The Fibronectin-Binding Protein EfbA Contributes to Pathogenesis and Protects against Infective Endocarditis Caused by Enterococcus faecalis Kavindra V. Singh,a,b Sabina Leanti La Rosa,a* Sudha R. Somarajan,a Jung Hyeob Roh,a Barbara E. Murraya,b,c Department of Internal Medicine, Division of Infectious Diseases, University of Texas Health Science Center at Houston, Houston, Texas, USAa; Center for the Study of Emerging and Re-emerging Pathogens, University of Texas Health Science Center at Houston, Houston, Texas, USAb; Department of Microbiology and Molecular Genetics, University of Texas Health Science Center at Houston, Houston, Texas, USAc

EfbA is a PavA-like fibronectin adhesin of Enterococcus faecalis previously shown to be important in experimental urinary tract infection. Here, we expressed and purified the E. faecalis OG1RF EfbA and confirmed that this protein binds with high affinity to immobilized fibronectin, collagen I, and collagen V. We constructed an efbA deletion mutant and demonstrated that its virulence was significantly attenuated (P < 0.0006) versus the wild type in a mixed inoculum rat endocarditis model. Furthermore, efbA deletion resulted in diminished ability to bind fibronectin (P < 0.0001) and reduced biofilm (P < 0.001). Reintroduction of efbA into the original chromosomal location restored virulence, adherence to fibronectin, and biofilm formation to wild-type levels. Finally, vaccination of rats with purified recombinant EfbA protein protected against OG1RF endocarditis (P ⴝ 0.008 versus control). Taken together, our results demonstrate that EfbA is an important factor involved in E. faecalis endocarditis and that rEfbA immunization is effective in preventing such infection, likely by interfering with bacterial adherence.

E

nterococci are among the first microbial colonizers of the infant gastrointestinal tract and, in healthy adults, they are part of the commensal intestinal microbiota (1). However, since the middle to late 1970s, enterococci have emerged as opportunistic pathogens and major causes of health care-associated infections, most notably urinary tract infections (UTIs) and bacteremia, second only to Staphylococcus spp. in the hospital setting (2–5). Among cases of enterococcal endocarditis, Enterococcus faecalis has been recognized as the most abundant species, accounting for more than 90% of isolates from this infection (2, 3, 6). Treatment of enterococcal endocarditis has been clinically challenging due to the emergence of strains resistant to commonly used antibiotics, i.e., aminoglycosides, although more recent use of ampicillin in combination with ceftriaxone seems to have alleviated this concern (7). The associated mortality remains high, with rates between 9 and 15% in Europe and the United States, respectively (8, 9). Adherence to and colonization of host tissue components, such as platelets and the extracellular matrix (ECM), are abilities that facilitate the early steps toward the development of infective endocarditis. To date, a variety of factors involved in the pathogenesis of experimental E. faecalis aortic valve infection have been identified; these include MSCRAMM (microbial surface components recognizing adhesive matrix molecules) proteins such as the adhesin to collagen, Ace, and the endocarditis and biofilm-associated pili, Ebp (10–12). Fibronectin is a glycoprotein consisting of two protein chains of approximately 250 kDa each, covalently linked by disulfide bonds, that is present in soluble forms in blood plasma and other body fluids. An insoluble, fibrillar form of fibronectin is present in the ECM and is known to support bacterial adherence (13). A recent study from Torelli et al. reported the presence of a fibronectin-binding protein in E. faecalis JH2-2, designated enterococcal fibronectin-binding protein A (EfbA) (14). Similar to the pneumococcal adherence and virulence factor A (PavA) of Streptococ-

December 2015 Volume 83 Number 12

cus pneumoniae, EfbA displays an unconventional structure for adhesins in that it lacks an N-terminal signal sequence for targeting the protein to the Sec system and a C-terminal LPXTG cell wall anchorage motif (14–16). Homologous proteins lacking these motifs include Fbp54 of Streptococcus pyogenes (17), FbpA of Streptococcus gordonii (18), the Streptococcus mutants SmFnB (19), and the group B streptococcus SfbA (20), collectively referred to as PavA-like proteins, several of which were shown to be involved in bacterial adherence to fibronectin and/or to mediate virulence in experimental models (15–21). E. faecalis JH2-2 EfbA consists of an N-terminal predicted domain involved in fibronectin binding (FbpA; amino acids 4 to 429), followed by a domain of unknown function (Duf814; amino acids 448 to 533) often observed in association with FbpA. EfbA was shown to be a serum-inducible protein displayed on the outer surface of E. faecalis JH2-2 and the derived recombinant protein (rEfbA) bound to immobilized human fibronectin. An efbA-deletion mutant showed impairment in adherence to fibronectin and was attenuated in a murine ascending UTI model (14). However,

Received 6 July 2015 Returned for modification 24 July 2015 Accepted 1 September 2015 Accepted manuscript posted online 8 September 2015 Citation Singh KV, La Rosa SL, Somarajan SR, Roh JH, Murray BE. 2015. The fibronectin-binding protein EfbA contributes to pathogenesis and protects against infective endocarditis caused by Enterococcus faecalis. Infect Immun 83:4487–4494. doi:10.1128/IAI.00884-15. Editor: A. Camilli Address correspondence to Barbara E. Murray, [email protected]. * Present address: Sabina Leanti La Rosa, Department of Chemistry, Biotechnology and Food Science, Norwegian University of Life Sciences, Aas, Norway. Supplemental material for this article may be found at http://dx.doi.org/10.1128 /IAI.00884-15. Copyright © 2015, American Society for Microbiology. All Rights Reserved.

Infection and Immunity

iai.asm.org

4487

Singh et al.

TABLE 1 Bacterial strains and plasmids used in this study Strain or plasmid

Relevant characteristicsa

Source or reference(s)

Strains E. coli DH5␣ EC1000 BL21(DE3)

Strain for routine cloning Host strain; provides RepA for replication of pHOU1 Strain for overexpression of recombinant proteins

Stratagene 46 Life Technologies

Laboratory strain; Fusr Rifr OG1RF⌬efbA, an efbA in-frame deletion mutant of OG1RF OG1RF⌬efbA::efbA*, reconstituted efbA* in the TX5707 genetic background; carries a silent nucleotide mutation in the Leu79 codon for differentiation from OG1RF Conjugative donor; carries repA in trans for pHOU1 replication

25, 47 This study This study

Plasmid for markerless allelic exchange in E. faecalis; carries the counterselectable pheS* gene encoding a phenylalanine tRNA synthetase; Genr Construct for efbA deletion; contains a 1,029-bp BamHI/PstII fragment encompassing the flanking region of the efbA gene cloned into pHOU1 Construct for reconstitution of efbA; contains a 2,700-bp BamHI/PstII fragment cloned into pHOU1 Expression vector for N-terminal histidine-tagged fusion proteins in E. coli; Ampr pET19b derivative carrying a 1,713-bp NdeI/BamHI fragment containing the efbA gene

22

E. faecalis OG1RF TX5707 TX5708 CK111 Plasmids pHOU1 pTX5707 pTX5708 pET19b pET19b-EfbA a

23

This study This study Novagen This study

Abbreviations: Ampr, ampicillin resistant; Fusr, fusidic acid resistant; Genr, gentamicin resistant; Rifr, rifampin resistant.

the role of EfbA in the pathogenesis of E. faecalis infective endocarditis has not previously been reported. In the present study, we generated a nonpolar efbA deletion mutant of E. faecalis OG1RF, restored the gene in its original chromosomal location, and evaluated the ability of these derivatives to infect damaged heart valves in a rat model, adhere to immobilized fibronectin, and form biofilm. Finally, we provide evidence that EfbA is a protective antigen that could be used as a vaccine to combat E. faecalis experimental endocarditis. MATERIALS AND METHODS Bacterial strains and growth conditions. The bacterial strains and plasmids used in the present study are described in Table 1. Unless otherwise specified, E. faecalis were cultured at 37°C in brain heart infusion (BHI; Difco Laboratories) broth and agar or in BHI supplemented with 40% horse serum (BHIS) (Sigma, St. Louis, MO). Escherichia coli strains were grown at 37°C in Luria-Bertani broth and agar (Difco Laboratories). The concentrations of the antibiotics used for selection were as follows: for E. faecalis, erythromycin (10 ␮g/ml), fusidic acid (25 ␮g/ml), and gentamicin (200 ␮g/ml) and for E. coli, ampicillin (100 ␮g/ml) and gentamicin (25 ␮g/ml). Enterococcosel agar (EA; Difco Laboratories) supplemented with rifampin at 100 ␮g/ml was used to recover bacteria from tissue homogenates. Construction of isogenic mutants. E. faecalis OG1RF isogenic mutants were constructed using pHOU1, a vector that contains a mutated pheS* allele conferring susceptibility to p-chloro-phenylalanine (p-ClPhe) (22). To generate an efbA deletion mutant, two fragments encompassing the upstream (507 bp) and downstream (522 bp) DNA region flanking efbA were amplified by PCR using the primer pair Up_efbF/ Up_efbR and Down_efbF/Down_efbR (see Table S1 in the supplemental material), respectively. The fragments were subsequently fused together by overlap extension PCR and the product cloned into BamHI/PstI-digested pHOU1, giving pTX5707. After transformation into electrocompetent E. faecalis CK111 using standard procedures (23, 24), pTX5707 was transferred to E. faecalis OG1RF by filter mating. Single-crossover integrants were selected on BHI plates containing gentamicin and fusidic acid and then replated onto MM9YEG medium supplemented with 7 mM p-Cl-Phe to select for vector excision. Resulting colonies were screened for

4488

iai.asm.org

the presence or absence of efbA by PCR. An efbA deletion strain was designated TX5707. To restore efbA in the native chromosomal site, a 2.7-kb fragment containing the efbA gene with its native promoter was PCR amplified with the primers Up_efbF and Down_efbR (see Table S1 in the supplemental material). To distinguish between the parent and the complemented strain, a silent mutation (TTG) was introduced into the leucine codon TTA corresponding to amino acid position 79 using the primers SmutF and SmutR (see Table S1 in the supplemental material). Leucine is the most abundant amino acid in EfbA (60 of 570 amino acids), and codon usage was as follows: 25 TTA, 13 CTT, 9 CTA, 6 TTG, 5 CTC, and 2 CTG. After digestion with BamHI and PstII, the PCR product was cloned into pHOU1, giving pTX5708. The construct was then electroporated into E. faecalis CK111, transferred to TX5707 via conjugation and processed as described above. The resulting efbA* reconstituted strain was named TX5708. Sequencing of PCR products amplified using primers EfbOutF and EfbOutR and pulsed-field gel electrophoresis (25) were performed to confirm the deletion and restoration of efbA in TX5707 and TX5708. In addition, excision of the pHOU1 derivatives was confirmed by absence of growth on BHI-gentamicin agar plates. Growth characteristics. Growth of the OG1RF derivatives in BHIS or tryptic soy broth supplemented with 0.25% glucose (TSBG) was monitored as described elsewhere (11). Heterologous expression, purification of recombinant EfbA and binding to ECM proteins. N-terminal 6⫻His-tagged recombinant EfbA was produced in Escherichia coli BL21(DE3) harboring the plasmid pET19b-EfbA. The construct was generated by amplifying the full-length efbA gene from OG1RF using the primer pair pET_efbF and pET_efbR after introduction into NdeI/BamHI-digested pET19b (see Table S1 in the supplemental material). BL21(DE3) cultures were grown at 37°C to the exponential phase, and protein overexpression was induced with 0.5 mM IPTG (isopropyl-␤-D-thiogalactopyranoside). Purification of rEfbA by affinity chromatography was carried out using His GraviTrap columns (GE Healthcare) according to the protocol of the manufacturer. The columns were washed with washing buffer (20 mM sodium phosphate [pH 7.4], 500 mM NaCl, and 45 mM imidazole), and then the protein was eluted in elution buffer (20 mM sodium phosphate [pH 8.0], 500 mM NaCl, and 500 mM imidazole). The protein was desalted in 50 mM TrisHCl buffer (pH 8.0) and 5% glycerol using PD-10 desalting columns (GE

Infection and Immunity

December 2015 Volume 83 Number 12

EfbA Is Important in E. faecalis Endocarditis

Healthcare), and the concentration was determined by the BCA method (Pierce); rEfbA was stored at ⫺70°C until used. Concentration-dependent binding of rEfbA to fibronectin and adherence to various ECM proteins was assessed by enzyme-linked immunosorbent assay (ELISA) as previously described (26). Briefly, Immulon 2B microwell plates (medium binding; Thermo Scientific, Woburn, MA) were coated overnight at 4°C with ECM proteins at 1 ␮g/well. The surfaces of the wells were blocked with 2% bovine serum albumin (BSA) at room temperature for 1 h and washed with phosphate-buffered saline (PBS [pH 7.4]), followed by the addition of different concentrations of rEfbA. After incubation for 2 h at room temperature, the plates were rinsed thrice with PBS-T (PBS with 0.05% Tween 20), and bound protein was detected by incubation with anti-His6 monoclonal antibodies (GE Healthcare) and alkaline phosphatase-conjugated anti-mouse IgG antibodies (Jackson ImmunoResearch Laboratories); signal was detected using p-nitrophenyl phosphate (Sigma). The absorbance at 405 nm was measured with a microplate reader (Thermo Scientific). Binding of E. faecalis to immobilized fibronectin. Overnight cultures in BHIS of E. faecalis were collected by centrifugation and then adjusted to an optical density at 600 nm (OD600) of 1.0 in PBS. A portion (100 ␮l) of the bacterial suspension was inoculated into Immulon 2B microwell plates coated overnight with 20 ␮g of fibronectin or BSA (control)/ml, as previously described (27). After 2 h of incubation at room temperature, the plates were washed with PBS to remove unbound bacteria, and adherent cells were fixed with Bouin’s fixation solution (Sigma-Aldrich, St. Louis, MO). Each well was then rinsed with PBS and stained with 1% (wt/vol) crystal violet for 30 min at room temperature. Adherent cells were dissolved in an ethanolacetone solution (80 and 20%, respectively), and the absorbance (OD570) was measured using a microplate reader (Thermo Scientific). To calculate the adherence of E. faecalis cells to fibronectin, the OG1RF value was set at 100% and binding of TX5707 and TX5708 was expressed as percentage of binding relative to OG1RF. Biofilm formation assay. The biofilm assay was performed as described previously (11). E. faecalis strains were cultured overnight in TSBG at 37°C and diluted to an OD600 of 0.05 in TSBG. Portions (200 ␮l) of the bacterial suspension were inoculated into 96-well polystyrene microtiter plates (BD Biosciences), followed by incubation for 24 h at 37°C. Biofilm density was determined by crystal violet staining (11). Rat infective endocarditis model. Animal experimental procedures were approved by the Animal Welfare Committee, University of Texas Health Science Center—Houston, and performed in accordance with institutional guidelines. A mixed infection competition assay was performed as we previously published (10). Briefly, groups of 7 to 11 catheterized Sprague-Dawley rats, weighing 200 to 300 g, were infected with a BHIS-grown mixture of TX5707 and either OG1RF or TX5708 (mixed in a 1:1 ratio, estimated by the OD600 and verified by CFU counts) via the tail vein, 24 h after catheter placement. To determine the CFU present in vegetations, animals were euthanized at 48 h postinfection, hearts were aseptically removed, and the aortic valves containing vegetations were excised, weighed, and homogenized in saline solution. Serial dilutions were plated onto EA-plus-rifampin (100 ␮g/ml) plates. The ratio of TX5707 and OG1RF or TX5708 was determined by hybridizing colony lysates under high-stringency conditions with intragenic probes of efbA and ace to confirm E. faecalis (10). Antibody titers and in vivo protection against E. faecalis OG1RF infection in the rat endocarditis model. The immunization protocol was adapted from our previous publication (10). For active immunization, a group of 10 Sprague-Dawley rats, weighing 200 to 300 g, received a dose of 100 ␮g/rat of rEfbA suspended in Freund complete adjuvant (FCA) at week 1, followed by another dose of 100 ␮g of rEfbA/rat in Freund incomplete adjuvant (FICA) at 3 and 5 weeks. The control group (n ⫽ 36) received FCA-FICA sham immunization at the same frequency. At week 7, rats were implanted with a catheter across the aortic valves. After 24 h, the animals were inoculated via the tail vein using BHIS-grown E. faecalis

December 2015 Volume 83 Number 12

FIG 1 Binding of rEfbA to immobilized fibronectin detected by ELISA. Im-

mulon 2B microwell plates coated with fibronectin at 1 ␮g/well were incubated with increasing concentrations of rEfbA. Binding of rEfbA to BSA or fibronectin was detected using anti-His6-tagged antibodies. The data points represent the means ⫾ the standard deviations from three independent experiments each with eight technical replicates.

OG1RF (2.7 ⫻ 108 CFU for the rEfbA immunized rats and 2 ⫻ 108 CFU for the sham-immunized rats) and sacrificed 48 h later, and the aortic valves processed as described above for CFU. At 24 h prior to infection with E. faecalis OG1RF, the presence of anti-EfbA antibody titers in the sera from vaccinated and sham-immunized rats was assessed by ELISA according to a previously described protocol (28). Statistical analysis. Statistical comparisons were performed by paired Student t test and by Mann-Whitney test using GraphPad Prism (v4.00 for Windows; GraphPad Software, San Diego, CA). The Fisher exact test was used to compare the numbers of infected versus noninfected animals in the active immunization experiments. Differences were considered significant at a P ⱕ 0.05.

RESULTS

Characterization of EfbA in E. faecalis OG1RF. A previous study showed that EfbA is a PavA-like fibronectin-binding protein localized to the outer cell surface of E. faecalis strain JH2-2 and functions as a bacterial adhesin to immobilized fibronectin (14). E. faecalis OG1RF encodes a 570-amino-acid protein that is 99% identical to JH2-2 EfbA, with five nonconserved residues; three changes (at positions 231, 259, and 361) are in the predicted Nterminal domain and another (position 453) is in the Duf814 domain (see Fig. S1 in the supplemental material). To confirm the function of the predicted homologue protein in E. faecalis OG1RF, the full-length OG1RF EfbA was expressed as histidine-tagged fusion protein in E. coli, purified using Ni-NTA affinity chromatography (see Fig. S2 in the supplemental material), and tested for the ability to bind immobilized human fibronectin using a microtiter plate-based assay. rEfbA showed concentration-dependent binding to fibronectin and low affinity to BSA (Fig. 1). The binding was saturable at high concentrations of rEfbA with a calculated Kd value of 0.58 pM (Fig. 1). As seen in Fig. 2, we detected high affinity of rEfbA to collagen I and collagen V (Fig. 2). Little binding to laminin, mucin, heparin, hyaluronic acid, collagen IV, and fibrinogen was observed (Fig. 2). Importance of efbA in a mixed infection endocarditis model. To test whether efbA contributes to E. faecalis ability to infect heart valves, we constructed a nonpolar efbA deletion mutant (TX5707) and a derivative strain in which the efbA gene was restored in its

Infection and Immunity

iai.asm.org

4489

Singh et al.

FIG 2 rEfbA binding to immobilized ECM components detected by ELISA. rEfbA binding to ECM (1 ␮g/well) was measured by ELISA. The data represent the means ⫾ the standard deviations from three independent experiments. BSA was used as a control.

original chromosomal location (TX5708). The growth kinetics of the deletion mutant and the restored efbA strain were similar to that of the parental strain OG1RF in broth cultures (BHIS and TSBG) (see Fig. S3 in the supplemental material). As shown in Fig. 3A, the mean percentages of total CFU from rats inoculated with an equal suspension of BHIS-grown OG1RF and TX5707 were 75% versus 25% (see Fig. S4 in the supplemental material), respectively, at 48 h postinfection, therefore showing a clear advantage of the parental strain compared to the efbA deletion mutant (P ⬍ 0.0006). In a similar experiment, we used a bacterial suspension that contained 59% TX5707 and 41% TX5708 and, again, significantly higher numbers of the EfbA-expressing strain (TX5708) were recovered than the EfbA-deficient mutant (Fig. 3B and see Fig. S5 in the supplemental material) (P ⬍ 0.0086), confirming that efbA contributes to this endovascular infection.

Role of EfbA in fibrinogen binding and biofilm formation. In order to further elucidate whether the effect of efbA deletion on E. faecalis pathogenicity might be via its ability to bind to fibronectin, we examined whole-cell binding of wild-type OG1RF and the efbA deletion mutant to immobilized fibronectin. As shown in Fig. 4A, deletion of efbA caused a 44% decrease in fibronectin adherence compared to the wild type (P ⬍ 0.0001). As expected, the efbA reconstituted strain showed fibronectin binding similar to that of OG1RF. Next, we examined whether the attenuated virulence phenotype of the efbA deletion mutant correlated with biofilm formation. As seen in Fig. 4B, the density of the biofilm produced by TX5707 showed a modest but significant reduction (15%) compared to the wild type (P ⬍ 0.0001). Restoration of the biofilm formation capacity was observed in the in situ efbA-reconstituted strain. In vivo protection against E. faecalis OG1RF rat endocarditis. The data described above demonstrated that EfbA play an important role in the development of infective endocarditis by E. faecalis. We therefore reasoned that vaccination with rEfbA might be protective against E. faecalis infection in the rat endocarditis model. After active immunization with three doses of 100 ␮g of rEfbA in FCA-FICA, we found that sera from vaccinated rats (n ⫽ 8) contained high levels of anti-EfbA titers (1:12,800), whereas the sera from sham-immunized animals (n ⫽ 8) were not reactive to EfbA (Fig. 5A), therefore demonstrating active production of anti-EfbA antibodies. Upon challenge with BHIS-grown OG1RF, 95% of the 36 sham-immunized animals developed E. faecalis endocarditis with a mean of 4.8 log10 CFU recovered from vegetations (Fig. 5B). Immunization with rEfbA was effective in reducing the number of infected animals to 60% (Fisher exact test; P ⫽ 0.015 versus sham immunized) and the bacterial burden recovered from vegetation to a mean of 2.1 log10 CFU (P ⫽ 0.008 versus sham immunized). These data therefore demonstrated that active immunization of rats with rEfbA protects from E. faecalis colonization of aortic valves.

FIG 3 Rat model of infective endocarditis using a mixed inoculum. The percentages of OG1RF and the efbA deletion mutant (TX5707; OG1RF⌬efbA) (A) and of efbA* chromosomally reconstituted (TX5708; OG1RF⌬efbA::efbA*) and TX5707 (B) present in the inoculum mix and recovered from aortic valve vegetations at 48 h postinfection were determined. (A) The CFU of the wild-type OG1RF (䊐) and mutant strain TX5707 (Œ) in the inoculum were 2.83 ⫻ 109 CFU (50%) and 2.88 ⫻ 109 CFU (50%), respectively. (B) Infection of rats with the chromosomally reconstituted (TX5708) and efbA deletion mutant (TX5707) bacteria. Rats were inoculated with 7.05 ⫻ 108 CFU (41%) of TX5708 (✳) and 9.75 ⫻ 108 CFU (59%) of TX5707 (Œ) cells, respectively. Horizontal lines indicate the means of percentages of OG1RF versus TX5707 (A) or of TX5708 versus TX5707 (B) in the inoculum and in vegetations, respectively. The data were analyzed using the Mann-Whitney test.

4490

iai.asm.org

Infection and Immunity

December 2015 Volume 83 Number 12

EfbA Is Important in E. faecalis Endocarditis

FIG 4 Effect of efbA deletion on E. faecalis OG1RF adherence to fibronectin and biofilm formation. (A) Whole-cell binding of E. faecalis OG1RF, TX5707, and

TX5708 to immobilized fibronectin. BHIS-grown cells were adjusted to an OD600 of 1 before inoculation in wells coated with 20 ␮g of fibronectin. Bars represent the adherence percentage compared to the wild-type strain OG1RF (defined as 100%). The data are presented as means ⫾ the standard deviations of at least three independent experiments. (B) Biofilm formation with overnight cultures of OG1RF, the deficient EfbA mutant strain TX5707, and the reconstituted derivative strain TX5708. Horizontal lines show the means of three independent experiments, each performed with eight technical replicates. Statistical analysis was performed using a Student t test.

DISCUSSION

Enterococci, particularly E. faecalis and, to a lesser extent, E. faecium, are the third most common pathogens isolated from native and prosthetic valve endocarditis worldwide (3). Enterococcal infection of the cardiac endothelium is believed to be a result of bacteremia with subsequent bacterial colonization of ECM components of the damaged heart tissues with subsequent formation of infected thrombotic vegetations (consisting of microbes, platelets, fibrin, and host cells) (29, 30). Fibronectin is an important component of the extracellular matrix (ECM) of the cardiac endothelium and is exposed when this tissue is injured. In addition, fibronectin binds to fibrin and platelets, thereby contributing to the thrombogenicity of surfaces

(31). Adherence to fibronectin by Gram-positive bacteria has been shown to be an important virulence factor that facilitates host colonization and supports disease progression (32–34). This ability is mediated by fibronectin-binding MSCRAMMs exhibiting typical structural features of cell wall anchored proteins, as well as anchorless but surface-located adhesins and invasins (35, 36). A previous report characterized EfbA, a PavA-like fibronectinbinding protein of E. faecalis JH2-2, and demonstrated that it plays a major role in in vitro adherence to ECM proteins and promotes urinary tract infections in a murine model (14). However, to date, no studies have demonstrated the involvement of E. faecalis fibronectin-binding proteins in the pathogenesis of infective endocarditis. Thus, the above observations led us to undertake the

FIG 5 Experimental endocarditis model in rEfbA-immunized rats. Rats were immunized with a dose of 100 ␮g of rEfbA, followed by two booster immunizations at weeks 3 and 5. (A) Antibody titers detected by ELISA in eight EfbA-immunized rats and in the pooled sera of eight sham-immunized animals prior to infection with E. faecalis OG1RF. Each line depicts the anti-EfbA titers for each antibody dilution tested. (B) Two weeks after the last vaccine administration, a catheter was placed across the aortic valves of the rats and left for 24 h prior to challenge with 2.7 ⫻ 108 CFU of E. faecalis OG1RF. Catheterized sham-immunized rats were inoculated with 2 ⫻ 108 CFU of E. faecalis OG1RF. Black triangles and circles represent OG1RF cells recovered from the vegetations of EfbA-immunized and sham-immunized rats, respectively. Horizontal lines indicate the geometric means. A value of 0 depicts an uninfected animal. The means of log10 CFU from vegetations were compared by unpaired t test.

December 2015 Volume 83 Number 12

Infection and Immunity

iai.asm.org

4491

Singh et al.

present study and investigate whether EfbA contributes to E. faecalis OG1RF endocarditis in rats with catheter-induced aortic vegetations. First, we confirmed that EfbA in OG1RF is a fibronectin-binding protein by showing that it adhered to immobilized human fibronectin in a concentration-dependent manner. EfbA supported adherence to other host ECM proteins such as collagen I and collagen V and modest levels of binding to laminin, mucin, heparin, hyaluronic acid, collagen IV, and fibrinogen. To assess the role of EfbA in the pathogenicity of experimental endocarditis, we used a mixed inoculum model in which OG1RF and TX5707 (OG1RF⌬efbA) were inoculated intravenously at a ratio of approximately 1:1. Deletion of efbA resulted in significant attenuation in the ability of the isogenic mutant strain to outcompete the parental OG1RF. Restoration of efbA in TX5707 resulted in increased colonization of the heart valves back to OG1RF levels, therefore indicating that efbA is important in this endovascular infection. To the best of our knowledge, this is the first demonstration that a fibronectin-binding protein contributes to the pathogenesis of E. faecalis infective endocarditis. Although another role, including an intracellular function, for efbA or EfbA cannot be excluded, we suggest that the difference in virulence of TX5707 is likely due to a reduced ability of the mutant to adhere to fibronectin at the site of damage, since the efbA deletion had no effect on growth but resulted in lower adherence to immobilized fibronectin compared to the wild-type cells, and this ability was restored in the in situ complemented strain TX5708. Together with attachment to and colonization of host surfaces, the capacity to form biofilm is considered important for enterococcal infections (37, 38). Interestingly, a study that evaluated the origin of 163 E. faecalis isolates in relation to biofilm production showed that endocarditis isolates are associated with greater biofilm densities than nonendocarditis isolates (39). Our results showed that the efbA deletion mutant had a significant reduction in its biofilm-forming capacity, therefore demonstrating that the EfbA fibronectin-binding protein contributes to the biofilm phenotype. Taken together, with the data on whole-cell binding to fibronectin, these results are consistent with the idea that the abrogation of EfbA from the cell surface of E. faecalis OG1RF causes reduced attachment/maintenance of adherence to aortic valves or to the sterile vegetation created by the catheter during infective endocarditis. Emergence of multi-antibiotic-resistant enterococci causing endocarditis poses a clinical challenge (7, 40). In this regard, there is increasing interest in the use of alternatives approaches to current antibiotic strategies such as the development of novel immunotherapies targeting proteins that are important for pathogenicity (41, 42). MSCRAMMs, which are factors ubiquitously present in traditional and opportunistic Gram-positive pathogens, have been used as a target to develop vaccine candidates against E. faecalis (10, 43–45). Three proteins have been identified thus far as potential targets against E. faecalis endocarditis. Singh et al. demonstrated that passive and active immunization against Ace diminished the bacterial counts recovered from aortic valve vegetations and protected against E. faecalis cardiac valve infection (10). IgG Fab fragments against the aggregation substance AS were shown to confer partial protection against enterococcal endocarditis (45). In addition, passive immunization using a monoclonal antibody raised against the major component of E. faecalis pili, EbpC, significantly prevented the establishment of rat endocardi-

4492

iai.asm.org

tis (44). Our results with active vaccination using EfbA showed a significant reduction (⬎35%) in the number of immunized rats that developed E. faecalis infective endocarditis. Consistent with these data, vaccination reduced the bacterial burden in vegetation by 2.28 logs compared to the control mice that received only sham immunization. High-titer antibody response to EfbA was demonstrated in rats immunized with rEfbA prior to E. faecalis experimental endocarditis, demonstrating that EfbA, when expressed during infection, can be recognized by the host immune system. Thus, targeting EfbA is a promising modality that may provide a prophylactic strategy for preventing enterococcal infection, presumably by interfering with bacterial adherence. However, the fact that OG1RF is still able to cause disease in some of rEfbAvaccinated rats indicates that complete protection against E. faecalis infective endocarditis would likely require targeting multiple adhesins. In conclusion, our study demonstrates that EfbA is an E. faecalis OG1RF virulence-associated factor, playing an important role in adherence to fibronectin, in biofilm formation and colonization of aortic valves in an experimental rat endocarditis model. Active rEfbA-based vaccination protected against OG1RF experimental endocarditis and was associated with a reduced number of bacteria in vegetations. Since EfbA homologues are found in a wide range of enterococci, streptococci, and staphylococci, it is conceivable that these immunogenic proteins can be used as a target for the development of effective immunotherapeutic strategies for prevention of Gram-positive infections. ACKNOWLEDGMENTS We thank Karen Jacques-Palaz for technical assistance. This study was supported by the grant R01 AI047923 from the National Institute of Allergy and Infectious Diseases to B.E.M.

REFERENCES 1. Lebreton F, Willems RJL, Gilmore MS. 2014. Enterococcus diversity, origins in nature, and gut colonization. In Gilmore MS, Clewell DB, Ike Y, Shankar N (ed), Enterococci: from commensals to leading causes of drugresistant infection. Massachusetts Eye and Ear Infirmary, Boston, MA. 2. Arias CA, Murray BE. 2012. The rise of the Enterococcus: beyond vancomycin resistance. Nat Rev Microbiol 10:266 –278. http://dx.doi.org/10 .1038/nrmicro2761. 3. Murray BE. 1990. The life and times of the Enterococcus. Clin Microbiol Rev 3:46 – 65. 4. Fernandez-Guerrero ML, Verdejo C, Azofra J, de Gorgolas M. 1995. Hospital-acquired infectious endocarditis not associated with cardiac surgery: an emerging problem. Clin Infect Dis 20:16 –23. 5. Giannitsioti E, Skiadas I, Antoniadou A, Tsiodras S, Kanavos K, Triantafyllidi H, Giamarellou H. 2007. Nosocomial versus communityacquired infective endocarditis in Greece: changing epidemiological profile and mortality risk. Clin Microbiol Infect 13:763–769. http://dx.doi.org /10.1111/j.1469-0691.2007.01746.x. 6. Tleyjeh IM, Steckelberg JM, Murad HS, Anavekar NS, Ghomrawi HM, Mirzoyev Z, Moustafa SE, Hoskin TL, Mandrekar JN, Wilson WR, Baddour LM. 2005. Temporal trends in infective endocarditis: a population-based study in Olmsted County, Minnesota. JAMA 293:3022–3028. http://dx.doi.org/10.1001/jama.293.24.3022. 7. Fernandez-Hidalgo N, Almirante B, Gavalda J, Gurgui M, Pena C, de Alarcon A, Ruiz J, Vilacosta I, Montejo M, Vallejo N, Lopez-Medrano F, Plata A, Lopez J, Hidalgo-Tenorio C, Galvez J, Saez C, Lomas JM, Falcone M, de la Torre J, Martinez-Lacasa X, Pahissa A. 2013. Ampicillin plus ceftriaxone is as effective as ampicillin plus gentamicin for treating Enterococcus faecalis infective endocarditis. Clin Infect Dis 56:1261– 1268. http://dx.doi.org/10.1093/cid/cit052. 8. Nigo M, Munita JM, Arias CA, Murray BE. 2014. What’s new in the treatment of enterococcal endocarditis? Curr Infect Dis Rep 16:431. http: //dx.doi.org/10.1007/s11908-014-0431-z.

Infection and Immunity

December 2015 Volume 83 Number 12

EfbA Is Important in E. faecalis Endocarditis

9. Werner G, Coque TM, Hammerum AM, Hope R, Hryniewicz W, Johnson A, Klare I, Kristinsson KG, Leclercq R, Lester CH, Lillie M, Novais C, Olsson-Liljequist B, Peixe LV, Sadowy E, Simonsen GS, Top J, Vuopio-Varkila J, Willems RJ, Witte W, Woodford N. 2008. Emergence and spread of vancomycin resistance among enterococci in Europe. Euro Surveill 13:pii.19046. 10. Singh KV, Nallapareddy SR, Sillanpaa J, Murray BE. 2010. Importance of the collagen adhesin ace in pathogenesis and protection against Enterococcus faecalis experimental endocarditis. PLoS Pathog 6:e1000716. http: //dx.doi.org/10.1371/journal.ppat.1000716. 11. Sillanpaa J, Chang C, Singh KV, Montealegre MC, Nallapareddy SR, Harvey BR, Ton-That H, Murray BE. 2013. Contribution of individual Ebp pilus subunits of Enterococcus faecalis OG1RF to pilus biogenesis, biofilm formation and urinary tract infection. PLoS One 8:e68813. http: //dx.doi.org/10.1371/journal.pone.68813. 12. Nallapareddy SR, Singh KV, Sillanpaa J, Garsin DA, Hook M, Erlandsen SL, Murray BE. 2006. Endocarditis and biofilm-associated pili of Enterococcus faecalis. J Clin Invest 116:2799 –2807. http://dx.doi.org/10 .1172/JCI29021. 13. Pankov R, Yamada KM. 2002. Fibronectin at a glance. J Cell Sci 115: 3861–3863. http://dx.doi.org/10.1242/jcs.00059. 14. Torelli R, Serror P, Bugli F, Paroni Sterbini F, Florio AR, Stringaro A, Colone M, De Carolis E, Martini C, Giard JC, Sanguinetti M, Posteraro B. 2012. The PavA-like fibronectin-binding protein of Enterococcus faecalis, EfbA, is important for virulence in a mouse model of ascending urinary tract infection. J Infect Dis 206:952–960. http://dx.doi.org/10.1093/infdis /jis440. 15. Pracht D, Elm C, Gerber J, Bergmann S, Rohde M, Seiler M, Kim KS, Jenkinson HF, Nau R, Hammerschmidt S. 2005. PavA of Streptococcus pneumoniae modulates adherence, invasion, and meningeal inflammation. Infect Immun 73:2680 –2689. http://dx.doi.org/10.1128/IAI.73.5 .2680-2689.2005. 16. Holmes AR, McNab R, Millsap KW, Rohde M, Hammerschmidt S, Mawdsley JL, Jenkinson HF. 2001. The pavA gene of Streptococcus pneumoniae encodes a fibronectin-binding protein that is essential for virulence. Mol Microbiol 41:1395–1408. http://dx.doi.org/10.1046/j.1365 -2958.2001.02610.x. 17. Kawabata S, Kunitomo E, Terao Y, Nakagawa I, Kikuchi K, Totsuka K, Hamada S. 2001. Systemic and mucosal immunizations with fibronectinbinding protein FBP54 induce protective immune responses against Streptococcus pyogenes challenge in mice. Infect Immun 69:924 –930. http: //dx.doi.org/10.1128/IAI.69.2.924-930.2001. 18. Christie J, McNab R, Jenkinson HF. 2002. Expression of fibronectinbinding protein FbpA modulates adhesion in Streptococcus gordonii. Microbiology 148:1615–1625. http://dx.doi.org/10.1099/00221287-148-6 -1615. 19. Miller-Torbert TA, Sharma S, Holt RG. 2008. Inactivation of a gene for a fibronectin-binding protein of the oral bacterium Streptococcus mutans partially impairs its adherence to fibronectin. Microb Pathog 45:53–59. http://dx.doi.org/10.1016/j.micpath.2008.02.001. 20. Mu R, Kim BJ, Paco C, Del Rosario Y, Courtney HS, Doran KS. 2014. Identification of a group B streptococcal fibronectin binding protein, SfbA, that contributes to invasion of brain endothelium and development of meningitis. Infect Immun 82:2276 –2286. http://dx.doi.org/10.1128 /IAI.01559-13. 21. Kadioglu A, Brewin H, Hartel T, Brittan JL, Klein M, Hammerschmidt S, Jenkinson HF. 2010. Pneumococcal protein PavA is important for nasopharyngeal carriage and development of sepsis. Mol Oral Microbiol 25:50 – 60. http://dx.doi.org/10.1111/j.2041-1014.2009.00561.x. 22. Panesso D, Montealegre MC, Rincon S, Mojica MF, Rice LB, Singh KV, Murray BE, Arias CA. 2011. The hylEfm gene in pHylEfm of Enterococcus faecium is not required in pathogenesis of murine peritonitis. BMC Microbiol 11:20. http://dx.doi.org/10.1186/1471-2180-11-20. 23. Kristich CJ, Chandler JR, Dunny GM. 2007. Development of a hostgenotype-independent counterselectable marker and a high-frequency conjugative delivery system and their use in genetic analysis of Enterococcus faecalis. Plasmid 57:131–144. http://dx.doi.org/10.1016/j.plasmid .2006.08.003. 24. Holo H, Nes IF. 1989. High-frequency transformation, by electroporation, of Lactococcus lactis subsp. cremoris grown with glycine in osmotically stabilized media. Appl Environ Microbiol 55:3119 –3123. 25. Murray BE, Singh KV, Ross RP, Heath JD, Dunny GM, Weinstock GM. 1993. Generation of restriction map of Enterococcus faecalis OG1 and in-

December 2015 Volume 83 Number 12

26.

27.

28.

29.

30. 31. 32.

33. 34. 35. 36. 37.

38. 39.

40. 41. 42.

43.

44.

vestigation of growth requirements and regions encoding biosynthetic function. J Bacteriol 175:5216 –5223. Nallapareddy SR, Singh KV, Sillanpaa J, Zhao M, Murray BE. 2011. Relative contributions of Ebp pili and the collagen adhesin ace to host extracellular matrix protein adherence and experimental urinary tract infection by Enterococcus faecalis OG1RF. Infect Immun 79:2901–2910. http://dx.doi.org/10.1128/IAI.00038-11. Nallapareddy SR, Qin X, Weinstock GM, Hook M, Murray BE. 2000. Enterococcus faecalis adhesin, ace, mediates attachment to extracellular matrix proteins collagen type IV and laminin as well as collagen type I. Infect Immun 68:5218 –5224. http://dx.doi.org/10.1128/IAI.68.9.5218 -5224.2000. Nallapareddy SR, Singh KV, Duh RW, Weinstock GM, Murray BE. 2000. Diversity of ace, a gene encoding a microbial surface component recognizing adhesive matrix molecules, from different strains of Enterococcus faecalis and evidence for production of ace during human infections. Infect Immun 68:5210 –5217. http://dx.doi.org/10.1128/IAI.68.9 .5210-5217.2000. Fernandez Guerrero ML, Goyenechea A, Verdejo C, Roblas RF, de Gorgolas M. 2007. Enterococcal endocarditis on native and prosthetic valves: a review of clinical and prognostic factors with emphasis on hospital-acquired infections as a major determinant of outcome. Medicine 86:363–377. http://dx.doi.org/10.1097/MD.0b013e31815d5386. Angrist AA, Oka M. 1963. Pathogenesis of bacterial endocarditis. JAMA 183:249 –252. Bergmeier W, Hynes RO. 2012. Extracellular matrix proteins in hemostasis and thrombosis. Cold Spring Harb Perspect Biol 4:pii.a005132. http: //dx.doi.org/10.1101/cshperspect.a005132. Heying R, van de Gevel J, Que YA, Moreillon P, Beekhuizen H. 2007. Fibronectin-binding proteins and clumping factor A in Staphylococcus aureus experimental endocarditis: FnBPA is sufficient to activate human endothelial cells. Thromb Haemost 97:617– 626. Lowrance JH, Baddour LM, Simpson WA. 1990. The role of fibronectin binding in the rat model of experimental endocarditis caused by Streptococcus sanguis. J Clin Invest 86:7–13. Schennings T, Heimdahl A, Coster K, Flock JI. 1993. Immunization with fibronectin binding protein from Staphylococcus aureus protects against experimental endocarditis in rats. Microb Pathog 15:227–236. Foster TJ, Geoghegan JA, Ganesh VK, Hook M. 2014. Adhesion, invasion and evasion: the many functions of the surface proteins of Staphylococcus aureus. Nat Rev Microbiol 12:49 – 62. Chhatwal GS. 2002. Anchorless adhesins and invasins of Gram-positive bacteria: a new class of virulence factors. Trends Microbiol 10:205–208. http://dx.doi.org/10.1016/S0966-842X(02)02351-X. Carniol K, Gilmore MS. 2004. Signal transduction, quorum-sensing, and extracellular protease activity in Enterococcus faecalis biofilm formation. J Bacteriol 186:8161– 8163. http://dx.doi.org/10.1128/JB.186.24.8161-8163 .2004. Donelli G, Guaglianone E. 2004. Emerging role of Enterococcus spp. in catheter-related infections: biofilm formation and novel mechanisms of antibiotic resistance. J Vasc Access 5:3–9. Mohamed JA, Huang W, Nallapareddy SR, Teng F, Murray BE. 2004. Influence of origin of isolates, especially endocarditis isolates, and various genes on biofilm formation by Enterococcus faecalis. Infect Immun 72: 3658 –3663. http://dx.doi.org/10.1128/IAI.72.6.3658-3663.2004. Murray BE. 2000. Vancomycin-resistant enterococcal infections. N Engl J Med 342:710 –721. http://dx.doi.org/10.1056/NEJM200003093421007. Koch S, Hufnagel M, Theilacker C, Huebner J. 2004. Enterococcal infections: host response, therapeutic, and prophylactic possibilities. Vaccine 22:822– 830. http://dx.doi.org/10.1016/j.vaccine.2003.11.027. Romero-Saavedra F, Laverde D, Wobser D, Michaux C, Budin-Verneuil A, Bernay B, Benachour A, Hartke A, Huebner J. 2014. Identification of peptidoglycan-associated proteins as vaccine candidates for enterococcal infections. PLoS One 9:e111880. http://dx.doi.org/10.1371/journal.pone .0111880. Flores-Mireles AL, Pinkner JS, Caparon MG, Hultgren SJ. 2014. EbpA vaccine antibodies block binding of Enterococcus faecalis to fibrinogen to prevent catheter-associated bladder infection in mice. Sci Transl Med 6:254ra127. http://dx.doi.org/10.1126/scitranslmed.3009384. Pinkston KL, Singh KV, Gao P, Wilganowski N, Robinson H, Ghosh S, Azhdarinia A, Sevick-Muraca EM, Murray BE, Harvey BR. 2014. Targeting pili in enterococcal pathogenesis. Infect Immun 82:1540 –1547. http://dx.doi.org/10.1128/IAI.01403-13.

Infection and Immunity

iai.asm.org

4493

Singh et al.

45. Schlievert PM, Chuang-Smith ON, Peterson ML, Cook LC, Dunny GM. 2010. Enterococcus faecalis endocarditis severity in rabbits is reduced by IgG Fabs interfering with aggregation substance. PLoS One 5:e13194. http: //dx.doi.org/10.1371/journal.pone.0013194. 46. Leenhouts K, Buist G, Bolhuis A, ten Berge A, Kiel J, Mierau I, Dabrowska M, Venema G, Kok J. 1996. A general system for generating unlabeled gene replacements in bacterial chromosomes. Mol Gen Genet 253:217–224.

4494

iai.asm.org

47. Bourgogne A, Garsin DA, Qin X, Singh KV, Sillanpaa J, Yerrapragada S, Ding Y, Dugan-Rocha S, Buhay C, Shen H, Chen G, Williams G, Muzny D, Maadani A, Fox KA, Gioia J, Chen L, Shang Y, Arias CA, Nallapareddy SR, Zhao M, Prakash VP, Chowdhury S, Jiang H, Gibbs RA, Murray BE, Highlander SK, Weinstock GM. 2008. Large scale variation in Enterococcus faecalis illustrated by the genome analysis of strain OG1RF. Genome Biol 9:R110. http://dx.doi.org/10.1186/gb-2008-9 -7-r110.

Infection and Immunity

December 2015 Volume 83 Number 12