lysed with 0.5% Triton X-100. Diluted lysates were plated onto blood agar (Blood-Agar Columbia, Becton Dickinson Microbiology Systems, USA) and incubated at 37uC for 24 h for enumeration of developed colonies and then the determination of the number of bacteria that attached to A549 cells.

Statistical analysis
Group data are presented as mean 6 SEM. Student t-test was used to determine differences between means. Difference was considered significant at P,0.05. SPSS (version 15.0) statistical package was used (SPSS Inc., Chicago, IL).

Supporting Information
Data S1 TonB-dependent copper receptor sequence information.