Staphylococcus aureus Proteins Sbi and Efb Recruit Human Plasmin to Degrade Complement C3 and C3b

Upon host infection, the human pathogenic microbe Staphylococcus aureus (S. aureus) immediately faces innate immune reactions such as the activated complement system. Here, a novel innate immune evasion strategy of S. aureus is described. The staphylococcal proteins surface immunoglobulin-binding protein (Sbi) and extracellular fibrinogen-binding protein (Efb) bind C3/C3b simultaneously with plasminogen. Bound plasminogen is converted by bacterial activator staphylokinase or by host-specific urokinase-type plasminogen activator to plasmin, which in turn leads to degradation of complement C3 and C3b. Efb and to a lesser extend Sbi enhance plasmin cleavage of C3/C3b, an effect which is explained by a conformational change in C3/C3b induced by Sbi and Efb. Furthermore, bound plasmin also degrades C3a, which exerts anaphylatoxic and antimicrobial activities. Thus, S. aureus Sbi and Efb comprise platforms to recruit plasmin(ogen) together with C3 and its activation product C3b for efficient degradation of these complement components in the local microbial environment and to protect S. aureus from host innate immune reactions.


Introduction
Staphylococcus aureus, a persisting human pathogen, can cause a variety of diseases including skin infections, pneumonia, endocarditis, and sepsis [1,2]. Upon infection, S. aureus is immediately recognized and targeted by innate immune responses, such as the complement system. Activation of complement by foreign surfaces (alternative pathway, AP), by antibodies (classical pathway, CP) or by mannan (lectin pathway, LP) causes microbial opsonization, leukocyte recruitment, and cell lysis. All three pathways lead to the cleavage of C3 and subsequent formation of anaphylatoxin C3a and opsonin C3b. C3a attracts and activates granulocytes; whereas C3b attaches covalently to the bacterial surface, amplifies complement activation, and thereby labels cells for phagocytosis. Furthermore, C3b deposition leads to inflammatory reactions and formation of the pore-forming terminal complement complex [3][4][5].
S. aureus evades the complement system by targeting C3 and the activity of C3 convertases [6]. At least three C3 binding proteins that exert complement inhibitory functions have been identified in S. aureus: staphylococcal immunoglobulin-binding protein (Sbi), extracellular fibrinogen-binding molecule (Efb), and the Efb homologue protein Ehp [7][8][9]. Sbi blocks the AP by induction of C3 consumption [10,11] and Efb inhibits binding of factor B to C3b and blocks the C3 and C5 convertases [12,13]. The interaction of Sbi or Efb with the thioester-containing domain (TED) in C3 induces a conformational change in the C3 molecule [11,14]. In a tripartite complex with the human complement regulator factor H, C3 [14] or C3b [15] is degraded by factor I. Sbi binds C3 via its structural domains 3 and 4 [11,15], which harbor a three-helix bundle motif and are structurally related to the C3 binding domain in Efb [14,16].
Previous work demonstrated that S. aureus binds the human plasma protein plasminogen (PLG) to its surface and expresses a plasminogen activator, staphylokinase (SAK), that converts plasminogen to the serine protease plasmin (PL) [17,18]. However, it was unclear how S. aureus binds plasminogen to target the host C3 molecule for complement evasion. Plasmin controls several processes such as fibrinolysis, wound healing and tissue remodeling. Thus pathogenic microbes often exploit the proteolytic activity of plasmin to degrade components of the extracellular matrix (ECM), as well as fibrinogen for dissemination in the host [19,20]. Beside fibrinolytic activity, plasmin can cleave native C3 leading to the formation of the anaphylatoxin C3a [21][22][23][24] and C3b, which is subsequently degraded and inactivated to iC3b and C3c [25][26][27][28].
This report demonstrates that plasminogen, which is bound by the staphylococcal proteins Sbi and Efb, is converted to plasmin by S. aureus secreted SAK. Bound plasmin subsequently cleaves C3 and C3b that are simultaneously bound by Sbi and Efb. This proteolytic process are enhanced by the conformational changes Efb and presumably Sbi induce in C3 and C3b. In addition, plasmin was shown to degrade C3a that harbors chemoattractant and antimicrobial activities. Thus, recruitment and activation of plasminogen by S. aureus is highly coordinated to maximize complement inhibition at the level of C3.

Bacterial strains
S. aureus strain Newman was cultured on blood agar (Invitrogen) or in Luria broth at 37uC.

Enzyme linked Immunosorbent Assay (ELISA)
Bacterial proteins were immobilized (equimolar) onto a microtiter plate (Maxisorb, Nunc) blocked with 0.4% gelatin in DPBS for 2 h at 37uC, PLG b was added for 1.5 h at 37uC, and bound proteins were detected with Streptavidin-Peroxidase For dose-dependent binding of PLG to Sbi 3-4/Efb-C, 5 mg/ml Sbi 3-4 or Efb-C were immobilized to a microtiter plate, blocked with Blocking Solution I (AppliChem) for 2 h at 37uC and 25-800 nM PLG were added for 2 h at 37uC. Bound proteins were Bacterial proteins (equimolar) were immobilized and PLG b was applied and detected using streptavidin-HRP. Plasminogen binding to Sbi and Efb was comparable to plasminogen binding to the borrelial protein CRASP-5. Plasminogen did not bind to HSA or the plate (buffer). Data represent mean values 6 standard deviations from three independent experiments. (C) Plasminogen was recruited to Sbi and Efb. Non-labeled plasminogen was applied to immobilized bacterial proteins. Bound proteins were eluted, separated by SDS-PAGE and plasminogen (92 kDa) was detected by Western blot analysis. CRASP-5 bound plasminogen in contrast to HSA and buffer. A representative experiment out of three is shown. doi:10.1371/journal.pone.0047638.g001 detected with anti-plasminogen (Acris Antibodies; 1:1000) in Cross Down Buffer (AppliChem) and anti-goat HRP (1:2000). In competition assays Sbi 3-4 or Efb-C were immobilized and incubated with concentrations of C3 (200 nM) and plasminogen (200 for Efb-C and 400 nM for Sbi 3-4) that saturate binding sites. C3 binding was detected using polyclonal anti-C3-antibody (CompTech; 1:1000) and plasminogen binding with polyclonal plasminogen-antiserum.

Combined ELISA-Western blot analysis (CEWA)
Combined ELISA-Western blot analysis was performed according to Haupt et al [15]. Briefly, equimolar amounts of Sbi, Efb, CRASP-5, and HSA were immobilized onto a microtiter plate. After blocking with gelatin, 50 nM plasma purified plasminogen was added and incubated overnight at 4uC. Bound plasminogen was eluted using SDS buffer, separated by SDS-PAGE, and analyzed by Western blotting using plasminogen antiserum (1:2000). Borrelia burgdorferi CRASP-5 protein was included as a positive control and HSA (Cleveland, Ohio) as a negative control.

Surface plasmon resonance (SPR)
Binding of plasminogen to Sbi 3-4 or Efb-C was analyzed by surface plasmon resonance (Biacore 3000, AB) at 25uC in 150 mM PBS. Sbi 3-4 or Efb-C was immobilized on a CMD 500 M sensor chip (Xantec) by standard amino coupling chemistry following the manufacturer's protocol. Plasminogen (400 nM) was injected at a constant flow rate of 5 ml/min.

Antimicrobial assay
C3a (1 mM) (CompTech) was incubated with S. aureus (3610 4 ) in 10 mM Tris-HCL with 5 mM Glucose (pH 7.4) with or without 5 mg PLG and/or 1 mg SAK at 37uC for 2 h with agitation. Bacteria were diluted 1:5, 20 ml were plated onto a LB plate in     4) and C3/C3b cleavage products appeared (marked with arrows). Bacterial proteins were immobilized (equimolar) and plasminogen was added together with C3 (A) or C3b (B). The activator SAK was applied and C3/C3b cleavage was followed by Western blot analysis using anti-C3-HRP (Fab). The mobility of the a (a) and b chains of C3/C3b and the cleavage products are indicated by arrows. CRASP-5 and HSA had no effect on cleavage (lane 7 and 8). Data shown are representative of three independent experiments. (C) Proposed schematic model of the C3 cleavage products generated by plasmin modified after [25,32]. The C3 dg cleavage product is not recognized by the C3-antibodies used in this study. doi:10.1371/journal.pone.0047638.g004 Roti-Load, separated by SDS-PAGE and analyzed by Western blotting using anti-C3-HRP. Bacteria surface-bound C3b was measured by flow cytometry with FITC-labeled anti-C3-Fab (Protos Immunoresearch).

Statistical analysis
Significant differences between two groups were analyzed by the unpaired Student's t-test. Values of *p,0.05, **p,0.01, ***p,0.001 were considered as statistically significant.

Plasminogen binds to Sbi and Efb
Sbi and Efb are C3 and C3b binding proteins. In order to investigate whether Sbi and Efb also recruit plasminogen, binding of plasminogen to recombinant Sbi and Efb was tested. The accuracy of recombinant Sbi and Efb was confirmed as both molecules showed binding to C3 and in addition Sbi bound to IgG (Fig. 1). Immobilized Sbi or Efb was incubated with biotin-labeled plasminogen (PLG b ) and binding was assayed by ELISA. Plasminogen bound to both staphylococcal proteins Sbi and Efb. Plasminogen binding to Sbi and Efb was similar to the previously identified plasminogen ligand CRASP-5 of Borrelia burgdorferi [30,31], (Fig. 1A+B). To exclude non-specific binding of plasminogen via its biotin-label, binding of unlabeled plasminogen to Sbi or Efb was tested using a combined ELISA-Western blot assay. Plasminogen was added to immobilized Sbi or Efb, washed, and all surface bound proteins were separated by SDS PAGE and immunoblotted for the detection of plasminogen. Also free plasminogen bound to Sbi, Efb, and CRASP-5 (Fig. 1C).

Characterization of the PLG:Sbi and PLG:Efb interactions
To determine the Sbi and Efb domains responsible for interaction with plasminogen, plasminogen binding to the structurally related domains Sbi 3-4 and Efb-C in Sbi and Efb, respectively, was assayed. Plasminogen added to the immobilized domain fragments bound to both Sbi 3-4 ( Fig. 2A) and Efb-C (Fig. 2B) in a dose-dependent manner. These binding interactions were also followed in real time by using surface plasmon resonance. Plasminogen added in fluid phase bound to immobilized Sbi 3-4 or Efb-C, confirming the interaction between plasminogen and these specific Sbi and Efb domains. To investigate whether plasminogen binding to Sbi and Efb occurs via lysine residues, eACA, a lysine analog which blocks lysine residues, was used in binding assays. eACA interfered with plasminogen binding to both Sbi 3-4 and Efb-C, demonstrating that lysine residues are involved in the interactions of these bacterial proteins with plasminogen.
To determine whether plasminogen and C3 bind simultaneously to Sbi and Efb, each plasma protein was added to immobilized Sbi 3-4 or Efb-C using concentrations of plasminogen (400 nM for Sbi 3-4 and 200 nM for EfbC) which saturate all binding sites. Plasminogen and C3 binding were measured in parallel. C3 binding to Sbi 3-4 and Efb-C, as well as plasminogen binding to Sbi 3-4 remained unchanged. Plasminogen binding to Efb-C was slightly reduced in the presence of C3 (Fig. 2E and 2F). The data indicate that plasminogen and C3 bind simultaneously to both staphylococcal molecules.

Plasminogen bound to Sbi or Efb is activated to plasmin
To exert proteolytic activity, plasminogen needs conversion to plasmin by an activator. To proof whether plasminogen bound to Sbi or Efb is converted to plasmin. Plasminogen was bound to immobilized Sbi or Efb and treated with the human activator uPa. Plasmin generation was followed by cleavage of a plasmin-specific chromogenic substrate. Plasminogen-bound Sbi or Efb was activated by uPa to plasmin, as seen by the increased conversion of the substrate S-2251. Similarly, plasminogen bound to CRASP-5 was activated to the protease plasmin (Fig. 3A).
SAK is known to activate plasminogen via a non-proteolytic mechanism. To determine whether SAK is also able to activate Sbi-or Efb-bound plasminogen SAK was used in similar experiments instead of uPa. Also SAK activated plasminogen bound to Sbi or Efb, and the resulting plasmin cleaved the chromogenic substrate ( Fig. 3B and C). In summary, plasminogen bound to Sbi and Efb is activated by both staphylococcal and human activators to form active plasmin.

Plasmin bound to Sbi or Efb degrades C3 and C3b
After plasmin(ogen) was shown to form complexes with C3 and Sbi or Efb, the plasmin activity was assayed to cleave complexed C3. Plasminogen was bound together with C3 to immobilized Sbi or Efb and activated by the addition of SAK. C3 degradation was investigated by separating the reaction mixture by SDS-PAGE and Western blot analysis using anti-C3 Fab'-fragments. Multiple C3-degradation products with mobilities of 115, 87, 68, 40, and 27 kDa were generated indicating that Sbi-or Efb-bound plasminogen was activated by SAK to plasmin which subsequently cleaved bound C3 (Fig. 4A). In a similar assay performed with C3b, cleavage products with mobilities of 87, 68, 40, and 27 kDa appeared and demonstrated that complexed plasmin also degraded C3b (Fig. 4B). When CRASP-5 or HSA was used in these assays, instead of Sbi or Efb, no cleavage of C3 was observed which is explained by the fact that CRASP-5 and HSA do not acquire plasmin together with C3/C3b. Thus, plasmin complexed together with C3/C3b and Sbi or Efb degrades the complement proteins C3 and C3b.

C3 degradation by plasmin is enhanced by Sbi and Efb
Upon binding, Efb changes the structural conformation of C3, leading to an increased susceptibility of C3 to degradation by trypsin [14]. To analyze whether C3 degradation by plasmin is also modulated by Sbi or Efb, C3 proteolysis by plasmin was compared in the presence or absence of Sbi or Efb. C3 degradation by plasmin was enhanced by both staphylococcal proteins, as demonstrated by the appearance of additional C3 cleavage products in Western blot analysis using anti-C3 Fab'fragments for protein detection (Fig. 5A). In the presence of Sbi, C3 cleavage products with mobilities of 114 kDa (a'), 68 kDa, and weakly 40 kDa appeared and with Efb, degradation products with mobilities of 68, 40, and 27 kDa were detected. By contrast, proteolytic cleavage of C3 by plasmin was not affected after addition of CRASP-5 or HSA. To localize the Sbi and Efb domains responsible for the degradation-enhancing activities, fragments Sbi1-2, Sbi 3-4, and Efb-C were investigated in the same assay. C3 cleavage by plasmin was enhanced in the presence of Sbi 3-4 and Efb-C, but no enhancement was observed in the presence of Sbi 1-2 (Fig. 5B). To confirm these results, C3 proteolysis was also examined using ELISA. C3 was immobilized on a microtiter plate and plasmin (generated by uPa or SAK) was added together with the bacterial proteins or with HSA. After incubation, the plate was washed and the amount of bound C3 was determined. Smaller amounts of C3 were detected in reactions containing plasmin than that of those containing plasminogen. The plasmin-dependent reduction of C3 deposition was enhanced by the addition of Sbi 3-4 and especially Efb-C to the reaction mixture, but not by the addition of CRASP-5 or HSA (Fig. 5C). This enhancement by Sbi 3-4 or Efb-C was more clearly demonstrated by using an assay in which increasing amounts of the bacterial proteins were used (Fig. 5D). The results indicate that plasmin cleaves C3, and that the staphylococcal proteins Sbi or Efb enhance the cleavage by plasmin. The two domains responsible for the degradation-enhancing effect were fragments Sbi 3-4 and Efb-C, each of which contains a three-helix bundle motif.

C3b degradation by plasmin is also enhanced by Sbi and Efb
Having shown that plasmin cleaves C3 more efficiently in the presence of Sbi or Efb, the same effect was analyzed for C3b. Plasmin-mediated C3b degradation was assayed using the deletion fragments Sbi 1-2, Sbi 3-4, or Efb-C, in the same assay as described for C3. The C3b degradation by plasmin was enhanced in the presence of Sbi 3-4 or Efb-C, as shown by the increase in C3b degradation products (Fig. 6A). However, the addition of Sbi 1-2 showed no enhanced degradative effect. To investigate whether Sbi and Efb also enhance plasmin activity on the bacterial surface, S. aureus was incubated with human serum to allow complement-mediated C3b deposition. The C3b-coated bacteria were then washed, incubated with plasmin plus Sbi 3-4 or Efb-C, and C3b deposition was analyzed by flow cytometry. Plasmin substantially reduced the C3b opsonization of S. aureus by about 37%. In the presence of Sbi 3-4, C3b deposition was further decreased to 50%; and in the presence of Efb-C, to 38% (Fig. 6B). In parallel analyses, supernatants containing the C3b degradation products were assessed by SDS-PAGE and Western blotting. C3b cleavage products with mobilities of 41 and 27 kDa were identified in the supernatants of those samples containing C3b-opsonized S. aureus treated with plasmin (Fig. 6C). Again C3b degradation was enhanced when plasmin cleaved surface-bound C3b in the presence of Sbi 3-4 and especially Efb-C (indicated by arrows), and the presence of Sbi 1-2 showed no effect. These results demonstrate that C3b degradation by plasmin is accelerated by interaction with the staphylococcal proteins Sbi and Efb on the bacterial surface.

Plasmin degrades C3a
Plasmin cleaves C3 and generates C3a [23,24] and degrades C3b and thereby inactivates the C3b molecule for complement C3b amplification [25,28,32]. As C3a exhibits antimicrobial activity [33,34] and thus can kill S. aureus, we investigated whether S. aureus recruited plasmin also degrades C3a. Therefore, S. aureus was treated with C3a alone or with C3a together with plasmin activated by SAK. The supernatants were analyzed by SDS- aureus to 85% and plasminogen alone to 50%. In parallel S. aureus treated with C3a and or plasmin(ogen) was cultivated overnight on LB agar plates. CFUs were counted and the survival without C3a was set to 100% (white columns). Data represent mean values 6 standard deviations from three independent experiments. ***p,0.001. doi:10.1371/journal.pone.0047638.g007 PAGE and Western blotting using C3a antisera. Plasmin completely degraded C3a, as demonstrated by the absence of detectable levels of C3a (Fig. 7A). When plasminogen was added without SAK to S. aureus, the amount of C3a was decreased, which is explained by synthesis and secretion of SAK by S. aureus during incubation time and subsequent activation of plasminogen to plasmin. In parallel, C3a antimicrobial activity against S. aureus was analyzed in survival assays. C3a added to growing S. aureus killed the bacteria, but in the presence of C3a and plasmin (PLG+SAK) 85% of S. aureus survived. Plasminogen added without SAK resulted in 50% survival of the bacteria, and addition of SAK without plasminogen had no effect on the bacteria (Fig. 7B). Thus, we conclude that plasmin inhibits the bactericidal activity of C3a by degradation of the C3a molecule.

Discussion
Presented here is a novel complement evasion strategy employed by S. aureus. Human plasminogen and C3/C3b are simultaneously bound by the microbial proteins Sbi and Efb. Recruited plasminogen remains accessible for the human activator uPa or the bacterial activator SAK for conversion to active plasmin. Plasmin bound to Sbi and Efb degrades and inactivates C3 and C3b. This degradation is enhanced by conformational changes exerted by Efb and to a lesser extent by Sbi on both, bound-C3 and bound-C3b. Moreover, we show that plasmin degrades C3a, and thus inactivates the antimicrobial activity of C3a. Consequently, Sbi and Efb-recruited plasmin inhibits complement cascade progression, opsonization, antimicrobial activities, and inflammatory reactions.
The importance of the complement system in defense against S. aureus infections is reflected by the multiple strategies this human pathogen has developed to circumvent the host complement attack. Thereby, S. aureus preferentially targets the central complement component C3 to inhibit the auto-amplification loop of the alternative pathway [7,35]. One of these strategies is the expression of small proteins that bind the C3 convertase, like SCIN [36,37], or C3/C3b such as Efb [38,39] and Sbi [11,15]. Additionally, S. aureus recruits complement regulators from human plasma like factor H that acts as cofactor for the serine protease factor I to cleave C3b [15]. Here, we show that plasminogen is recruited by Sbi and Efb. To date, three staphylococcal plasminogen binding proteins have been described: inosine 59monophosphate, a-enolase, and ribonucleotide reductase [17]. However, in contrast to these plasminogen ligands, Sbi and Efb bind C3/C3b in addition to plasminogen. Plasminogen-binding is mediated by the domains Sbi 3-4 in Sbi and Efb-C in Efb. Although the sequence identity between Sbi 3-4 and Efb-C is rather low (only 19%), both fragments contain a three-helix bundle motif [16], in which amino acids of the a2-helixes contribute to C3d binding [10,12]. Plasminogen and C3 bind simultaneously to Sbi 3-4 or Efb-C and are located closely to each other, which results in accelerated degradation of C3/C3b, once plasminogen is converted to plasmin. The C3/C3b degradationenhancing activity of Efb-C is more pronounced as compared to Sbi 3-4. This difference is likely based on the higher binding affinity of C3d to Efb than Sbi (K D [Efb-C:C3d] = 0.8 nM, K D [Sbi 3-4:C3d] = 1.4 mM) [11,14].
Plasmin cleaves C3b at several sites and generates C3c and C3d [25][26][27][28]. The interactions of plasminogen with both staphylococcal proteins Sbi or Efb depend on lysine residues. This is the same method of plasminogen binding observed in other microbial proteins such as PE from H. influenzae [27], CRASP-1 to -5 from B. burgdoferi [30,31], Pra1 [40] and Gpm1 from C. albicans [41], and indicates a similar binding strategy, which ensures activation of plasminogen to plasmin for tissue evasion and complement escape.
Complexed with C3 and Sbi or Efb, plasminogen remains accessible for the staphylococcal activator SAK and the human activator uPa to be converted to plasmin. More than 67% of S. aureus strains express the sak gene and produce the non-proteolytic staphylokinase [42]. In addition, the pathogen can enhance uPa production in mammalian cells [43] or activate conversion of pro-uPa to uPa by metalloprotease aureolysin [44] to generate plasmin. Plasmin-mediated C3b degradation is accelerated by Sbi or Efb. These functions are mediated by the C3 binding domains Sbi 3-4 and Efb-C, and concur with studies showing that Efb-C-binding to C3 and C3b leads to a conformational change in C3 which enhances proteolytic cleavage of C3 by trypsin and factor I [14]. Thus, a similar effect of Efb-C on C3 for plasmin degradation is anticipated.
Plasmin has also been described as a complement activator, because it cleaves C3 to C3a and C3b [21][22][23][24]. However, as shown here and in previous studies, plasmin further degrades C3b and demonstrates anti-opsonic as well as complement-inhibitory activity [25][26][27][28]. Here, we show for the first time that purified C3a is also degraded by plasmin, and that in such cases, antimicrobial activities against S. aureus are decreased. These results are in accordance with those of Amara et al, who observed that while plasmin generates C3a, the C3a amount decreased when plasmin concentrations increased [24]. Thus, binding of plasmin by S. aureus helps the pathogen inactivate host antimicrobial activities and obviously represents a further survival strategy.
The powerful inhibitory effect of acquired plasmin(ogen) against several immune defense mechanisms apparently explains why many pathogenic microbes attach human plasminogen to their surface and activate plasminogen to plasmin. S. aureus is remarkably effective, using several plasmin activation pathways and expressing at least five plasminogen binding proteins. Among these proteins Sbi and Efb act like an amplifier of plasminmediated C3 cleavage to reduce local complement activity.