|Home | About | Journals | Submit | Contact Us | Français|
In addition to its pivotal role in hemostasis, fibrinogen (Fg) and provisional fibrin matrices play important roles in inflammation and regulate innate immune responses by interacting with leukocytes. Efb (the extracellular fibrinogen-binding protein) is a secreted Staphylococcus aureus protein that engages host Fg and complement C3. However, the molecular details underlying the Efb-Fg interaction and the biological relevance of this interaction have not been determined. In the present study, we characterize the interaction of Efb with Fg. We demonstrate that the Fg binding activity is located within the intrinsically disordered N-terminal half of Efb (Efb-N) and that the D fragment of Fg is the region that mediates Efb-N binding. More detailed studies of the Efb-N-Fg interactions using ELISA and surface plasmon resonance analyses revealed that Efb-N exhibits a much higher affinity for Fg than typically observed with Fg-binding MSCRAMMs (microbial surface components recognizing adhesive matrix molecules), and data obtained from ELISA analyses using truncated Efb-N constructs demonstrate that Efb-N contains two binding sites located within residues 30–67 and 68–98, respectively. Efb-N inhibits neutrophil adhesion to immobilized Fg by binding to Fg and blocking the interaction of the protein with the leukocyte integrin receptor, αMβ2. A motif in the Fg γ chain previously shown to be central to the αMβ2 interaction was shown to be functionally distinguishable from the Efb-N binding site, suggesting that the Fg-Efb interaction indirectly impedes Fg engagement by αMβ2. Taken together, these studies provide insights into how Efb interacts with Fg and suggest that Efb may support bacterial virulence at least in part by impeding Fg-driven leukocyte adhesion events.
Fibrinogen (Fg)2 is an acute phase plasma protein that is the fundamental building block of insoluble fibrin clots and is understood to play an important role in hemostasis and inflammatory processes (1). Mice lacking Fg exhibit a delayed inflammatory response and defects in wound healing (2, 3). In addition to providing substrates for bacterial colonization, Fg also participates in the innate immune defense through interactions with leukocytes that support leukocyte activation events and delay apoptosis (4). The interaction of Fg/fibrin with the leukocyte integrin αMβ2 contributes to the development of inflammatory disease processes, including inflammatory joint disease and neuroinflammatory disease (5, 6) as well as physiological processes, such as the clearance of Staphylococcus aureus within the peritoneal cavity (7).
Numerous bacterial proteins interact with Fg, including ClfA (clumping factor A), FnbpA, Eap, and coagulase from S. aureus; SdrG from Staphylococcus epidermidis; M protein from group A, C, and G streptococci; and FbsA from Staphylococcus agalactiae (8). In many cases, these Fg-binding proteins act as virulence factors in animal models of infection. For example, ClfA promotes sepsis in a mouse model (9), and the streptococcal M proteins cause lung lesions in mice (10) in part by altering leukocyte function (11). The goal of the present study was to further characterize the interaction of the secreted S. aureus Efb (extracellular fibrinogen-binding protein) with Fg and to determine how Efb may alter Fg-mediated leukocyte adhesion events.
S. aureus is a versatile, opportunistic Gram-positive pathogen and a leading cause of bacterial infections worldwide (12). It causes a diverse array of diseases, ranging from minor skin infections to life-threatening diseases such as pneumonia, endocarditis, and sepsis (13). S. aureus produces a large arsenal of virulence factors, including cell surface-associated adhesins, and secreted proteins like proteases, toxins, and superantigens (14, 15). Several secreted staphylococcal proteins, including Eap (extracellular adhesive protein), CHIPS (chemotaxis inhibitory protein of staphylococci), and SCIN (staphylococcus complement inhibitor) appear to target and counter the host innate immune response (16) by inhibiting neutrophil trafficking (17), recruitment, chemotaxis (17, 18), and complement activation and preventing opsonization and phagocytosis of the bacteria (19).
Efb is another innate immune evasion molecule secreted by S. aureus, which is reported to inhibit complement activation (20, 21), block platelet aggregation (22, 23), and delay wound healing in a rat wound infection model (24). The 16-kDa Efb protein has a disordered N-terminal region (Efb-N), spanning the sequence Efb(30–104), and a folded domain in the C-terminal region (Efb-C), spanning the sequence Efb(105–165) (25, 26). Earlier studies in our laboratory (20) and those of others (21) established that Efb-C binds to C3 and blocks C3b-containing convertase activities, resulting in decreased C3b deposition on the bacterial surface and down-regulation of the neutrophil response mediated by C5a. The crystal structure of the Efb-C·C3d complex revealed that the binding of Efb-C to C3 alters the conformation of C3 such that it cannot be processed into C3b (26). Interestingly, Efb-C also blocks the interaction of C3d with complement receptor 2 (CR2), which plays a critical role in B cell maturation and activation, suggesting an additional role of Efb as an inhibitor of the adaptive immunity (27).
Efb also interacts with Fg (25, 28), but the precise motifs on bacterial Efb and host Fg as well as the precise biological significance of this interaction remain unclear (25, 29). Our laboratory previously demonstrated that the intrinsically disordered Efb-N region is important for the Fg binding activity, whereas Efb-C does not contribute to Fg binding (25). Here, we have characterized the interaction of Efb with Fg in detail. Our results demonstrate that Efb binds with unusually high affinity to Fg, and we identify two distinct Fg-binding sequences within the Efb-N domain. Furthermore, we show that this interaction results in an inhibition of neutrophil adherence to immobilized Fg.
For studies with neutrophils from human subjects, written informed consent was obtained from healthy adult individuals with approval of the Baylor College of Medicine Institutional Review Board for Human Subject Research.
Fresh human peripheral blood neutrophils obtained from healthy adult individuals were isolated from citrate anticoagulated venous blood sedimented in 6% dextran (Mr 250,000) and centrifuged over Ficoll-Hypaque gradients at room temperature. Isolated neutrophils were suspended in Dulbecco's PBS (Invitrogen) and maintained at 4 °C for up to 4 h at a concentration of 106 cells/ml.
Neutrophil adherence to Fg-coated surfaces was quantitated using a Muntz-static adhesion chamber (30, 31), consisting of two metal plates holding two 25-mm round glass coverslips separated by a rubber O-ring. The coverslips, cleaned with 10% NaOH, 57% EtOH for 2 h at room temperature and washed with Dulbecco's PBS, were coated with Fg (100 μg/ml in Dulbecco's PBS) overnight at 4 °C. Freshly isolated neutrophils (106/ml) were injected into the chamber through a 25-gauge needle connected to a 1-ml syringe. Neutrophil adherence was quantified by counting the number of neutrophils that settled and attached during an initial 500-s observation period. The adhesion chamber was then inverted for an additional 500 s to allow nonadherent neutrophils to fall off. The remaining adherent neutrophils were counted and expressed as a percentage of adherent cells. All experiments were conducted at 37 °C using a Nikon Diaphot inverted phase-contrast microscope. To study the effect of staphylococcus proteins on neutrophil adherence, neutrophils were mixed with 100 μl of 5 μm His-tagged recombinant SdrG or Efb protein (0.5 μm final concentrations) immediately before injection into the adhesion chamber. In the case of preincubation experiments, either neutrophils were preincubated with recombinant Efb for 20 min, collected by centrifugation, washed with PBS, and added to an Fg-coated chamber, or Fg-coated glass coverslips were preincubated with Efb for 1 h and washed with PBS before adding neutrophils.
A human embryonic kidney 293 (HEK 293) cell line stably expressing αMβ2 (HEK 293-Mac-1) (kindly provided by Dr. Edward F. Plow) (32) was maintained in DMEM/F-12 (Lonza) supplemented with 10% fetal bovine serum (FBS), 2 μm l-glutamine, 100 units/ml penicillin, 100 μg/ml streptomycin, and 600 μg/ml G418. The HEK 293 cell line was maintained in 5% FBS, DMEM (Thermo Scientific). Prior to use, cells were harvested with cell dissociation buffer (Invitrogen), washed, and suspended in TBS containing 1 mm CaCl2, 2 mm MgCl2, and 2 mm MnCl2. For cell adherence assays, 96-well plates (Thermo Scientific) were coated with 100 μl of Fg (10 μg/ml; Enzyme Research) overnight at 4 °C and then postcoated with 0.05% polyvinyl alcohol (30–70 kDa; Sigma) for 1 h at 37 °C. Subsequently, the cells were seeded at 50,000 cells/well and incubated at 37 °C for 1 h. Non-adherent cells were removed by washing gently twice with PBS. Adherent cells were fixed with 200 μl of 0.25% glutaraldehyde for 15 min at room temperature, followed by 100 μl/well 100 mm glycine for 10 min. The numbers of adherent cells were determined by staining with 0.2% crystal violet for 15 min (33). The optical density of crystal violet was measured at 590 nm after solubilizing the dye in 10% acetic acid.
Genomic DNA isolated from S. aureus strain Newman was used as template for all PCRs using the oligonucleotide primers described in supplemental Table 1. PCR products were digested with BamHI and EcoRI and ligated into the pGEX-5x-1 vector (GE Healthcare) or digested with BamHI and PstI and ligated into the pQE30 (Invitrogen). The ligation mixture was transformed into Escherichia coli XL-1 blue (Stratagene), grown on an LB agar plate containing 100 μg/ml ampicillin to select for transformants. Insertions were confirmed by DNA sequencing.
E. coli strain M15 (pREP4) (Qiagen) containing plasmids encoding N-terminal glutathione S-transferase (GST) or N-terminal His6-tagged Efb fusion proteins were grown overnight at 37 °C in LB containing 100 μg/ml ampicillin and 25 μg/ml kanamycin. The overnight cultures were diluted 1:20 into fresh LB medium, and recombinant protein expression was induced with 0.2 mm isopropyl 1-thio-β-d-galactopyranoside for 2–3 h. Bacteria were harvested by centrifugation and lysed using a French press (SLM Aminco). Soluble proteins were purified through a glutathione-Sepharose 4B column (GE Healthcare) or by nickel-chelating chromatography (GE Healthcare) according to the manufacturer's manual. Purified proteins were dialyzed into PBS and stored at −20 °C. Protein concentrations were determined by the Bradford assay (Pierce).
Fg D fragments were generated by digestion of Fg (Enzyme Research) with plasmin (Enzyme Research; 10 μg/15 mg Fg) in TBS containing 10 mm CaCl2 for 4 h at 37 °C as described earlier (34) with modifications. D fragments were obtained by column gel filtration on Sephacryl S-200 (GE Healthcare). Purified D fragments were analyzed on SDS-PAGE with a molecular mass of 85 kDa. Fg E fragments were purchased from Calbiochem.
96-well immulon 4HBX microtiter plates (Thermo Scientific) were coated overnight at 4 °C with 100 μl of 2.5 μg/ml full-length human Fg (diluted in PBS; Enzyme Research) unless otherwise indicated. After blocking the wells with 3% BSA in PBS, recombinant Efb proteins were added, and the plates were incubated for 1 h. Bound Efb proteins were detected through incubation with horseradish peroxidase (HRP)-conjugated anti-His antibodies (10,000× dilution) or HRP-conjugated anti-GST polyclonal antibodies (5000× dilution) for 1 h and quantified after adding the substrate o-phenylenediamine dihydrochloride (Sigma) by measuring the resulting absorbance at 450 nm in an ELISA microplate reader (Thermomax).
In the case of the Efb peptide inhibition assay, various concentrations of Efb peptides were mixed with a fixed concentration of Efb-GST fusion proteins (10 nm) in PBS, and the bound Efb-GST fusion proteins were detected through incubation with HRP-conjugated rabbit anti-GST polyclonal antibodies (Abcam) (5000× dilution). All proteins were diluted in PBS containing 1% BSA and 0.05% Tween 20, and the ELISAs were carried out at room temperature. 50% inhibition concentration (IC50) is obtained by fitting raw data into a one-site competition equation using GraphPad Prism version 4.0.
2 μm human Fg (Enzyme Research) or isolated Fg D fragment, generated by plasmin digestion, was incubated with different concentrations of Efb-His fusion proteins in 20 mm HEPES buffer (pH 7.4) with 150 mm NaCl and 0.1% Tween 20 in a 96-well assay plate (Falcon3912) for 2 h at room temperature. The turbidity of the protein solution mixture was monitored at an optical density of 405 nm.
The interaction between Efb proteins and the soluble, isolated D fragment of Fg was further characterized by isothermal titration calorimetry (ITC) using a VP-ITC microcalorimeter (MicroCal). The Fg D fragment used in these studies was generated by digesting full-length Fg with plasmin for 4 h and fractionating the digestion products by gel filtration chromatography. The ITC cell contained 6 μm Fg D fragments, and the syringe contained 50–70 μm Efb-GST fusion proteins in TBS (25 mm Tris, 3.0 mm KCl, and 140 mm NaCl, pH 7.4). All proteins were filtered through 0.22-μm membranes and degassed for 20 min before use. The titrations were performed at 27 °C using a single preliminary injection of 2 μl of Efb fusion proteins followed by 30–40 injections of 5 μl with an injection speed of 0.5 μl/s. Injections were spaced over 5-min intervals at a stirring speed of 260 rpm. Raw titration data were fit to a one-site model of binding using MicroCal Origin version 5.0.
SPR-based Biacore binding experiments were performed at 25 °C on a Biacore 3000 (GE Healthcare/Biacore). Ligand surfaces were prepared using an amine coupling procedure as recommended by the manufacturer. HEPES-buffered saline (HBS-T: 10 mm HEPES, pH 7.3, 150 mm NaCl, and 0.005% Tween 20) was used as the running buffer at a flow rate of 5 μl/min. Twenty microliters of Fg D fragment (15 μg/ml in sodium acetate, pH 5.0) were injected into an activated (7 min) flow cell on a CM5 chip. After deactivation, the flow cell surface was washed with running buffer until the surface was stable. His-tagged Efb-N, which has a predicted pI of 8.83, was immobilized onto the C1 chip. After a 2-min activation, 10 μl of protein (2 μg/ml in water) was injected, and the flow cell was subsequently deactivated. The immobilization procedures mentioned above created a ligand density of about 1900 response units for Fg D fragment and 50 response units for His-tagged Efb-N. A reference surface lacking coupled protein was prepared on each sensor chip. Binding studies were performed at a flow rate of 30 μl/min with HBS-T as the running buffer. To regenerate the ligand surfaces, bound proteins were removed by 10- or 20-s injections of 5 mm NaOH. All SPR responses were base line-corrected by subtracting the response generated from the reference surface. In order to obtain the binding kinetics data, 2-fold increasing concentrations of proteins in running buffer were injected over the ligand and reference surfaces. Base line-corrected SPR response curves (with buffer blank run subtracted) were globally fitted (except Rmax locally) to the 1:1 (Langmuir) binding model using BIAevaluation software (version 4.1). Association and dissociation rate constants (ka and kd, respectively) for the interactions were derived from the fitting, and the dissociation constant KD was calculated (KD = kd/ka).
We initially explored the binding of Efb to immobilized Fg using purified His-tagged full-length Efb, Efb-N and Efb-C in ELISA-type binding assays. Binding of full-length Efb and Efb-N, but not Efb-C, to Fg was dose-dependent and exhibited saturation kinetics. Half-maximum binding for Efb and Efb-N was observed at concentrations of 1.21 and 1.16 nm, respectively (Fig. 1A). These observations corroborate our previous results (25). More detailed studies revealed that Efb binds to both immobilized and soluble forms of Fg (Fig. 1, B and C). In a competition ELISA-type assay, soluble Fg efficiently inhibited the binding of Efb to surface-bound Fg, with 50% inhibition achieved at 142 ± 1 nm (Fig. 1C). Soluble Fg also bound to immobilized Efb coated onto microtiter plates (Fig. 1B). To locate the Efb binding site(s) in Fg, we tested isolated D and E fragments of Fg in an ELISA-type binding assay. Efb-N bound equally well to the Fg D fragment (Fg-D) and to intact Fg but did not bind to the Fg E fragment (Fg-E) (Fig. 1D).
The Efb-Fg interaction was further characterized by SPR using a Biacore 3000 system with either a CM5 sensor chip derivatized with Fg-D or a C1 sensor chip derivatized with His-tagged Efb. The results obtained were consistent with our observations from the ELISA-type binding assays; full-length Efb and Efb-N, but not Efb-C, bound to Fg-D (Fig. 2A). We determined that Efb-N bound to Fg-D in a dose-dependent manner with a calculated KD of 0.23 nm based on kinetic data (Fig. 2B). The interaction of Efb-N with Fg-D exhibits a reasonably fast on rate (ka = 5.03 × 105 m−1 s−1) but a very slow off rate (kd = 1.14 × 10−4 s−1), suggesting that once formed, the complex is very stable (Fig. 2B). Interestingly, when the system was reversed and the Efb-N fusion protein was immobilized, Fg-D bound to the Efb-N chip in a dose-dependent manner but with a higher KD (2.67 nm) due to a faster off rate (kd = 7.9 × 10−4 s−1) (Fig. 2C). The molecular explanation for this difference is presently not clear, but regardless of the experimental design, the binding affinities between this bacterial product and host fibrinogen appear to be remarkably high.
We observed that a solution containing an equimolar ratio of Efb and Fg (a symmetrical dimer composed of six polypeptide chains; Aα2Bβ2γ2) turned cloudy, indicating the formation of aggregates. To elucidate whether this was due to the formation of multimers or precipitation as a result of conformational changes, we incubated Fg, at a set concentration (2 μm), with increasing amounts of the different Efb proteins and assessed turbidity. Interestingly, combining either full-length Efb or Efb-N with Fg resulted in a concentration-dependent and bell-shaped change in light scatter, whereby turbidity initially increased and then subsequently decreased as the concentration of the Efb proteins increased (Fig. 3A). This behavior is reminiscent of an immunoprecipitation curve, suggesting that the observed turbidity results from the formation of multimers. The addition of Efb-C, which does not bind Fg, did not induce a measurable turbidity change (Fig. 3A). The ability of full-length Efb and Efb-N to induce the formation of multimers suggests that each protein contains at least two Fg-binding sites. Notably, incubating the monomeric D fragment of Fg with either full-length Efb or Efb-N did not yield changes in turbidity (Fig. 3B), suggesting that the two proposed binding sites on each Efb-N molecule do not bind to independent motifs within individual D region.
To identify the two putative Fg-binding sites located within Efb, we generated a panel of recombinant truncated Efb-N proteins fused to GST and evaluated their Fg binding ability using the ELISA-type assay. Efb-N contains two nearly identical repeat segments (Efb(46–67) (Efb-L) and Efb(77–98) (Efb-I)) composed of 22 residues each. These repeats were previously shown to contribute to Fg binding (35), but neither element appears to be sufficient to support the high affinity Fg interaction. Recombinant forms containing these two segments (Efb-L and Efb-I; Fig. 4A) did not bind to Fg in the ELISA (Fig. 4B). However, the addition of short N-terminal extensions to generate Efb(30–67) (Efb-A) and Efb(68–98) (Efb-O) resulted in proteins with Fg binding activity where half-maximal binding was observed at a concentration of 1.1 and 0.64 nm, respectively (Fig. 4B). Interestingly, Efb fragments covering residues 30–45 (Efb-K) and 68–76 (Efb-M) did not show Fg binding activity (Fig. 4B), suggesting that the Fg-binding sites require longer sequences of Efb. Taken together, it appears that the Efb-K and Efb-M segments are insufficient for Fg binding, but residues in these regions are necessary for conferring Fg binding activity to Efb-A and Efb-O.
To characterize the Fg binding affinity of Efb-A and Efb-O in greater detail, ITC was conducted with plasmin-generated Fg-D fragment and GST-tagged Efb-A and Efb-O. GST fusion proteins were used in these experiments because they are more soluble than the corresponding synthetic peptides. Strikingly, in solution, Efb-O bound to Fg-D with significantly higher affinity than Efb-A with dissociation constants of 4.66 nm and 1.0 μm, respectively (Fig. 5, A and B). Thus, the affinity of Efb-O for Fg-D is 200 times higher than that of Efb-A, suggesting that Efb-O contains the premiere Fg binding motif in Efb (Fig. 5). Purified GST did not interact with Fg-D and served as the control (data not shown). The remarkable feature of both interactions is the excess favorable binding enthalpy (ΔH ~−27 kcal mol−1; Table 1), presumably due to a large number of hydrogen bonds and van der Waals interactions formed in the complex. However, the favorable interaction enthalpy is countered by very large unfavorable entropy changes (ΔS ~−30 to −60 cal mol−1K−1; Table 1), particularly for Efb-A (ΔS ~−60 cal mol−1K−1; Table 1), indicating a vast loss in the degree of freedom upon binding. This thermodynamic unfavorable entropy change most likely results from reduction of conformational flexibility in the two interacting proteins upon binding (36, 37).
In the competition ELISAs, the synthetic peptide Efb-o (a 31-residue peptide corresponding to Efb amino acids 68–98) effectively inhibited GST-tagged Efb-N protein binding to Fg, whereas synthetic peptide Efb-a (a 38-residue peptide corresponding to Efb amino acids 30–67) had no effect on the Fg binding activity of recombinant Efb-N even when the peptide was used at a concentration of 100 μm (Fig. 6A) (data not shown). These results further demonstrated that Efb-O contains the major Fg-binding site. We also compared the ability of peptides Efb-a, Efb-o, and Efb-h (a 28-residue peptide corresponding to Efb amino acids 77–104) to compete with the Fg binding activity of GST-tagged Efb-A and Efb-O. Peptide Efb-o successfully inhibited GST-Efb-A (10 nm) binding to Fg with a half-maximal inhibitory concentration (IC50) of 24 ± 1 nm (Fig. 6B). In contrast, synthetic Efb-a peptide was not able to compete with GST-Efb-O for binding to Fg even at a concentration of 100 μm, which is in 10,000-fold molar excess of the peptide compared with the GST-tagged protein (Fig. 6C) (data not shown), again suggesting that the Efb-O binds much more strongly to Fg than Efb-A and confirming that Efb-O contains the major binding site for Fg. The ability of Efb-o to compete with Efb-A in Fg binding suggests that these two Efb fragments bind to the same or partially overlapping sites in Fg. Peptide Efb-h did not interfere with the Fg binding of GST-tagged Efb-N (Fig. 6A), Efb-A (Fig. 6B), and Efb-O (Fig. 6C), suggesting that residues 68–76 are essential for the Fg binding activity and that residues 99–104 may not significantly contribute to Fg binding. To further determine the precise sequences in Efb-O that binds to Fg, truncated Efb-o peptides with deletions at either the N terminus or C terminus were synthesized, and their ability to inhibit GST-tagged Efb-N binding to Fg was evaluated. Truncated Efb-o peptides lacking either three N-terminal residues (Efb-o-N3), six N-terminal residues (Efb-o-N6), 10 C-terminal residues (Efb-o-C10), or 15 C-terminal residues (Efb-o-C15) each lost the ability to block GST-Efb-N binding to Fg. However, a five-residue deletion at the C terminus of Efb-o (Efb-o-C5) still effectively inhibited the Fg binding activity of GST-Efb-N, suggesting that the 26-residue peptide Efb-o-C5, corresponding to Efb amino acids 68–93, encompasses the minimal binding site for Fg (Fig. 6D).
We then examined the effect of Efb on neutrophil adherence to Fg using the static Muntz adhesion chambers described previously (30, 31, 38). The recombinant full-length Efb effectively inhibited neutrophil attachment to immobilized Fg, reducing the number of attached cells by more than 80% compared with the non-treated control group (Fig. 7A). To determine which segment of Efb is responsible for this inhibitory effect, we examined the ability of purified recombinant Efb-N and Efb-C to inhibit neutrophil attachment. We found that Efb-N reduced neutrophil attachment to Fg to the same extent as full-length Efb, whereas Efb-C had no effect (Fig. 7A). A recombinant Fg-binding fragment of SdrG, a cell wall-anchored Fg-binding MSCRAMM (microbial surface component recognizing adhesive matrix molecules) from S. epidermidis (39), did not affect neutrophil attachment to Fg and served as a negative control (Fig. 7A).
The morphology of neutrophil changes from a spherical to a bipolar shape in response to stimulation by chemotactic factors or to adherence to a substratum (40). Microscopic analysis of neutrophils 10 min after treatment with Efb-C exhibited a bipolar morphology similar to the non-treated control group, suggesting that neutrophils that attached to Fg under these conditions also can respond with shape changes and that Efb-C does not interfere with this process (Fig. 7B). However, the few neutrophils attached in the presence of full-length Efb or Efb-N retained a spherical shape (Fig. 7B), indicating that these neutrophils did not progress with the adherence-associated morphological changes.
To further explore the mechanism by which Efb blocks neutrophil adherence to Fg, human neutrophils were preincubated with recombinant Efb for 20 min, collected by centrifugation, washed with PBS, and added to an Fg-coated chamber. Neutrophils pre-incubation with full-length Efb or Efb-N, adhered to Fg to the same extent as the non-treated control group in which no Efb was added (Fig. 7C). These data suggest that neither Efb nor Efb-N directly engage and stably interfere with the neutrophil surface molecules required for adhesion to Fg.
In contrast, neutrophils were prevented from adhering to Fg when the Fg-coated surface was pretreated with full-length Efb or Efb-N (Fig. 7D), indicating that a direct interaction between Fg and Efb/Efb-N is responsible for inhibiting subsequent neutrophil adherence to immobilized Fg. Pretreatment of the Fg-coated surface with SdrG did not inhibit neutrophil adherence to Fg, which is consistent with the earlier finding that SdrG does not affect neutrophil attachment to Fg (Fig. 7D). Taken together, these results demonstrate that the binding of Efb to Fg interferes with neutrophil-Fg interaction, resulting in an inhibition of neutrophil adherence to Fg. These results do not formally exclude the possibility that Efb interacts with a neutrophil surface molecule, but these interactions, if they occur, apparently do not affect neutrophil adhesion to Fg.
Because the neutrophil-Fg interaction is primarily an αMβ2-dependent process (7), we sought to directly evaluate the role of αMβ2 in Efb-mediated inhibition of neutrophil adherence. Consistent with prior observations, we found that a monoclonal antibody against the αM-I domain effectively blocked the adhesion of HEK293 cells expressing human αMβ2 (kindly provided by Dr. Edward F. Plow) to immobilized Fg (Fig. 8A). Control HEK293 cells lacking αMβ2 cannot attach to immobilized Fg (Fig. 8A). Moreover, two synthetic Fg peptides based on the αM binding sites in Fg (denoted Fg-P1 (41) and Fg-P2 (42), respectively) abrogated cell attachment to immobilized Fg surface (Fig. 8A). These results were consistent with earlier studies (7) and demonstrated that the attachment of the transfected HEK293 cells to Fg is a specific αMβ2-dependent process. The addition of either full-length Efb or Efb-N to the adhesion assay effectively blocked attachment of the αMβ2-expressing cells to Fg (Fig. 8B), suggesting that the Efb or Efb-N interferes with the αMβ2-Fg interaction. The possibility that Efb competes with αMβ2 for a common site in Fg was investigated. Here, a mutant form of mouse Fg, termed Fgγ390–396A, that retains the capacity to form fibrin in vitro and in vivo but lacks seven amino acid residues required for αMβ2-mediated cell adhesion (7) was compared with wild-type fibrinogen for its capacity to bind Efb. As shown in Fig. 8C, Fgγ390–396A was indistinguishable from wild-type mouse and human Fg in the Efb binding assay. Consistent with these findings, the synthetic peptides covering the sequences of P1 and P2-C, respectively, do not interfere with Fg binding to Efb (Fig. 8D). These results suggest that the Efb binding site is functionally separable from the motifs in Fg critical to αMβ2 binding. These observations also suggest that the mechanism by which Efb blocks neutrophil adherence to Fg is not through a direct competitive interaction.
The inhibition of neutrophil adherence to Fg described in this study is a novel function for Efb that may be central to overall bacterial evasion of immune surveillance. Neutrophils are professional phagocytes that play a pivotal role in innate immunity and inflammatory responses by virtue of their ability to phagocytose microorganisms, produce inflammatory mediators, and elicit both intracellular and extracellular microbial killing strategies. The leukocyte-specific surface integrin, αMβ2/Mac-1, is one of several β2 integrins known to be central to both leukocyte trafficking and leukocyte activation events that support bacterial clearance. The αMβ2 integrin recognizes numerous ligands, including ICAM-1, C3bi, factor X, kininogen, and Fg (43). Several ex vivo and in vitro studies have shown that the interaction between Fg and neutrophils promotes neutrophil-endothelial transmigration (44), enhances phagocytosis (45), activates NF-κB signaling (4), promotes degranulation (46), produces chemokines and cytokines (47), and delays apoptosis (45). The biological significance of the Fg-αMβ2 interaction was substantiated in gene-targeted mice expressing Fgγ390–396A, a mutant form of Fg retaining clotting function but lacking the αMβ2 binding motif. Fgγ390–396A mice exhibit a significant diminution in inflammatory disease processes relative to control mice challenged in parallel, including an amelioration in inflammatory joint disease (5) and diminished inflammatory demyelination in the CNS and retained motor function in the context of autoimmune encephalomyelitis (6). Similarly, in a mouse model of S. aureus peritonitis, Fgγ390–396A showed an impaired bacterial clearance (7). The ability of Efb to block neutrophil adherence to Fg could have an effect on the inflammatory response similar to that seen in the Fg mutant mice.
Adhesion of neutrophils to immobilized Fg and fibrin is primarily an αMβ2-dependent process (32, 44, 48). We demonstrate here that the binding of Efb to Fg effectively blocks αMβ2 binding to Fg and strongly impedes neutrophil adhesion to Fg. However, this effect is not the consequence of a direct competition between Efb and αMβ2 for the same binding sites in Fg (Fig. 8, C and D) because (a) Efb binds equally well to WT Fg and Fgγ390–396A and (b) synthetic Fg peptides representing the αMβ2 binding site(s) do not affect the Efb/Fg interaction. Efb binding is also not dependent on the C-terminal five amino acids of the Fg γ chain (residues 407–411) that are known to support binding to the platelet integrin receptor αIIbβ3 (49). A mutant form of mouse Fg (termed FgγΔ5) that lacks these last five residues (49) is still recognized by Efb (data not shown). The mechanism by which Efb binding to Fg blocks the Fg-αMβ2 interaction remains to be defined, but two potential mechanisms explaining the observed indirect inhibition are as follows: (a) binding of Efb to Fg causes a structural rearrangement within Fg, making the αMβ2 binding site(s) inaccessible or non-functional, or (b) Efb binds in the vicinity of the αMβ2 binding site and sterically blocks the accessibility of αMβ2 to its binding site. The Efb-binding site(s) was located to the D fragment of Fg, but we have not further defined the site. Previously, the Fg Aα chain was proposed to interact with Efb using a far Western blot assay (50); however, our attempts to locate an Efb binding site to a specific Fg chain using the same methods were unsuccessful and possibly hampered by a requirement for the native fibrinogen protein fold for high affinity binding to Efb. We believe that defining the crystal structure of the D fragment in complex with the Efb peptide will best reveal the molecular details of this intriguing interaction, and we have initiated these studies.
In this study, we identified two sequences, Efb-A and Efb-O, within the Efb-N segment, that bind Fg. The two binding sites interact with Fg individually with significantly different affinity. The results from the competition ELISA assay and ITC demonstrate that Efb-O is the major binding site. This observation raises the question of whether two Fg-binding sites are both needed to fully support the biological function of Efb in vivo. Further investigation will yield better insight into this complex mechanism of binding and reveal how a microbial protein can interact with a host molecule to circumvent a host defense mechanism. Furthermore, our studies confirm that Efb-C does not bind Fg, which is consistent with our previous study (25) but is at odds with the report of Palma et al. (29).
Fg binding and aggregation induced by bacterial protein binding is a critical component of the virulence strategy used by streptococcal M protein, a potent anti-phagocytic molecule (11). Fg aggregated by M protein activates neutrophils and induces vascular leakage and lung damage (10, 51). Moreover, Fg·M protein complexes were identified in tissue lesions of patients with necrotizing fasciitis and toxic shock syndrome (10), consistent with a pathogenic significance of the Fg· bacterial protein complex in the disease process. Also, the staphylococcal Fg-binding proteins Coa (coagulase) and vWbp (von Willebrand factor-binding protein) were recently shown to act as virulence factors in a mouse model of abscess formation, and Fg was co-localized with Coa and vWbp proteins in the abscess lesions (52). The precise role of Efb in the pathogenesis of S. aureus infection and the contribution of the Efb-Fg interaction to virulence is not understood. Efb was previously shown to act as a virulence factor in a wound infection model (24); however, the mechanisms of action remain unclear. Our observation that Efb, like streptococcal M protein, can aggregate Fg raises the possibility that the two proteins use similar virulence strategies.
In conclusion, Efb is a powerful evader of human host defense systems. Efb not only blocks neutrophil recruitment to the site of infection by inhibiting C5a generation (21), it also inhibits activation of the adaptive immune system by preventing C3d from binding to CR2 (27). Here we show that Efb also effectively inhibits Fg-neutrophil interactions in vitro, and this property may serve the interest of the microbe by impeding neutrophil transmigration and/or impeding fibrin-supported neutrophil activation, resulting in an ineffective antimicrobial response in vivo.
We thank Dr. Caná L. Ross and Dr. Vannakambadi K. Ganesh for critical reading of the manuscript and Dr. Edward F. Plow and Dr. Dmitry A. Solovjov (Department of Molecular Cardiology, Cleveland Clinic) for the HEK293-Mac-1 cell line.
2The abbreviations used are: