THE JOURNAL OF BIOLOGICAL CHEMISTRY 2004 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 279, No. 49, Issue of December 3, pp. 50710 50716, 2004 Printed in U.S.A.
Identification and Characterization of the C3 Binding Domain of the Staphylococcus aureus Extracellular Fibrinogen-binding Protein (Efb)*
Received for publication, July 28, 2004, and in revised form, August 16, 2004 Published, JBC Papers in Press, August 26, 2004, DOI 10.1074/jbc.M408570200
Lawrence Y. L. Lee, Xiaowen Liang, Magnus Hook, and Eric L. Brown
From The Center for Extracellular Matrix Biology, Texas A&M University System Health Science Center, Albert B. Alkek Institute of Biosciences and Technology, Houston, Texas, 77030-7552
The secreted Staphylococcus aureus extracellular fibrinogen-binding protein (Efb) is a virulence factor that binds to both the complement component C3b and fibrinogen. Our laboratory previously reported that by binding to C3b, Efb inhibited complement activation and blocked opsonophagocytosis. We have now located the Efb binding domain in C3b to the C3d fragment and determined a disassociation constant (Kd) of 0.24 M for the Efb-C3d binding using intrinsic fluorescence quenching assays. Using truncated, recombinant forms of Efb, we also demonstrate that the C3b binding region of Efb is located within the C terminus, in contrast to the fibrinogen binding domains that are located at the N-terminal end of the protein. Enzyme-linked immunosorbent assay-type binding assays demonstrated that recombinant Efb could bind to both C3b and fibrinogen simultaneously, forming a trimolecular complex and that the C-terminal region of Efb could inhibit complement activity in vitro. In addition, secondary structure analysis using circular dichroism spectroscopy revealed that the C-terminal, C3b binding region of Efb is composed primarily of -helices, suggesting that this domain of Efb represents a novel type of C3b-binding protein.
Staphylococcus aureus is an important human pathogen that causes a wide range of infections and is the most common bacterium associated with septic arthritis. The disease often results in severe and persistent joint damage, and mortality rates from this type of infection are high (1, 2). Furthermore, S. aureus remains one of the most common causes of infections in the industrialized world, and since the 1980s S. aureus has become the most common nosocomial pathogen due in part to an increasing resistance to multiple antibiotics and to the recent worldwide emergence of highly virulent strains (1, 311). In addition, S. aureus infections can result in a variety of diseases, including skin infections, endocarditis, arthritis, osteomyelitis, or sepsis and is a reflection of the capacity of this organism to colonize a variety of different tissues and of their ability to circumvent a variety of immune surveillance systems resulting in the persistence of S. aureus in different environments within the host organism (1, 1214).

* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Present address and to whom correspondence and reprint requests should be addressed: Wyle Laboratories Bioastronautics Team, Human Adaptation and Countermeasures, Wyle/HAC/W1, 1290 Hercules Dr., 120, Office 408, Houston, TX 77058. Tel.: 281-212-1421; Fax: 281-2121316; E-mail: ebrown@wylehou.com.
The ability of S. aureus to affect such diverse tissues and cause persistent infections is in part related to their ability to manipulate or evade multiple defense mechanisms (1, 1217). Pathogens associated with persistent infections often produce multiple microbial immunomodulatory molecules (12, 17) to counter the immune systems of the host (i.e. adaptive and innate immunity) (12, 17), and this strategy results in the effective avoidance or delay of the pathogen clearance while concomitantly affecting the development of a protective memory response. Well characterized members of the S. aureus microbial immunomodulatory molecule family of proteins are the super antigens and protein A, which can affect T cell and antibody responses, respectively (1). Recent additions to this rapidly growing protein family are the secreted extracellular fibrinogen-binding protein (Efb)1 (18 23), which we identified as an inhibitor of various complement-mediated processes (17), the major histocompatibility complex class II analog protein (Map), which interferes with T cell function and neutrophil migration (12, 24), and the chemotaxis inhibitory protein from S. aureus (CHIPS) (13, 25), which interferes with the inflammatory response early in the infection process by binding to the C5a- and formylated peptide receptors. One of the central components of innate immunity is the complement system. Activation of the complement pathways (i.e. classical, mannose binding lectin, or alternative pathways) can lead to opsonization and phagocytosis of invading pathogens, and in a majority of cases, complement mediates lysis and cell death (26). Furthermore, complement component byproducts (e.g. C3a and C5a) can serve as potent chemoattractants for numerous inflammatory cells (27). A central component to all three pathways is the complement protein C3 (26, 28). C3 is not only critical for complement pathway activation, but it also plays a pivotal role in the interface between innate and acquired immune responses (28). Patients with a C3 deficiency have a significantly impaired complement system and have an increased susceptibility to infections from a variety of organisms (29 31). Many successful human pathogens have evolved either direct C3 inactivation strategies by producing C3-binding proteins (e.g. S. aureus, Trypanosoma cruzi, and Pseudomonas aeruginosa) (17, 3234) or indirect mechanisms by targeting and manipulating the host endogenous C3-inhibitor, Factor H (or Factor H-like proteins) (e.g. Borrelia burgdorferi, Candida albicans, and Streptococcus pyogenes) (3538). A parallel strategy employed primarily by intracellular pathogens involves membrane proteins with C3 binding capabilities that

1 The abbreviations used are: Efb, extracellular fibrinogen-binding protein; rEfb, recombinant Efb; PVDF, polyvinylidene difluoride; ELISA, enzyme-linked immunosorbent assay; PBS, phosphate-buffered saline; AP, alkaline phosphatase; RBC, red blood cell; SCR, short consensus repeat.
Downloaded from www.jbc.org by guest, on June 5, 2011
This paper is available on line at http://www.jbc.org
The C3 Binding Domain of the S. aureus Efb
facilitate cellular internalization via the complement receptors (CR1 or CR2) (e.g. Mycobacterium tuberculosis, Leishmania sp., Streptococcus pneumoniae, Legionella pneumophila, Chlamydia tracomatis, and herpes simplex 1) (39 49). To date most work describing the interactions between S. aureus and complement components has primarily focused on the role of the capsule as a means of preventing C3 deposition and C3-mediated opsonization of S. aureus (50 52). However, many virulent S. aureus strains do not form a capsule (53, 54), suggesting that other complement-inhibitory strategies are likely. A recent report from our laboratory demonstrated that the S. aureus protein Efb binds to C3b and inhibits both complement-mediated lysis and opsonophagocytosis via a mechanism that inhibits C3b deposition onto activator surfaces (17). Efb is a secreted, constitutively expressed protein that also binds fibrinogen. In addition, the efb gene was present in all S. aureus isolates examined but not identified in any other staphylococcal species (21). Investigations into the roles of Efb in vivo have resulted in various hypotheses describing functions for Efb and include roles for Efb in potentiating S. aureus survival by delaying wound healing and by inhibiting platelet aggregation via its interactions with fibrinogen and the platelet receptor GPIIb/IIIa (23, 55). Furthermore, these studies suggest that the fibrinogen binding domains were located at the N terminus of Efb and consist of two 22-amino acid repeats with homology to the fibrinogen binding domains of coagulase from S. aureus (20). A third fibrinogen binding domain was also reported to reside in the C-terminal end of the protein; however, the fibrinogen binding activity of this site was less well defined since binding was dependent on whether fibrinogen was soluble or plate-bound (22, 57, 58). Data presented in this report suggest that only the N-terminal half of Efb contains fibrinogen binding activity and that this domain is not involved in C3b binding. To define the C3b binding region(s), we engineered four recombinant forms of Efb that spanned the N- and C-terminal regions, respectively, and examined them in various binding assays for C3b and fibrinogen. Data presented in this report describe the C3 binding region in Efb to be located in the C terminus and that Efb binds specifically to the C3d fragment of C3 with high affinity. Furthermore, secondary structure analysis of recombinant Efb (rEfb) using circular dichroism spectroscopy suggested that the C3 binding region is primarily composed of -helices and may represent a novel C3 binding motif.

EXPERIMENTAL PROCEDURES

Cloning of and Expression of the Efb Truncations from S. aureus Strain NewmanThe efb gene and efb truncations were amplified by PCR using S. aureus strain Newman DNA as a template. The following oligonucleotide primers were used for efb, efb104, efb120, and efb165, respectively (IDT Inc., Coralville, IA): 5-CGC GGA TCC CCA AGA GAA AAG AAA CCA GTG AGT A-3 (forward primer) and 5-AAC TGC AGA GTT TTA TTT AAC TAA TCC TTG-3 (reverse primer); 5-CCAGCAGCGAAAACTGATGCAA-3 (forward primer) and 5-AAGTTTTATTTAACTAATCCTTG-3 (reverse primer); 5-CGC GGA TCC CCA AGA GAA AAG AAA CCA GTG AGT A-3 (forward primer) and 5-AAC TGC AGT TAT TCT CTC ACA AGA TTT TGA GCT TG-3 (reverse primer); 5-CCA GCA GCG AAA ACT GAT GCA ACT-3 (forward primer) and 5-AAC TGC AGA GTT TTA TTT AAC TAA TCC TTG-3 (reverse primer). The resulting PCR products were subsequently cloned using the TA expression kit into the pCRT7/NT-TOPO expression vector (Invitrogen) and designated pCRT7/NT-rEfb, pCRT7/NT-rEfb104, pCRT7/NT-rEfb120, and pCRT7/NT-rEfb165. Nucleotide sequencing of efb, efb104, efb120, and efb165 was performed by automated sequencing (Molecular Genetics Core Facility, University of Texas-Houston Medical School). The recombinant proteins rEfb (17), rEfb104, rEfb120, and rEfb165 were expressed as recombinant N-terminal His-tagged proteins that allowed for purification using metal ion-chelating chromatography as described previously (12, 59). Proteins were expressed and purified as
previously described (59, 60). Protein concentrations were determined by UV spectroscopy, and proteins were stored at 20 C until use. Western Blot AnalysisRecombinant proteins C3b, iC3b, C3c, or C3d (Advanced Research Technologies, San Diego, CA) (4 g each) were subjected to SDS-PAGE and examined by staining with 0.05% Coomassie Brilliant Blue or electrotransferred onto a 0.45-m Immobilon-P PVDF (polyvinylidene fluoride) membrane (Millipore, Bedford, MA) as described previously (17). Membranes subjected to Western blot analysis were blocked overnight at 4 C in 5% nonfat dry milk in TBST (0.15 M NaCl, 20 mM Tris-HCl, 0.05% Tween 20 (Sigma-Aldrich), pH 7.4) and probed accordingly and then developed with 10 ml of 1-StepTM nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate solution (Pierce). All incubations were performed in 15 ml of 1% TBST for 1 h with shaking at room temperature, and membranes were washed in TBST between all steps. Labeling with digoxigenin was performed as described previously according to the manufacturers instructions (17). ELISA-type Binding AssaysImmulon-1B microtiter plate wells (Dynatech Laboratories, Chantilly, VA) were coated with either 0.25 g of C3b or fibrinogen in 100 l of PBS overnight at 4 C. The plates were washed and blocked with 200 l of Super Block (Pierce) for 1 h. Recombinant proteins (nM in 100 l final volume/well) were added to the wells and incubated for 1 h. In the next step 100 l of anti-His antibodies (Amersham Biosciences) (1:5000) were added and incubated for 1 h followed by 100 l of goat anti-mouse alkaline phosphatase (AP)-conjugated antibodies (1:5000). Alternatively, for digoxigenin-labeled proteins, 100 l of AP-conjugated anti-digoxigenin antibodies ((Fab fragment) Roche Diagnostics) (1:5000) were added. Next, 100 l of a 1 mg/ml Sigma 104 phosphatase substrate (Sigma) dissolved in 1 M diethanolamine, 0.5 mM MgCl2, pH 9.8, was added, and the plates were allowed to develop for 1 h. Plates were read at 405 nm using a microplate reader (Molecular Devices, Menlo Park, CA). Plates were washed between all steps with PBS, 0.05% Tween 20, and all incubations took place at 37 C. All dilutions were made using Super Block unless otherwise specified. Inhibition ELISAImmulon-1B microtiter plate wells (Dynatech Laboratories) were coated with either 0.25 g of C3b or C3d (Advanced Research Technologies) in 50 l of PBS overnight at 4 C. The plates were washed and blocked with 200 l of Super Block (Pierce) for 1 h. Increasing concentrations of either rEfb or the control protein decorinbinding protein A (DbpA) from B. burgdorferi (nM in 50 l final volume/well) were added to C3b- or C3d-coated wells and incubated for 1 h followed by the addition of 50 l of human Factor H (Advanced Research Technologies) (1 or 5 g) and allowed to incubate for an additional hour. A 1:1000 dilution of goat-anti human Factor H was subsequently added (100 l) and incubated for 1 h followed by 100 l of rabbit anti-goat alkaline phosphatase (AP)-conjugated antibodies (1: 5000). Plates were washed between all steps with PBS, 0.05% Tween 20 unless otherwise indicated, and all incubations took place at 37 C. All dilutions were made using Super Block unless otherwise specified. Complement Activity AssaysThe EZ Complement CH50 clinical diagnostic assay kit (Diamedix, Miami, FL) was used to evaluate the effects of the rEfb truncations on classical complement pathway activation and used as described by the manufacturer. Briefly, human serum (5 l of complement reference serum) was incubated in the presence of 5 g of each recombinant protein at a final volume of 20 l at 37 C for 1 h before a 1-h incubation at room temperature with antibody-coated sheep red blood cells (RBCs) (3 ml). After incubating, the RBCs were centrifuged (1800 rpm for 10 min), and the absorbance of the supernatants (150 l) was measured at 405 nm using a microplate reader as described above to determine the percent lysis of each sample. The data are expressed as percent lysis of the standard reference serum, and the values were derived using the equation, absorbance of sample/absorbance of reference serum CH50 value of reference (Diamedix) (17). Circular Dichroism (CD)rEfb, rEfb104, rEfb120, and rEfb165 were dialyzed twice in 2 liters of 10 mM potassium phosphate buffer, pH 7.4, and their respective concentrations were determined using UV spectroscopy. CD spectra were obtained using a Jasco J-720 spectropolarimeter (Jasco, Inc., Easton, MD) calibrated with 10 mM potassium phosphate buffer, pH 7.4, using a round 0.2-mm quartz cell. The CD spectra were collected at a scan speed of 20 nm/min at 1-nm intervals with a 1-s response and a bandwidth of 1 nm. CD data were collected across a wavelength range of nm and repeated through 20 iterations. The rEfb CD spectra were normalized by subtracting the CD spectra of buffer only, presented as units of mean residue ellipticity MRW, and calculated by using the equation, MRW 100 MRW/c d, where is the ellipticity in millidegrees, MRW is the mean residue weight of the recombinant protein, c is the concentration of the protein in mg/ml,

FIG. 1. rEfb binds to the C3d fragment of C3. C3b (lane 1), iC3b (lane 2), C3c (lane 3), C3d (lane 4) were subjected to SDS-PAGE and Coomassie-stained (A) or transferred to PVDF membranes and analyzed by Western ligand blot by probing with 10 g of rEfb (B). rEfb was detected using anti-His antibodies (1:15,000) followed by a secondary antibody conjugated to AP (1:15,000) and then developed. MW, molecular weight. and d is the cell path length in centimeters. Prediction of the secondary structure was obtained by combining data derived from three deconvolution programs (mean and S.E. were determined): self-consistent (Selcon) method, neural network (NN), and CONTIN (Sofsec version 1.2, Softwood Co., Brooksfield, CT) as described previously (61, 62). Fluorescence MeasurementsThe intrinsic tryptophan fluorescence of 1.3 M C3d in PBS was examined using a LS50B spectrofluorimeter (PerkinElmer Life Sciences) at 25 C. The excitation wavelength was set at 295 nm (5-nm slit width) while monitoring emissions from nm (4-nm slit width). Quenching of tryptophan fluorescence after the addition of rEfb (from 60 nM to 3.5 M) was analyzed according to the modified plot of Stern-Volmer (63) the F0/(F0 F) ratio, where F0 and F are the fluorescence intensities at 335 nm in the absence or presence of rEfb, respectively. The F0/(F0 F) ratio determined for each concentration of rEfb was plotted against the reciprocal of the rEfb concentrations, which yielded a straight line whose x intercept equaled the value of the association constant (Ka) for rEfb. The disassociation constant (Kd) for rEfb was determined by the slope of the straight line divided by the value of the y intercept (63).

RESULTS

FIG. 2. Circular dichroism spectroscopy. The resulting spectra (with the buffer-only spectra subtracted) for rEfb (thin line), rEfb104 (thin dashed line), rEfb120 (think line), rEfb165 (thick dashed line) are the results of 20 iterations.
rEfb Binds to the C3d Fragment of C3Identifying the binding site(s) in C3b for rEfb was performed by testing the binding of rEfb to various C3b -chain fragments generated during the process of complement activation. During this process C3 undergoes multiple cleavage events that generate various fragments which include C3a, C3b, iC3b, C3c, C3d, C3dg, and C3g (64). To determine which C3 fragment bound to rEfb, the C3 protein fragments C3b, iC3b, C3c, and C3d were subjected to SDS-PAGE and Coomassie-stained (Fig. 1A) or transferred onto a PVDF membrane and probed with rEfb (Fig. 1B). Western ligand blot analysis demonstrated that rEfb bound to C3d and all C3d-containing fragments. C3c, which does not contain the C3d region, did not bind rEfb (Fig. 1B). The additional -chain fragment of 58 kDa (Fig. 1, lane 3) is a byproduct of the trypsin digestion used to generate C3c from C3b and does not contain the C3d region. Blots probed with secondary antibody or with secondary antibody only revealed no color change after the addition of substrate (data not shown). These experiments defined C3d as the minimal C3 degradation product recognized by rEfb. Secondary Structural Analysis of rEfbThe C3d region contains binding sites for two complement regulatory proteins of the Factor H family, CR2 and Factor H. Factor H is a soluble 150-kDa multifunctional plasma protein that is the central activation inhibitor of the alternative complement pathway (65), and members of this protein family (66) are composed exclusively of globular protein domains termed short consensus repeats (SCRs) (67). Although this family of proteins differs in their number of SCRs, the individual SCR domains poses a high degree of identity to each other, which explains why these proteins share many similar functions (65). Because Factor H

contains 3 different C3 binding regions (SCR1 4, which binds to intact C3b; SCR6 10, which binds to C3c; SCR16 20, which binds to C3d), and each domain has been defined as a sixstranded anti-parallel -sheet structure (65, 68), we attempted to determine, using CD spectroscopy, if rEfb shared any structural similarity to Factor H. The secondary structure of rEfb was analyzed using CD spectroscopy and the PHD (Profile network prediction Heidelberg) program (69, 70). Both analyses revealed that rEfb was largely helical (Fig. 2 and Table I) and composed of two domains, neither of which had any structural similarity to the SCR motif. To identify the region(s) of rEfb containing the -helices, two constructs (rEfb104 and rEfb120) spanning the N-terminal half of rEfb and one C-terminal construct (rEfb165) (Fig. 3) were generated and examined in a similar manner (Fig. 2 and Table I). There is a 25-amino acid overlap between the C terminus of rEfb120 and the N terminus of rEfb165 (Fig. 3). The results of the CD analyses and the secondary structure predictions for rEfb and the rEfb constructs are shown in Table I. The CD analysis data suggested that rEfb contains an -helical C-terminal region and an unordered Nterminal domain (which contains the two fibrinogen binding repeats) (19) since rEfb165 contains almost the same percent of -helix as rEfb. Conversely, rEfb104 is highly unordered with some -sheet (Table I). The CD data collected for rEfb120 suggested that the 25 amino acids at its C terminus (shared with rEfb165) are responsible for some of the -helix observed in this construct. The predicted secondary structure predictions mirror the CD in that this analysis also indicated that the -helices were located at the C terminus (rEfb165) with little or no structure at the N terminus (rEfb104) (Table I). The differences in the nature of the predicted structure and the structural data collected by CD spectroscopy for some of the rEfb constructs suggest that secondary structure prediction programs should be used cautiously since some disparity exists between the structural data collected by CD and the predicted structures for some of the rEfb constructs. The C Terminus of rEfb Binds to C3b but Not Fibrinogen Determining the location of the C3b binding region in rEfb was initially examined by Western ligand blot analysis. rEfb, rEfb120, and rEfb165 were subjected to SDS-PAGE and Coomassie-stained (Fig. 4A) or transferred onto a PVDF membrane and probed with either digoxigenin-labeled C3b (Fig. 4B) or digoxigenin-labeled fibrinogen (Fig. 4C). Western blots probed with C3b demonstrated that rEfb and rEfb165 bound C3b but
TABLE I Secondary structural analysis of rEfb and rEfb constructs Secondary structure values obtained by CD analysis of rEfb and rEfb constructs are indicated in bold, and the predicted secondary structure values are in parentheses. The sequences submitted for the secondary structure predictions did not contain the leader (Efb, Efb104, and Efb120) or the His-tag sequences (Fig. 3).

Efb construct % -Helix % -Sheet % Other kDaa
rEfb rEfb(104) rEfb(120) rEfb(165)
(40.77) 15 2.0 (0.0) 25 0.76 (19.54) 36 1.2 (69.57)
16 1.1 (13.08) 35 3.1 (22.86) 21 4.8 (18.39) 15 4.4 (8.70)
47 4.1 (46.15) 48 1.8 (77.14) 54 5.6 (62.07) (21.74)
kDa values for the respective proteins do not include signal or His-tag sequences.
FIG. 3. Putative structural organization of Efb. S, signal sequence; R, N-terminal region of the protein containing two homologous repeat regions known to bind fibrinogen; U, undefined region. rEfb, full-length Efb protein without the signal sequence. rEfb104, contains only the R region. rEfb120 contains the R region and 24 amino acids of the U region, and rEfb165 contains the U-region. All recombinant proteins are expressed as N-terminal His6-tag proteins.
that rEfb120 did not (Fig. 4B). Conversely, Western blots probed with fibrinogen demonstrated binding to rEfb and rEfb120 but not to rEfb165 (Fig. 4C). Blots probed with an AP-labeled secondary antibody alone revealed no color change after the addition of substrate (data not shown). rEfb104 bound to fibrinogen but did not bind C3b (data not shown). Similar results were obtained using ELISA-type binding assays. rEfb and rEfb165 bound to plate-bound C3b and rEfb, rEfb104 (data not show), and rEfb120 bound plate-bound fibrinogen in a dose-dependent and saturable manner (Fig. 5, A and B). rEfb120 and not rEfb104 was used in these experiments and in all subsequent studies. The C-terminal Domain of rEfb Inhibits Complement ActivationWe previously reported that rEfb interfered with both classical and alternative pathways of complement activation (17). To determine whether the rEfb165-C3b interaction could also affect complement activity, we examined the effect of rEfb165 in a functional assay measuring classical pathway activity. This assay utilizes RBC lyses as a read-out for complement activity. Data presented in Fig. 6 demonstrated that both full-length rEfb and rEfb165 inhibited complement-mediated RBC lysis to the same degree; in contrast, rEfb120 had no inhibitory activity. rEfb Binds to C3b and Fibrinogen SimultaneouslyBecause the binding regions for C3b and fibrinogen are located on different regions of rEfb, we investigated the possibility that rEfb could bind to both C3b and fibrinogen simultaneously. A sandwich-type ELISA was performed where fibrinogen was immobilized on microtiter wells and probed first with either rEfb, rEfb120, or rEfb165 followed by a subsequent incubation with digoxigenin-labeled C3b (Fig. 7A). These results demonstrated that rEfb but not rEfb120 or rEfb165 could bind to C3b and fibrinogen simultaneously (Fig. 7A). As a control, anti-His antibodies were used to detect binding of the recombinant proteins to fibrinogen (Fig. 7B). The anti-His antibodies recognized the fibrinogen-bound rEfb and rEfb120; however, rEfb165, which does not bind fibrinogen, was not detected (Fig. 7B and Fig. 4C). Determination of the KdIntrinsic tryptophan fluorescent

quenching of rEfb was used to determine the dissociation constant (Kd) for the rEfb-C3b interaction. The conformational changes that occur when a protein binds to its ligand result in measurable fluctuations in the intrinsic fluorescence of aromatic amino acids (e.g. tryptophan, tyrosine, and phenylalanine) present in the protein (or ligand) of interest. These three aromatic residues have distinct absorption and emission wavelengths that allow the fluorescence spectrum of a protein to be defined. Tryptophan has a much stronger fluorescence and higher quantum yield that the other two aromatic amino acids, and it is, therefore, possible to use fluorescence quenching (of tryptophan) to evaluate binding kinetics. This approach was facilitated by the knowledge that the 35-kDa C3d fragment (which contains 5 tryptophan residues compared with zero tryptophan residues in rEfb) contained the rEfb-binding site(s). The intrinsic fluorescence of C3d was examined in the presence of increasing concentrations of rEfb, and the intrinsic fluorescence of C3d in the absence of rEfb is represented as the relative fluorescence intensity equal to 100. rEfb at various concentrations (60 nM3.5 M) quenched the intrinsic fluorescence of C3d in a dose-dependent manner (Fig. 8), and the modified Stern-Volmer plot (71) was applied to the fluorescence data to determine the Kd. A dose-dependent reduction in quenching was not observed when a recombinant control Histag protein, DbpA (decorin-binding protein) (59), was used (data not shown). The data plotted produced a linear slope allowing for the determination of the Kd that was calculated to be 0.24 M (Fig. 8).

DISCUSSION

Depending on the pathogen and the nature of the infection (local versus systemic), a series of events resulting in pathogen survival need to occur. In the case of S. aureus and many other bacteria, gaining access to and adhering to host extracellular matrix while concomitantly avoiding the immune system are critical first steps in the infection process followed by dissemination to other tissues. S. aureus has developed various strategies for avoiding host immunity. Super antigens and protein A can affect cellular and humoral responses, respectively (1), and the major histocompatibility complex analog protein (Map) can negatively affect both T cell function and neutrophil migration (12, 14). Interestingly, both Map and Efb were originally described as MSCRAMMS (microbial surface components recognizing adhesive matrix molecules) (72), a family of adhesins that mediate adhesion to host extracellular matrix components (18 20, 24), emphasizing the importance of both parameters in establishing an infection. In the case of Efb, which inactivates complement-mediated lysis and phagocytosis (by binding to C3b) (17) and to delay clot formation and wound healing via a mechanism that involves fibrinogen binding (23, 73), further illustrate the pleiotropic properties of many of these molecules. In this study we have identified the region of rEfb associated with C3b binding to be within the C terminus, in contrast to the fibrinogen binding domains that reside within the N-terminal half of the protein. Western ligand blot and ELISA-type bind-

FIG. 4. The C-terminal domain of rEfb binds to C3b. rEfb (lane 1), rEfb120 (lane 2), and rEfb165 (lane 3) were subjected to SDS-PAGE and Coomassie-stained (A) or transferred to PVDF for Western ligand blot analysis and probed with 5 g of digoxigenin-labeled C3b (B) or digoxigenin-labeled fibrinogen (C). Bound proteins were detected using AP-conjugated anti-digoxigenin Fab fragments (1:15,000) and then developed.
FIG. 5. rEfb constructs binding to plate-bound C3b or fibrinogen. Microtiter wells were coated with 0.25 g of C3b (A) or fibrinogen (B). rEfb (open squares), rEfb120 (open circles), or rEfb165 (closed circles) were plated at various concentrations to determine their binding specificity for C3b or fibrinogen. The rEfb proteins were detected using anti-His antibodies (1:5000) followed by a secondary antibody conjugated to AP (1:5000) and then developed. The data are expressed as the mean absorbance (405 nm) S.E. of the mean of triplicate samples.
FIG. 6. The C-terminal rEfb165 construct inhibits complementmediated RBC lysis. The effect of rEfb, rEfb120, or rEfb165 on complement activation was examined using an assay to measure the classical complement activation pathway. The data are expressed as percent RBC lysis of the complement standard reference serum.
ing assays demonstrated that the C-terminal construct rEfb165 bound exclusively to C3b but not to fibrinogen; conversely, the N-terminal constructs rEfb104 and rEfb120 bound exclusively to fibrinogen but not to C3b. This finding is somewhat contradictory to previous data that suggested that Efb contained a third fibrinogen-binding site in the C-terminal region. DNA sequence analysis of the cloned refb165 used in this study was consistent with sequence data identified in NCBI nucleotide data base (Q08691) and did not contain any
mismatches or mutations. One explanation for this difference in fibrinogen binding is that the previous studies examined synthetic Efb peptides for fibrinogen binding in addition to using soluble fibrinogen as a probe in ELISA-type binding assays (22). Earlier work from the same laboratory demonstrated that binding of rEfb to fibrinogen occurred only to solid phase but not soluble fibrinogen (20), suggesting that the conformation of fibrinogen may be important for Efb binding; however, in our hands neither solid or soluble phase fibrinogen (data not shown) bound to rEfb165 in ELISA-type binding assays. Defining the rEfb-binding site to the C3d fragment was significant because this fragment contains important binding sites for both factor H and CR2 (64). CR2 is expressed on B cells, T cells, and follicular dendritic cells and is implicated in the regulation of B and T cell responses (74, 75). If rEfb-bound C3d can block binding to CR1/CR2; it may also serve to negatively affect B cell, T cell, and/or antigen processing functions by diminishing the potent adjuvant properties of antigenlinked C3d (76). Although no similarity between the predicted rEfb structure and the -sheet-rich Factor H was observed, we examined the possibility that rEfb may bind to the Factor H-binding site in C3d since Factor H also inhibits complement activity by binding to various domains in C3. ELISA-type binding assays demonstrated that rEfb could not inhibit Factor H binding to plate-bound C3d or C3b (data not shown), suggesting that rEfb does not bind to the Factor H-binding site(s). Whether rEfb can block the binding between CR2 and C3b/C3d is not known and will be the subject for future work. However, since rEfb can also inhibit classical pathway activation (17) and Factor H is an inhibitor of the alternative pathway (65), it is

FIG. 7. rEfb simultaneously binds to C3b and fibrinogen. Microtiter wells were coated with 0.25 g of fibrinogen and then incubated with 1 g of either rEfb, rEfrb120, or rEfb165 and then probed with 0.25 g of digoxigenin-labeled C3b. A, digoxigenin-labeled C3b was detected using AP-conjugated antidigoxigenin (Fab fragment) (1:5000) and then developed. B, rEfb bound to fibrinogen was detected using anti-His antibodies (1:15,000) followed by an anti-mouse AP-conjugated secondary antibody and then developed. The data are expressed as the mean absorbance (405 nm) S.E. of the mean of triplicate wells.
FIG. 8. Fluorescence-quenching analysis. C3d alone represents a relative fluorescence intensity of 100. Increasing concentrations of rEfb (60 nM--3.5 M) were added, and the relative fluorescence intensity was determined (A) and then expressed as percent change in fluorescence (B). C, representative modified Stern-Volmer plot of the data acquired from the fluorescence quenching assay was used to determine the Kd value of 0.24 M. TABLE II Secondary structure predictions of bacterial C3-binding proteins Secondary structure prediction values are indicated (in parentheses). SA, surface-associated protein; CRP, complement regulatory protein; CbpA, choline-binding protein A; HbhA, heparin binding hemagglutinin; MOMP, major outer membrane protein; LPG, lipophosphoglycan; gp63, glycoprotein 63; gC, complement-interacting glycoprotein C; AP, alkaline phosphatase; E/I, extracellular/intracellular organism.
Organism Protein % -Helix % -Sheet % Other SA kDaa E/I Reference
T. cruzi S. pneumoniae M. tuberculosis L. pneumophila C. tracomatis Leishmania major Leishmania mexicana Herpes simplex 1 P. aeruginosa
CRP CbpA HbhA MOMP MOMP LPG gp63 gC AP
(7.48) (37.10) (87.94) (38.94) (21.99) (40.54) (28.95) (1.76) (32.77)
(21.93) (12.67) (1.51) (26.55) (25.09) (12.06) (16.87) (33.27) (11.97)
(70.59) (50.23) (10.51) (34.51) (52.92) (47.40) (54.18) (64.97) (55.25)
No Yes Yes Yes Yes Yes Yes Yes No

E I I I I I I I E

33, 44, 40, 43, 32, 78
kDa values obtained were from the selected reference(s).
not surprising that rEfb could not bind to the Factor H-binding sites and compete for binding with an endogenous inhibitor of complement activation. Comparing Efb to various other C3binding proteins from numerous pathogens (Table II) revealed that the majority of these proteins are surface-associated and from intracellular pathogens (exceptions being T. cruzi and P. aeruginosa) (32, 34, 77) (Table II), and all except P. aeruginosa (77) utilize these proteins primarily as a means of gaining entry into cells via CR1/2 (Table II). In addition, these proteins range in size, and their predicted secondary structures vary significantly. Even though proteins such as CbpA, MOMP, and LPG (Table II) have a similar predicted secondary structure to that of rEfb, the specific C3-binding sites in these proteins have not yet been defined, and it would, therefore, be unwise to speculate as to the nature of the secondary structure(s) involved with C3 binding for these and the other C3-binding

proteins listed in Table II. No amino acid sequence similarity between Efb and any of the proteins listed in Table II was identified. The data presented in this report suggest that the -helixrich C3b binding domain of Efb may represent a novel structural motif involved in complement regulation. Identification of the C3 binding region in rEfb and characterization of the complement inhibitory properties of this region may lead to the development of clinical therapeutics against complement-mediated disorders (56) and to the development of novel strategies to prevent or treat S. aureus infections.

REFERENCES

1. Lowy, F. D. (1998) N. Engl. J. Med. 339, 2. Kissane, J. M. (1997) in Pathology of Infectious Diseases (Connor, D. H., Chandler, F. W., Schwartz, H. J., Manz, H. J., and Lack, E. E., eds) Vol. I, pp. 805 816, Appleton and Lange, Samford, CT
39. Garnier, T., Eiglmeier, K., Camus, J. C., Medina, N., Mansoor, H., Pryor, M., Duthoy, S., Grondin, S., Lacroix, C., Monsempe, C., Simon, S., Harris, B., Atkin, R., Doggett, J., Mayes, R., Keating, L., Wheeler, P. R., Parkhill, J., Barrell, B. G., Cole, S. T., Gordon, S. V., and Hewinson, R. G. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 78777882 40. Hall, R. T., Strugnell, T., Wu, X., Devine, D. V., and Stiver, H. G. (1993) Infect. Immun. 61, 41. High, A. S., Torosian, S. D., and Rodgers, F. G. (1993) J. Gen. Microbiol. 139, 17151721 42. Lubinski, J. M., Wang, L., Soulika, A. M., Burger, R., Wetsel, R. A., Colten, H., Cohen, G. H., Eisenberg, R. J., Lambris, J. D., and Friedman, H. M. (1998) J. Virol. 72, 43. Medina-Acosta, E., Karess, R. E., and Russell, D. G. (1993) Mol. Biochem. Parasitol. 57, 44. Muller-Ortiz, S. L., Wanger, A. R., and Norris, S. J. (2001) Infect. Immun. 69, 45. Russell, D. G. (1987) Eur. J. Biochem. 164, 213221 46. Rosenow, C., Ryan, P., Weiser, J. N., Johnson, S., Fontan, P., Ortqvist, A., and Masure, H. R. (1997) Mol. Microbiol. 25, 47. Stevens, M. P., Tabrizi, S. N., Muller, R., Krause, V., and Garland, S. M. (2004) J. Clin. Microbiol. 42, 25012507 48. Schlesinger, L. S., and Horwitz, M. A. (1990) J. Clin. Investig. 85, 49. Dobson, D. E., Scholtes, L. D., Valdez, K. E., Sullivan, D. R., Mengeling, B. J., Cilmi, S., Turco, S. J., and Beverley, S. M. (2003) J. Biol. Chem. 278, 1552315531 50. Cunnion, K. M., Zhang, H. M., and Frank, M. M. (2003) Infect. Immun. 71, 51. Cunnion, K. M., Lee, J. C., and Frank, M. M. (2001) Infect. Immun. 69, 52. Wilkinson, B. J., Sisson, S. P., Kim, Y., and Peterson, P. K. (1979) Infect. Immun. 26, 53. Sopolinsky, D., Samra, Z., Karakawa, W. W., Vann, W. F., Schneerson, R., and Malik, Z. (1985) J. Clin. Microbiol. 22, 54. Nawas, T., Hawwari, A., Hendrix, E., Hebden, J., Edelman, R., Martin, M., Campbell, W., Naso, R., and Fattom, A. I. (1998) J. Clin. Microbiol. 36, 55. Heilmann, C., Herrmann, M., Kehrel, B. E., and Peters, G. (2002) J. Infect. Dis. 186, 3239 56. Kirschfink, M. (2001) Immunol. Rev. 180, 177189 57. Shannon, O., and Flock, J. I. (2004) Thromb. Haemostasis 91, 58. Wade, D., Palma, M., Lofving-Arvholm, I., Sallberg, M., Silberring, J., and Flock, J. I. (1998) Biochem. Biophys. Res. Commun. 248, 59. Guo, B. P., Brown, E. L., Dorward, D. W., Rosenberg, L. C., and Hook, M. (1998) Mol. Microbiol. 30, 711723 60. Maniatis, T., Fritsch, E. F., and Sambrook, J. (1989) Molecular Cloning: A Laboratory Manual, P.A.3, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY 61. Sreerama, N., and Woody, R. W. (1993) Anal. Biochem. 209, 62. Pancoska, P., Blazek, M., and Keiderling, T. A. (1992) Biochemistry 31, 63. Eftink, M. R., and Ghiron, C. A. (1981) Anal. Biochem. 114, 64. Sahu, A., and Lambris, J. D. (2001) Immunol. Rev. 180, 65. Zipfel, P. F., Skerka, C., Hellwage, J., Jokiranta, S. T., Meri, S., Brade, V., Kraiczy, P., Noris, M., and Remuzzi, G. (2002) Biochem. Soc. Trans. 30, 971978 66. Male, D. A., Ormsby, R. J., Ranganathan, S., Giannakis, E., and Gordon, D. L. (2000) Mol. Immunol. 37, 4152 67. Perkins, S. J., and Goodship, T. H. (2002) J. Mol. Biol. 316, 217224 68. Perkins, S. J., Gilbert, H. E., Aslam, M., Hannan, J., Holers, V. M., and Goodship, T. H. (2002) Biochem. Soc Trans 30, 69. Rost, B. (1997) Proteins, Vol. 29, Suppl. 1, 192197 70. Rost, B. (1996) Methods Enzymol. 266, 525539 71. Hall, A. V., Solanski, K. K., Vinjamuri, S., Britton, K. E., and Das, S. S. (1998) J. Clin. Pathol. 51, 215219 72. Patti, J. M., Allen, B. L., McGavin, M. J., and Hook, M. (1994) Annu. Rev. Microbiol. 48, 73. Palma, M., Nozohoor, S., Schennings, T., Heimdahl, A., and Flock, J. I. (1996) Infect. Immun. 64, 74. Boackle, S. A. (2003) Biomed. Pharmacother. 57, 75. Boackle, S. A., and Holers, V. M. (2003) Curr. Dir. Autoimmun. 6, 76. Haas, K. M., Toapanta, F. R., Oliver, J. A., Poe, J. C., Weis, J. H., Karp, D. R., Bower, J. F., Ross, T. M., and Tedder, T. F. (2004) J. Immunol. 172, 58335837 77. Hong, Y. Q., and Ghebrehiwet, B. (1992) Clin. Immunol. Immunopathol. 62, 133138