Immune Evasion of Enterococcus faecalis by an Extracellular Gelatinase That Cleaves C3 and iC3b

This information is current as Shin Yong Park, Yong Pyo Shin, Chong Han Kim, Ho Jin of October 1, 2021. Park, Yeon Sun Seong, Byung Sam Kim, Sook Jae Seo and In Hee Lee J Immunol 2008; 181:6328-6336; ; doi: 10.4049/jimmunol.181.9.6328 http://www.jimmunol.org/content/181/9/6328 Downloaded from

References This article cites 44 articles, 13 of which you can access for free at: http://www.jimmunol.org/content/181/9/6328.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on October 1, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2008 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Immune Evasion of Enterococcus faecalis by an Extracellular Gelatinase That Cleaves C3 and iC3b1

Shin Yong Park,* Yong Pyo Shin,* Chong Han Kim,* Ho Jin Park,* Yeon Sun Seong,† Byung Sam Kim,‡ Sook Jae Seo,§ and In Hee Lee2*

Enterococcus faecalis (Ef) accounts for most cases of enterococcal bacteremia, which is one of the principal causes of nosocomial bloodstream infections (BSI). Among several virulence factors associated with the pathogenesis of Ef, an extracellular gelatinase (GelE) has been known to be the most common factor, although its virulence mechanisms, especially in association with human BSI, have yet to be demonstrated. In this study, we describe the complement resistance mechanism of Ef mediated by GelE. Using purified GelE, we determined that it cleaved the C3 occurring in human serum into a -like molecule, which was inactivated rapidly via reaction with water. This C3 convertase-like activity of GelE was shown to result in a consumption of C3 and thus inhibited the activation of the . Also, GelE was confirmed to degrade an iC3b that was deposited on the Ag surfaces without affecting the bound Downloaded from C3b. This proteolytic effect of GelE against the major complement resulted in a substantial reduction in Ef phagocytosis by human polymorphonuclear leukocytes. In addition, we verified that the action of GelE against C3, which is a central component of the complement cascade, was human specific. Taken together, it was suggested that GelE may represent a promising molecule for targeting human BSI associated with Ef. The Journal of Immunology, 2008, 181: 6328–6336.

he complement system is an essential part of the innate sumption of C3 induced by the excessive assembly of C3 conver- , and performs crucial functions in the rec- tase on microbial surfaces (3). As the major complement opsonin http://www.jimmunol.org/ T ognition and elimination of invading microorganisms in for phagocytosis by polymorphonuclear leukocytes (PMN), iC3b the tissue fluid, as well as the blood of mammals. It is activated by is generated by protease factor (f) I with the aid of its cofactor, three distinct pathways (the alternative, classical, and lectin path- which are fH in serum (4), (CR) 1 or mem- ways), which differ in terms of their modes of activation, but com- brane cofactor protein (5). CR3 on PMN recognizes iC3b as a monly result in the deposition of iC3b (inactive form of C3b) and ligand displayed on the surfaces of microorganisms (6), and this 3 the formation of the membrane attack complex (C5b-9) (MAC) receptor-ligand interaction is a prerequisite for the complement- on microbial surfaces via the activity of C3 convertase (C4b2a or mediated clearance of circulating pathogens.

C3bBb) (1, 2). In the activation of three complement pathways, Pathogenic microorganisms have developed a variety of strategies by guest on October 1, 2021 C3b is a central component, as it behaves as a recognition protein to overcome or evade the complement system. Previous studies con- that attaches covalently to microbial surfaces via its reactive thio- cerning complement evasion by microbial virulence factors will have ester. Bound C3b provides a platform for the generation of more a marked impact on the search for more efficient and specific clinical C3 convertase (C3bBb) in the alternative pathway, and C5 con- treatment for infectious disease. In recent years, several complement vertase (C4b2a3b or C3bBb3b) in all three pathways. The majority evasion strategies employed by pathogenic microbes have been elu- of bound C3b is changed to iC3b to prevent the unnecessary con- cidated, and many pathogen-encoded proteins that contribute to those strategies have also been identified as novel virulence factors (1, 7, 8). Microbial pathogens utilize two types of strategies for complement *Department of Biotechnology, Hoseo University, Asan City, Chungnam, South Ko- system inhibition. One is to resist the deposition of C3b on their sur- rea; †Department of Biochemistry, College of Medicine, Dankook University, Cheo- nan, South Korea; ‡Immunomodulation Research Center, University of Ulsan, Ulsan, faces. To do this, 1) they derange the initial recognition steps for South Korea; and §Division of Life Science, College of Natural Sciences, Gyeong- complement activation (9, 10), 2) inactivate C3 convertase, thereby sang National University, Chinju, South Korea preventing further C3b deposition on their surfaces (11), 3) modulate Received for publication March 31, 2008. Accepted for publication August 21, 2008. C3 and its split products (C3b and iC3b) (12, 13). The other strategy The costs of publication of this article were defrayed in part by the payment of page involves the blockage of the terminal step in the complement cascade, charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. namely the assembly of cytolytic molecules (C5b-9) on the microbial 1 This work was supported by a grant (2005-015-C00447) from Korea Research surfaces (14, 15). Foundation (KRF). S.Y.P. and C.H.K. received a scholarship from the World-Class Enterococci are the third most common pathogens isolated from 2030 Project of Hoseo University. S.J.S. and B.S.K. were supported by a scholarship human bloodstream infections (BSIs) (16, 17). It has been previ- (or grant) from the BK21 Program, the Ministry of Education and Human Resources Development, Korea. ously reported that up to 90% of enterococcal infections in human 2 Address correspondence and reprint requests to Dr. In Hee Lee, Department of are caused by Enterococcus faecalis (Ef) (18). Whereas the mech- Biotechnology, Hoseo University, 165 Sechuli, Baebangmyun, Asan City, Chung- anisms of their antibiotic resistance and spread have been exten- nam, South Korea. E-mail address: [email protected] sively studied, virulence mechanisms for enterococcal infections 3 Abbreviations used in this paper: MAC, membrane attack complex; PMN, poly- remain largely unknown. In our previous study (19), we purified an morphonuclear leukocyte; f, factor; CR, complement receptor; BSI, bloodstream in- fection; Ef, Enterococcus faecalis; GelE, gelatinase; NHS, normal human serum; extracellular gelatinase (GelE) (GenBank EF105504) of Ef GM C3-def HS, C3-deficient human serum; O-Ef, pre-opsonized Ef; CVF, cobra venom (GenBank EF120452), which has been described previously as an factor. important virulence factor of Ef. In addition, we demonstrated that the Copyright © 2008 by The American Association of Immunologists, Inc. 0022-1767/08/$2.00 GelE of Ef GM evidenced an identity of 99% at the nucleotide and www.jimmunol.org The Journal of Immunology 6329 Downloaded from FIGURE 1. Proteolytic effects of GelE on complement components in the human serum. A, Each reaction mixture was subjected to SDS-PAGE (upper panel) and immunoblotting analysis (lower panel). B, Proteolytic activity of GelE to each purified complement factor. One microgram of each complement factor was treated with GelE at the indicated concentrations for 30 min at room temperature and then subjected to SDS-PAGE. C, The assembly of fluid-phase C3 convertase with fB, fD, and C3bЈ generated via GelE action. Validity of C3bЈ as a component of C3 convertase was monitored by the appearance of Bb after the incubation of the mixture. Lane 1, human C3b; lane 2, human C3 treated with GelE; lane 3, mixture of C3b, fB, and fD; lane 4, mixture of C3, GelE, fB, and fD. D, Effects of fH and fI on assembled C3 convertase. Lane 1, C3 treated with assembled C3 convertase for control; lane 2, C3 treated with assembled C3 convertase in the presence of fH and fI; lane 3, mixture of C3, GelE, fB, fD, fH, and fI. E, Native SDS-PAGE analysis http://www.jimmunol.org/ to examine the fD cofactor activity of C3b and C3bЈ. Of note, whereas the band for C3 was not changed upon incubation with fB (third lane), those for C3b and C3bЈ were moved upward in the presence of fB. In A and E, symbols ϩ and Ϫ indicate addition and no addition of corresponding protein into the sample, respectively.

amino acid sequence levels with the GelE of Ef V583. It was also sence of GelE. The reaction volume was 10 ␮l of PBS and the quantity of shown that proteolytic activity of GelE affected the complement sys- each protein used in these tests was adjusted to 1 ␮g. Each mixture was incubated for 30 min at room temperature and dried down completely tem in human serum, although the mechanisms underlying its anti- ␮

via vacuum centrifugation. The sample was then resuspended in 10 lof by guest on October 1, 2021 complement effect have yet to be elucidated in detail. In the current SDS-PAGE sample buffer (0.5 M Tris-HCl (pH 6.8), containing 10% (w/v) study, we attempted to uncover the exact mechanism underlying the SDS and 40 mM DTT) and subjected to SDS-PAGE analysis. Duplicate SDS- anticomplement activity of GelE. In particular, we assessed its effects PAGE gel was employed for immunoblot assay conducted with anti-human on the complement-mediated phagocytosis of Ef by human PMNs. C3 Ab. In an effort to detect the proteolytic activity of GelE to C3, fH, fI, fB, or fD, each complement factor was incubated with increasing concen- Furthermore, we also endeavored to determine whether GelE acts tration of GelE under conditions identical to those listed above, then as- specifically on the human complement system. sessed via SDS-PAGE analyses. In addition, C3 (0.5 ␮g), fB (0.6 ␮g), and fD (4 ng) were incubated with GelE (0.5 ␮g) in 10 ␮l of PBS containing Materials and Methods 5 mM MgCl2 for 30 min at room temperature to examine whether the C3b-like molecule generated by GelE is capable of functioning as a com- GelE, Abs, and complement components C3, C3b, iC3b, factors ponent of soluble C3 convertase. For the control assembly of C3 conver- B, D, I, and H tase, commercial C3b (0.5 ␮g) was incubated with fB (0.6 ␮g) and fD (4 GelE was purified from the Ef GM culture media in accordance with the ng) under the same condition as above. The assembly of C3 convertase was verified on SDS-PAGE gel by monitoring a band of fB fragment (Bb) procedures described in our previous study (19). Anti-human C3 Ab was generated via proteolysis by fD. In an effort to assess the activity of soluble purchased from Sigma-Aldrich (Cat. No. C7761). In the case of the anti- C3 convertase against the purified C3, C3 convertase was assembled via GelE Ab, we utilized an antiserum that had been previously generated via the 30-min incubation of C3b (2 ␮ ␮ ␮ an injection of GelE into a rabbit (19). All human complement components g) with fB (2 g) and fD (0.4 g) in 50 ␮l of PBS containing 5 mM MgCl and mixed with 10 ␮l of PBS con- used in this study were commercially obtained from three companies as 2 taining C3 (10 ␮ ␮ follows: C3, Sigma-Aldrich C2910; C3b, Calbiochem 204860; iC3b, Cal- g). Alternatively, 50 l of the assembled C3 convertase was mixed with 10 ␮l of PBS containing C3 (10 ␮g), fI (2 ␮g), and fH biochem 204863; fB, Quidel 902985; fD, Sigma-Aldrich C5688; fI, Sigma- ␮ Aldrich C5938; fH, Sigma-Aldrich C5813. (4 g), then incubated for 30 min at room temperature. In addition, GelE (2 ␮g) was added to the mixture of fB (2 ␮g), fD (0.4 ␮g), fI (2 ␮g), fH Normal human serum and PMNs (4 ␮g), and C3 (10 ␮g), and then incubated for 30 min at room temperature. Thereafter, 6 ␮l of each sample was subjected to SDS-PAGE analysis. Fresh normal human serum (NHS) and PMNs were prepared from blood To examine the fD cofactor activity of C3b or C3bЈ, C3b (2 ␮g) or C3 samples that were collected from three healthy human donors after obtain- (2 ␮g) preincubated for 30 min with GelE, was incubated for 10 min with ing informed consent. PMNs were isolated from 15 ml of EDTA-antico- fB (4 ␮g) at room temperature. Then, each sample was loaded onto 10% agulated blood via three-step procedures consisting of dextran sedimenta- native SDS-PAGE gel. C3 (2 ␮g) was incubated with fB (4 ␮g) under the tion, Ficoll-Hypaque centrifugation, and hypotonic lysis of residual same condition as above and used for control. erythrocytes as previously described by Rakita et al. (20). They were then suspended in HBSS containing 1% BSA without Ca2ϩ and Mg2ϩ at ap- Immunofluorescence microscopic observation proximately 6 ϫ 106 cells/ml. Using zymosan (Sigma-Aldrich, Z4025) and C3-deficient human serum Degradation of C3 in NHS by GelE (C3-def HS) (Sigma-Aldrich, C8788), we assessed the opsonization of for- eign particles by the complement system. Sixty microliters of PBS con- Purified C3 was mixed with GelE in the presence of both fH and fI or each taining C3 (10 ␮g) was mixed with 40 ␮l of Gelatin Veronal Buffer (GVB, of two factors. Alternatively, the same reaction was conducted in the ab- Sigma-Aldrich, G6514) containing 50% (v/v) C3-def HS and 100 ␮gof 6330 COMPLEMENT EVASION OF ENTEROCOCCUS FAECALIS Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 2. Immunofluorescence analysis for an effect of GelE on C3 opsonization of zymosan. A, Zymosan incubated with C3-def HS supplemented with C3. B, Zymosan incubated with C3-def HS supplemented with C3 treated by C3 convertase. C, Zymosan incubated with C3-def HS supplemented with C3 treated by C3 convertase in the presence of fI and fH. D, Zymosan incubated with C3-def HS supplemented with C3 treated by GelE in the presence of fI and fH. For immunostaining, FITC-labeled anti-C3c Ab was diluted to a concentration of 1/25 in PBS.

zymosan. The mixture was then incubated for1hat37°C in a rotary shaker experiment, we utilized PBS containing 1 mM CaCl2 and 1 mM MgCl2. and washed twice in PBS. After centrifugation, the sediment was resus- This PMN-NHS sample was mixed with log-phase Ef (6 ϫ 105 cells) in 10 pended in 100 ␮l of PBS containing 4% (v/v) FITC-labeled C3c comple- ␮l of PBS. After 30 min of incubation at 37°C in a rotary shaker, 10 ␮lof ment Ab (GeneTex, 16262), then incubated for an additional 30 min at the mixture was removed and added to 1.49 ml of distilled water to lyse the 37°C in a rotary shaker. The sample was then washed twice in PBS and PMNs. Then, 10 ␮l of the sample was plated on tryptic soy agar (Difco). resuspended in 100 ␮l of PBS. In brief, 20 ␮l of sample was removed and The resultant colonies were counted following overnight incubation at observed under the fluorescence microscopy (Leica DMBL). To assess an 37°C. For a control, 10 ␮l of Ef solution (6 ϫ 105 cells) was incubated in effect of soluble C3 convertase assembled with C3b (2 ␮g), fB (2 ␮g), and buffer mixture (100 ␮l of HBSS plus 90 ␮l of PBS) in the absence of NHS fD (0.4 ␮g) on C3, C3 (10 ␮g) was incubated in 60 ␮l of PBS containing and PMNs under the same condition described above. According to the C3 convertase for 30 min at room temperature. The sample was then same procedure, bacteria were numerated after overnight incubation. In an treated sequentially with C3-def HS and zymosan in accordance with the effort to assess the effects of GelE on the Ef killing mediated by NHS and procedure described above, and then subjected to microscopic analysis. PMNs, 100 ␮l of PMN solution was added to 90 ␮l of PBS containing Alternatively, C3 was incubated with C3 convertase in the presence of fH various quantities of GelE (0.5, 1, or 2 ␮M) and 10 ␮l of NHS and each ␮ ␮ (4 g) and fI (2 g) under the same condition as above and the sample was sample was then subjected to the same procedure as described above. Al- subjected to the same microscopic observation procedure. To assess the ternatively, to evaluate the effects of GelE on the killing of pre-opsonized effects of GelE on C3 opsonization, C3 (10 ␮g) was incubated for 30 min ␮ ϫ 9 ␮ ␮ ␮ Ef (O-Ef) by PMN, 900 l of Ef solution (6 10 cells/ml) were first with GelE (2 g) in the presence of fH (4 g) and fI (2 g) at room incubated with 100 ␮l of NHS for 15 min at 37°C in a rotary shaker. The temperature. After further incubation in Gelatin Veronal Buffer containing reaction was stopped by the addition of 50 mM EDTA solution in PBS, zymosan and C3-def HS, the sample was subjected to microscopic after which the mixture was washed twice in PBS. After a brief centrifu- analysis. gation, the O-Ef sediment was resuspended in 1 ml of PBS. Then, 10 ␮lof ϫ 7 ␮ PMN-mediated bactericidal assay O-Ef (6 10 cells/ml) was mixed with 90 l of PBS containing varying concentration of GelE (0.5, 1 or 2 ␮M) or no GelE. Following 30 min of One hundred microliters of PMN solution (6 ϫ 106 cells/ml) prepared in incubation at 37°C in a rotary shaker, 100 ␮l of PMN solution (6 ϫ 106 HBSS were added to 90 ␮l of PBS containing 10 ␮l of NHS. In this cells/ml) in HBSS was added to the mixture. After 30 min of incubation, The Journal of Immunology 6331

each sample was subjected to the same procedure as described above to count the recovered colonies. For a 100% survival control, O-Ef was in- cubated under the same conditions in HBSS containing no PMNs. In addition, we attempted to examine whether GelE affected PMNs and damaged their phagocytic activity against O-Ef. GelE at different concen- trations (0.5, 1, or 2 ␮M) were initially incubated for 30 min with PMNs in HBSS at 37°C. After three times washing in HBSS to remove the re- maining GelE, 190 ␮l of GelE-treated PMN sample (6 ϫ 105 cells) was mixed with 10 ␮l of PBS containing O-Ef (6 ϫ 105 cells), and incubated for 30 min at 37°C. The survival colonies from each sample were numer- ated in accordance with the same procedure as described above.

Degradation of iC3b by GelE and microscopic observation for effect of GelE on CR3 (CD11b/CD18) on human PMNs It has been determined that GelE is capable of degrading purified iC3b and surface-bound iC3b, both of which were generated via the activity of fI in the presence of fH. Purified C3b (2 ␮g) was incubated for 30 min with GelE (0.5 ␮g) in 10 ␮l of PBS containing fI (1 ␮g) and fH (2 ␮g) at 37°C. For a control, C3b was treated with an identical amount of fI and fH in PBS without GelE. Each sample was subjected to SDS-PAGE and immunob- loting analyses were conducted with anti-C3 Ab. The proteolytic activity of

GelE to iC3b was also evaluated using purified iC3b. One microgram of Downloaded from iC3b was incubated with GelE of 2-fold serially diluted concentrations from 0.8 ␮M for 30 min at 37°C, then subjected to SDS-PAGE analysis. To assess the effect of GelE on surface-bound iC3b, 900 ␮l of PBS con- taining 200 ␮g of zymosan was mixed with 100 ␮l of NHS and incubated for 20 min at 37°C. After two washings in PBS, zymosan was treated with GelE (2 ␮g) for 30 min at 37°C. The zymosan sample was then extensively washed in PBS and boiled for 5 min in 50 ␮l of SDS-sample buffer to elute zymosan-bound proteins. After a brief centrifugation, the supernatant was http://www.jimmunol.org/ subjected to SDS-PAGE and duplicate gels were utilized for immunoblot analysis. The protein sample detached from zymosan that was incubated with NHS but not treated with GelE was utilized as a control. Furthermore, zymosan was incubated with NHS for different times, and each sample was treated with GelE. Then, the eluted proteins were also examined by im- munobloting analysis. To examine the effect of GelE on CR3, 100 ␮lof PMNs suspension (2 ϫ 107 cells/ml) in HBSS was mixed with 100 ␮lof FIGURE 3. Effects of GelE on PMN-mediated bacterial killing. A, All ␮ 2 M GelE in PBS and incubated for 30 min at 37°C. After washing twice elements (NHS, PMN, GelE, and Ef) were incubated together in the reac- with HBSS, PMNs were treated with FITC-conjugated anti-CD11b Ab tion mixture. B, O-Ef was treated with GelE before incubation with PMN.

(sc-52686, Santa Cruz Biotechnology) for 30 min and observed under flu- by guest on October 1, 2021 C, PMN was treated with GelE before incubation with O-Ef. In control orescence microscopy. sample, Ef bacteria (N-Ef) that were not pre-opsonized by NHS were in- cubated with PMN, after which the recovered CFU was numerated. In all Phagocytosis and flow cytometry analysis tests, PMN-mediated Ef killing was expressed as the percent of bacterial Methanol-fixed Ef was first labeled with FITC (Sigma-Aldrich F7250). survival in each sample as compared with that of the control sample, ac- FITC-labeled O-Ef was treated with 2 ␮M GelE for 30 min at 37°C and cording to the following equation: [(number of colony forming units (CFU) then incubated for 30 min with PMNs at 37°C. For a control, FITC-labeled in test sample/number of CFU in control sample) ϫ 100]. Data were ob- O-Ef untreated with 2 ␮M GelE was incubated with PMNs. According to tained from three independent experiments and expressed as the mean val- the same procedure as described above, each sample was then visualized ues Ϯ SDs (p Ͻ 0.01). via fluorescence microscopy. To assess the effects of GelE on PMNs, PMN samples were treated with 2 ␮M GelE for 30 min at 37°C and washed twice in HBSS before incubation with FITC-labeled O-Ef. The phagocytosis of O-Ef by GelE-treated PMNs was then observed in accordance with the Results same procedures as above. Alternatively, the phagocytosis of FITC-labeled Inactivation of C3 in human serum by GelE O-Ef by PMNs was assessed via flow cytometry using a FACS flow Using purified components of the human complement, the effect of cytometer (BD Biosciences) with computer-assisted evaluation of data (FACScan software). In these experiments, two times more PMNs (6 ϫ GelE on C3 was assessed via SDS-PAGE and immunoblotting 105 cells) and FITC-labeled O-Ef (1.2 ϫ 106 cells) than those employed in assays. As shown in Fig. 1A, the C3 ␣-chain vanished from the gel the microscopic experiments were used, and 20,000 PMNs of each sample when it was incubated with GelE in the presence of fI and fH. were analyzed for FACS analyses. Also, before cell counting, the samples However, without fI or fH in the reaction mixture, C3 was changed were consistently fixed with 4% paraformaldehyde. into a C3b-like molecule (C3bЈ) by GelE. In addition, C3 was not affected by fI and fH in the absence of GelE. Therefore, it was Hemolysis assay concluded that GelE acted on C3 in a manner similar to that of C3 Complement activation in sera from human and other animals was assessed convertase and generated a C3b-like protein which was degraded by measuring their hemolytic activity against RBC obtained from a rabbit. further by fI in the presence of its cofactor, fH. We also attempted ␮ ␮ ␮ ␮ Human (20 l), dog (25 l), chicken (25 l), mouse (55 l), or guinea pig to determine whether this C3bЈ can behave like an original C3b as (45 ␮l) sera were mixed with PBS to a total volume of 90 ␮l. Each serum sample was supplemented with 10 ␮l of rabbit RBC solution (109 cells/ml) a component of C3 convertase (C3bBb). When it was incubated in HBSS, then incubated for 30 min at 37°C. After 3 min centrifugation at with fB and fD, it was confirmed that an fB fragment (Bb) was 10,000 ϫ g,50␮l of the supernatant was removed and its OD was mea- generated via the action of fD with the aid of its cofactor as in the sured at 405 nm using a 96-well plate reader (Bio-Tek Instruments ELx control mixture, to which an original C3b was added (Fig. 1C). 800). Alternatively, each serum sample was treated with 2 ␮M GelE for 30 Ј min at 37°C before the addition of rabbit RBC, and the hemolysis of each This result demonstrated that C3b retained an fD cofactor activity sample was then examined according to the same procedure as described for the proteolysis of fB. The proteolytic activity of the assembled above. C3 convertase (C3b, fB, and fD) against C3 was evaluated in the 6332 COMPLEMENT EVASION OF ENTEROCOCCUS FAECALIS Downloaded from FIGURE 4. Proteolysis of human iC3b by GelE. A, Immunoblotting assay for proteolytic activity of GelE against iC3b generated from purified C3b via the action of fI and fH. B, Purified iC3b was treated with varying concentrations of GelE. The sample was then electrophoresed on 10% SDS-PAGE gel and the gel was stained with Coomassie blue. C, Immu- noblotting analysis for the proteolytic activity of GelE against C3 frag-

ments bound to the zymosan surfaces. Protein samples were subjected to http://www.jimmunol.org/ 8% SDS-PAGE analysis and duplicate gels were used for immunoblotting assay. Lane 1, purified C3b; lane 2, purified iC3b; lane 3, protein sample eluted from zymosan opsonized by NHS; lane 4, proteins eluted from zymosan after the treatment of opsonized-zymosan with 2 ␮M GelE. D, GelE (2 ␮M) was added to the reaction mixture at different postincubation times of zymosan and 20% NHS, and protein samples eluted from zymosan were subjected to immunoblotting analysis with anti-C3 Ab as described by Towbin et al. (44). In A and D, symbols ϩ and Ϫ indicate addition and no addition of corresponding protein into the reaction mixture, respectively. by guest on October 1, 2021 absence or presence of fI and fH (Fig. 1D). As a result, the C3 ␣-chain was found to be intact in the mixture containing C3 con- vertase, fI and fH, thereby indicating that C3 convertase was in- activated by fI and fH (lane 2 in Fig. 1D). In contrast, once C3 was FIGURE 5. Effect of GelE on phagocytosis of Ef by human PMNs. incubated with fB, fD and GelE in the presence of fI and fH, the Effects of GelE on phagocytosis of Ef by PMN under different conditions C3 ␣-chain disappeared but Bb was detected, thereby suggesting were assessed by FACScan flow cytometry. A, Phagocytosis of Ef by that C3bЈ could exert its fD cofactor activity before degradation by PMNs once FITC-labeled Ef and PMNs were coincubated in 5% NHS fI. In addition, the fD cofactor activity of C3bЈ was confirmed via containing GelE or no GelE. B, Phagocytosis of O-Ef by PMNs once O-Ef an SDS-PAGE analysis under non-reducing condition (Fig. 1E). was incubated with PMNs after O-Ef was treated with GelE. C, Phagocy- When GelE-treated C3 was incubated with fB, the band for C3bЈ tosis by PMNs that were treated with GelE before incubation with O-Ef. In all data, the left ones show the results of the positive controls conducted was shifted to upper position (sixth lane) as did an original C3b Ј without the addition of GelE into the sample, and the right ones show the incubated with fB (fourth lane), indicating C3b was capable of results of the test samples treated with GelE. binding to fB and acting as a cofactor of fD. In parallel, it was verified that the assembled C3 convertase did not affect C3 in NHS upon its addition to NHS but that the addition of GelE to NHS result indicates that the activity of soluble C3 convertase was abol- resulted in the degradation of the C3 ␣-chain as had been observed ished by fI and fH, and thus an intact C3 was provided to C3-def previously (data not shown). From these results, it was concluded HS. In contrast, as expected from the results shown in Fig. 1, that GelE functioned as a soluble C3 convertase which was resis- C3-def HS supplemented with GelE-treated C3 evidenced no op- tant to proteolysis by fI and fH. Using C3-deficient human serum sonization effects on zymosan (Fig. 2D). (C3-def HS) and zymosan, we also determined the effects of GelE on the opsonization of complement on zymosan surfaces (Fig. 2). Effect of GelE on PMN-mediated bactericidal activity When C3-def HS was supplemented with C3 that had been prein- When Ef was incubated at 5% NHS containing constant quantities cubated with the assembled C3 convertase, fluorescence resulting of human PMNs and different concentrations of GelE, the rate of from opsonization by C3 fragments, such as C3b and iC3b, was Ef survival was enhanced with increasing concentration of GelE in not detected on zymosan surfaces (Fig. 2B). However, C3-def HS the reaction mixture (Fig. 3A). From this result, it was concluded supplemented with C3 pre-treated with soluble C3 convertase in that GelE inhibited an opsonization by NHS and substantially at- the presence of fI and fH evidenced a significant opsonization tenuated PMN-mediated Ef killing. In the following experiments, effect (Fig. 2C) as in the case of the control sample (Fig. 2A). This we also attempted to determine whether this inhibitory effect of The Journal of Immunology 6333

FIGURE 6. Sera from human and four other animals were tested for their hemolytic activity against rabbit RBCs and each hemolysis was compared with that of each GelE-treated serum sample (A). Experiments were iterated three times on different days, and the mean val- ues were utilized for the preparation of data (p Ͻ 0.01). Downloaded from B, The proteolytic effects of GelE against human and mouse C3 were assessed via immunoblotting analyses. One microgram of purified human C3 (left panel), 10% NHS (center panel), or 10% normal mouse serum (NMS) (right panel) was treated with 2 ␮M GelE for 30 min at 37°C. Each sample was electrophoresed on 8% SDS-PAGE gels under non-reducing conditions and http://www.jimmunol.org/ each duplicate gel was probed with anti-human C3 Ab (left and center) or anti-mouse C3 Ab (right). C, Joining regions of and C3b in C3 molecules from human indicate the cleavage (ء) and other animals. Asterisks sites by C3 convertase and the action site of GelE is indicated on the sequence of human C3 by an arrow. It is worthy of note that, whereas P1 site of human C3 for action of GelE is an Asn, the putative site in other C3

proteins is an acidic amino acid, such as Asp or Glu. by guest on October 1, 2021 Sequences cited were from the database of the Na- tional Center for Biotechnology Information (human, AAA85332; chicken, NP_990736; dog, XP_533932; mouse, AAC42013; guinea pig, P12387).

GelE could be induced by its action against pre-opsonized Ef (O- concluded that GelE affected O-Ef but not PMNs, and thereby Ef) with NHS or against PMNs. Firstly, O-Ef was incubated at damaged the PMN-mediated Ef killing process. varying concentrations of GelE and then the PMN-mediated bac- tericidal effect was assessed. As a result, the higher the concen- Effects of GelE on C3 fragments bound to zymosan and CR3 on tration of GelE that was applied to O-Ef, the higher the survival PMN surfaces rates of Ef were (Fig. 3B). Secondly, before incubation with O-Ef, As it has been determined that iC3b is a predominant C3 fragment PMNs were treated with varying concentrations of GelE. As bound to microbial surfaces and performs a pivotal function as a shown in Fig. 3C, all PMN samples treated with different concen- major complement opsonin in phagocytosis by PMNs, we assessed trations of GelE evidenced an equivalent antibacterial activity the proteolytic effects of GelE on iC3b. As shown in Fig. 4A, against O-Ef as was observed with the untreated PMN. Thus, we treatment of C3b with fI and fH generated two fragments (62 and 6334 COMPLEMENT EVASION OF ENTEROCOCCUS FAECALIS

43 kDa) on SDS-PAGE gel. The appearance of two fragments is molecules of four animals were resistant to GelE-mediated prote- indicative of the generation of iC3b, which resulted from the re- olysis; consequently, the MAC was assembled successfully on rab- moval of the C3f fragment from C3b ␣-chain. Two fragments of bit RBC. Furthermore, the proteolytic effect of GelE on C3 in the iC3b ␣-chain were further degraded upon the addition of GelE mouse serum was assessed via an immunoblotting assay conducted into the mixture, a finding which differed from the result that the with an anti-mouse C3 mAb (CL7503AP: Cedarlane Laboratories) C3b ␣-chain was intact against GelE attack. To further verify it, (Fig. 6B). As a result, the band of human C3 was detected at a we directly assessed the proteolytic activity of GelE against a pu- slightly lower position upon treatment with GelE than was an un- rified iC3b. As shown in Fig. 4B, two bands of iC3b corresponding treated C3, thereby indicating that C3a was removed from C3 via to 62 and 43 kDa vanished from the gel as the concentration of degradation by GelE. In addition, it was noted that the C3 band GelE increased in the reaction mixture. In addition, we attempted was moved to a much lower position when NHS was treated with to evaluate the effects of GelE on C3 fragments bound to the zy- GelE. This result shows that C3 was changed to C3bЈ by GelE, mosan surface. The protein sample eluted from zymosan pre-op- which was subsequently cleaved into iC3b by fI and fH occurring sonized by NHS was subjected to immunoblotting analysis with in NHS, and iC3b was degraded further by GelE. In contrast, it was anti-C3 Ab (Fig. 4C). As a result, iC3b was detected principally as shown that the position of the C3 band in the mouse serum re- C3 fragments, although some C3b was also detected. However, mained unchanged after incubation with GelE. when opsonized zymosan was treated with GelE, only C3b was shown to have been eluted from zymosan. This result demon- strated that iC3b was stripped off from the zymosan surfaces, as Discussion two fragments of the iC3b ␣-chain were degraded by GelE. Al- BSIs induced by pathogenic bacteria are a principal cause of in- ternatively, GelE was added to the incubation mixture (zymosan creasing morbidity and mortality in hospitalized patients. Over the Downloaded from and NHS) at different postincubation times (0, 5, and 10 min). past three decades, worldwide epidemiologic studies with human After 10 min of further incubation, C3 fragments eluted from the clinical bacteria have shown that enterococci are one of the pre- zymosan surfaces were also assessed via an immunoblotting anal- dominant pathogens of nosocomial BSIs (17, 18). Among several ysis. As a result, both C3b and iC3b were detected in the zymosan causes of human BSIs, enterococcal bacteremia has been shown to samples not treated with GelE. In contrast, only C3b was detected result in the highest rates of patient mortality (21, 22) although in all samples, to which GelE was added (Fig. 4D). It is worthy of enterococci are known to be intrinsically not as virulent as other http://www.jimmunol.org/ note that the longer zymosan and NHS were incubated before ad- Gram-positive cocci or Gram-negative bacilli detected in human dition of GelE, the more C3b was detected in the sample eluted BSIs. Enterococcal bacteremia has been considered to evidence from zymosan, although iC3b was not detected in any of the GelE- features of poly-microbial infection, as enterococci are frequently treated samples. According to these results, it was concluded that isolated from poly-microbial flora (21, 23). Accordingly, entero- GelE degraded iC3b, but not C3b bound to the surfaces of foreign cocci have been proposed to perform a critical function in bacterial particles, and consequently paralyzed an opsonization induced by synergy and the development of diseases caused by other microbes the complement activation. In addition, the proteolytic effect of with higher virulence (24). That is to say, it is reasonable to sup- GelE against CR3 on the surfaces of human PMNs was examined pose that enterococci may contribute to an evasion of other patho- by guest on October 1, 2021 by fluorescence microscopy (data not shown). Fluorescence result- gens from the human immune system, and thus facilitating their ing from binding of FITC-conjugated anti-CD11b Ab to the PMN survival and proliferation in plasma. Nonetheless, the mechanism surfaces was consistently detected after treatment with GelE, sug- underlying the virulence or the immune evasion of enterococci that gesting that CR3 on the PMN surfaces was not affected by is associated with human BSIs remains to be clearly elucidated. Ef GelE. GelE is a member of the matrix metalloprotease family, and can hydrolyze a variety of substrates, including collagen, fibrinogen, GelE inhibits complement-mediated phagocytosis of Ef by insulin, and bioactive peptides (25). As for the virulence of GelE, human PMN a growing body of evidences has been collected in animal studies To further assess the functional role of GelE in protecting Ef (26) and etiological studies conducted with human clinical isolates against the PMN-mediated bactericidal effects, we conducted a of Ef (27–29). In more detail, it has been shown that GelE is series of experiments regarding the phagocytosis of Ef by PMNs crucial for biofilm formation (30, 31), for the pathogenesis of peri- under conditions similar to those in the PMN-mediated bacteri- apical inflammation (18) and for translocation across the human cidal assay. FACS analysis showed that the phagocytosis of Ef was intestine (32). However, the exact mode of action for the virulence severely disrupted upon the incubation of PMN with NHS con- of GelE still remains to be explained. Our findings indicated that taining GelE as compared with the control samples (Fig. 5A). In GelE inhibited the complement cascade by consuming C3 in NHS addition, when O-Ef was treated with GelE and then incubated via an excessive activation of the alternative pathway, or paralyzed with PMN, phagocytosis was reduced significantly (Fig. 5B). In the PMN-mediated killing of bacteria by clearing surface-depos- contrast, GelE-treated PMN was confirmed to maintain its activity ited iC3b from the bacterial cells. As has already been established, for the phagocytosis of O-Ef (Fig. 5C). Collectively, these results C3 is a central component of the complement-mediated defense were generally consistent with our previous data obtained from reactions against altered host cells, such as cancer cells, as well as PMN-mediated bactericidal assays (Fig. 3). against invading pathogens. It has been reported that neoplasms and pathogenic microbes have developed several strategies to spe- Effect of GelE on complement activation in sera from other cifically target C3 fragments and C3 convertase that was assem- animals bled on their surfaces (33, 34). A variety of neoplasms have been To get further information about the actual site at which GelE shown to employ C3-cleaving proteases such as matrix metallo- cleaved C3, we tested its activity against sera of animals whose C3 protease (33), serine protease (35), and cysteine proteases (36), has defined differences in sequence from human C3. Whereas the which were equipped primarily on their membrane surfaces. As a hemolytic activity of NHS was reduced significantly after GelE result of the proteolysis, the deposited C3 fragments were liberated treatment, those of sera from four animals against rabbit RBC were from the cell surface, thereby protecting neoplasms against attack not affected by GelE (Fig. 6A). This result suggested that the C3 from the host complement system. The elastase of Pseudomonas The Journal of Immunology 6335 aeruginosa has been demonstrated as a bacterial protease possess- ing C3-cleaving activity (37, 38). The protease was demonstrated to degrade the C3 ␣-chain without affecting the C3 ␤-chain, which appeared to be similar to the results obtained from our experiment, in which the purified C3 was treated with GelE (Fig. 1B). How- ever, further explanation must be provided as to whether the cleav- age of the C3 ␣-chain by an elastase is associated with comple- ment inactivation and then contributes to the immune evasion activities of P. aeruginosa. Native C3 in the human plasma is inert, and does not bind to any receptor. Once activated by C3 convertase assembled onto the Ag surfaces, C3 is transformed to C3b (39). During this activation, the protein undergoes a major conformational change, which resulted in the exposure of its internal thioester, which became liable to attack by nucleophiles within its immediate surroundings. Conse- quently, some C3b bind covalently to nearby molecules on the target surfaces on which complement activation was occurring. However, it was confirmed that this binding of C3b was not effi- cient: typically, ϳ90% of C3b molecules were unable to attach to the target surfaces, and were inactivated rapidly as the result of Downloaded from reaction with water (39). In our previous study (19), we demon- strated that GelE cleaved C3 at a position quite close to the original site for C3 convertase (see in Fig. 6D). In addition, the result of the present study showed that the C3b-like fragment generated by FIGURE 7. Schematic illustration depicting inhibitory effects of GelE on the human complement system. A, Circulating C3 was inactivated via GelE became a component of C3 convertase and was degraded Ј Ј transformation into a C3b-like fragment (C3b ) by GelE and C3b was http://www.jimmunol.org/ further by fI and fH, as in cases of an original C3b (Fig. 1). Col- cleaved into iC3b, the ␣-chain of which was degraded further by GelE. lectively, our results led to the conclusion that GelE acted on C3 Also, C3bЈ bound to fB before degradation by fI, and played a role as an in the fluid phase, as if it were a soluble C3 convertase. As a result fD cofactor in the generation of Bb. In addition, C3a was degraded by of the activity of GelE, circulating C3 was transformed continu- GelE, as reported in our previous work. B, Major complement opsonin ously into soluble C3b-like molecules, which were then immedi- (iC3b) was cleaved further by GelE and was released from the Ag surfaces, ately reacted with water and then lost their binding capacity to the thereby inhibiting phagocytosis mediated by an interaction between iC3b target surfaces. Therefore, it was suggested that GelE rendered and CR3 on PMN. AP, CP, and LP denote alternative, classical and lectin native C3 inactive, which resulted in the consumption of a central pathways, respectively. component in all complement activation routes. This activity of by guest on October 1, 2021 GelE against the human complement system was reminiscent of that of cobra venom factor (CVF), which has been well studied as tion of opsonized-microbes into PMNs, but rather facilitate the a functional C3b homologue (40, 41). Upon the introduction of conversion of C3b to iC3b on microbial surfaces (42, 43). This CVF into the human serum, it was shown to behave like a C3b conversion progressively increased the proportion of iC3b/CR3 molecule in the formation of a C3 convertase with fB in the pres- interactions, which further improved the complement-mediated ence of fD. The resultant CVF-Bb complex could convert native phagocytosis by PMNs. Accordingly, CR1 is considered to func- C3 into C3b in the fluid phase. Unlike an original C3 convertase tion as a cofactor of fI, rather than as a phagocytic receptor to that was very labile, CVF-Bb convertase proved to be rather stable, complement opsonin (42). As shown in Fig. 4, GelE degraded with a long half-life in the fluid phase (40). Moreover, whereas C3 iC3b on opsonized Ef without affecting bound C3b. Moreover, it convertase proved vulnerable to proteolysis by fI and fH, CVF-Bb was determined that bound iC3b was stripped off from the Ef sur- was confirmed to be completely resistant to them (40). Therefore, faces, and this was attributable to the cleavage of 62 kDa fragment as a consequence of its stability, CVF-Bb convertase consumed C3 containing a thioester binding domain (39) by GelE. In addition, it via the continuous activation of the alternative pathway, and oblit- was observed that the treatment of PMNs with GelE neither dis- erated the functional complement. Similarly, it was evidenced that rupted to the phagocytosis of O-Ef by PMNs, nor their capacity for Ef GelE behaved like a stable C3 convertase in the fluid phase as Ef killing. Taken together, our results indicated that GelE impaired did CVF-Bb, although it remains to be clarified whether or not an iC3b among the four members participating in opsonin-receptor GelE was expressed in a serum environment in such amounts that matches (C3b/CR1 and iC3b/CR3) between O-Ef and PMN. This led to complete depletion of C3. In addition, it was interesting to impairment of iC3b by GelE was sufficient to precipitate a signif- note that CVF-induced C3 convertase activity was detected in the icant reduction in PMN-mediated Ef killing, although it was not hemolymph of Galleria mellonella (41), from which our bacterium clear whether the C3b/CR1 pair was also affected as we have not (E. faecalis GM) was isolated (19). We now consider that the C3 specifically studied the effect of GelE on PMN surface molecules convertase activity was probably attributed to the presence of Ef other than CR3. GelE in the G. mellonella hemolymph. At the inception of this work, we surmised that GelE might Complement-mediated phagocytosis is accomplished via the influence the complement system of other animals in such a way specific recognition of bound C3 fragments (C3b and iC3b) by the that it acted on the human complement system, as it was previously corresponding receptors (CR1 and CR3) on human PMNs. It has verified that the cleavage sites in other animals’ C3 molecules by been suggested that CR1 mediates an initial transient adhesion to C3 convertase were highly homologous with that of human C3, C3b, and that CR3 mediates the stable interaction between mi- and GelE cleaved a peptide bond adjacent to the action site of C3 crobes and PMNs via specific binding to iC3b (6). It was verified convertase (Fig. 6C). However, our hemolytic assay showed that that the C3b/CR1 interaction did not contribute to the internaliza- GelE, at least, did not affect the formation of MAC resulting from 6336 COMPLEMENT EVASION OF ENTEROCOCCUS FAECALIS complement activation in the sera of the other four animals. Upon 18. Kayaoglu, G., and D. Ørstavik. 2004. Virulence factors of Enterococcus faecalis: consideration of the upstream of the complement cascade, it was relationship to endodontic disease. Crit. Rev. Oral Biol. Med. 15: 308–320. 19. Park, S. Y., K. M. Kim, J. H. Lee, S. J. Seo, and I. H. Lee. 2007. Extracellular likely that the C3b molecules of sera from other animals were gelatinase of Enterococcus faecalis destroys a defense system in insect hemo- sufficiently deposited on the rabbit RBC surfaces, and subse- lymph and human serum. Infect. Immun. 75: 1861–1869. 20. Rakita, R. M., N. N. Vanek, K. Jacques-Palaz, M. Mee, M. M. Mariscalco, quently afforded the assembly of C3/C5 convertases which re- G. M. Dunny, M. Snuggs, W. B. Van Winkle, and S. I. Simon. 1999. Entero- sulted in MAC formation. This assumption was bolstered by the coccus faecalis bearing aggregation substance is resistant to killing by human fact that the putative sites for the activity of GelE in the C3 of other neutrophils despite phagocytosis and neutrophil activation. Infect. Immun. 67: 6067–6075. animals differed from that associated with human C3. 21. Shaked, H., Y. Carmeli, D. Schwartz, and Y. Siegman-Igra. 2006. Enterococcal In summary, the results of this study demonstrate that GelE, bacteraemia: epidemiological, microbiological, clinical and prognostic character- known as a virulence factor of Ef, paralyzed the human comple- istics, and the impact of high level gentamicin resistance. Scand. J. Infect. Dis. 38: 995–1000. ment system, via a two-stage process (see Fig. 7). Firstly, it func- 22. Tacconelli, E. 2006. New strategies to identify patients harbouring antibiotic- tioned as a functional homologue of soluble C3 convertase that resistant bacteria at hospital admission. Clin. Microbiol. Infect. 12: 102–109. 23. Koch, S., M. Hufnagel, C. Theilacker, and J. Huebner. 2004. Enterococcal in- converted circulating C3 to C3b, which was instantaneously inac- fections: host response, therapeutic, and prophylactic possibilities. Vaccine 22: tivated via reaction with water. Secondly, GelE degraded two frag- 822–830. ments of the ␣-chain of iC3b, and removed it from the Ef surfaces. 24. Montravers, P., A. Andremont, L. Massias, and C. Carbon. 1994. Investigation of the potential role of Enterococcus faecalis in the pathophysiology of experimen- Consequently, the iC3b/CR3 match was not made between Ef and tal peritonitis. J. Infect. Dis. 169: 821–830. PMN, and thus PMN-mediated Ef killing was severely disrupted. 25. Ma¨kinen, P. L., D. B. Clewell, F. An, and K. K. Ma¨kinen. 1989. Purification and Finally, our findings regarding the complement evasion of Ef by substrate specificity of a strongly hydrophobic extracellular metalloendopeptidase (“gelatinase”) from Streptococcus faecalis (strain 0G1–10). J. Biol. Chem. 264: GelE make GelE a promising molecule for the targeting of infec- 3325–3334. tious diseases associated with Ef. 26. Singh, K. V., X. Qin, G. M. Weinstock, and B. E. Murray. 1998. Generation and Downloaded from testing of mutants of Enterococcus faecalis in a mouse peritonitis model. J. In- fect. Dis. 178: 1416–1420. Disclosures 27. Coque, T. M., J. E. Patterson, J. M. Steckelberg, and B. E. Murray. 1995. Inci- dence of hemolysin, gelatinase, and aggregation substance among enterococci The authors have no financial conflict of interest. isolated from patients with endocarditis and other infections and from feces of hospitalized and community-based persons. J. Infect. Dis. 171: 1223–1229. References 28. Elsner, H. A., I. Sobottka, D. Mack, M. Claussen, R. Laufs, and R. Wirth. 2000. Virulence factors of Enterococcus faecalis and Enterococcus faecium blood cul- http://www.jimmunol.org/ 1. Rooijakkers, S. H., M. Ruyken, A. Roos, M. R. Daha, J. S. Presanis, R. B. Sim, ture isolates. Eur. J. Clin. Microbiol. Infect. Dis. 19: 39–42. W. J. van Wamel, K. P. van Kessel, and J. A. van Strijp. 2005. Immune evasion 29. Kanemitsu, K., T. Nishino, H. Kunishima, N. Okamura, H. Takemura, by a staphylococcal complement inhibitor that acts on C3 convertases. Nat. Im- H. Yamamoto, and M. Kaku. 2001. Quantitative determination of gelatinase ac- munol. 6: 920–927. tivity among enterococci. J. Microbiol. Methods 47: 11–16. 2. Kozel, T. R. 1996. Activation of the complement system by pathogenic fungi. 30. Arciola, C. R., L. Baldassarri, D. Campoccia, R. Creti, V. Pirini, J. Huebner, and Clin. Microbiol. Rev. 9: 34–46. L. Montanaro. 2008. Strong biofilm production, antibiotic multi-resistance and 3. Gasque, P. 2004. Complement: a unique innate immune sensor for danger signals. high gelE expression in epidemic clones of Enterococcus faecalis from ortho- Mol. Immunol. 41: 1089–1098. paedic implant infections. Biomaterials 29: 580–586. 4. McRae, J. L., T. G. Duthy, K. M. Griggs, R. J. Ormsby, P. J. Cowan, 31. Mohamed, J. A., and B. E. Murray. 2006. Influence of the fsr locus on biofilm B. A. Cromer, W. J. McKinstry, M. W. Parker, B. F. Murphy, and D. L. Gordon. formation by Enterococcus faecalis lacking gelE. J. Med. Microbiol. 55: 2005. Human -related protein 5 has cofactor activity, inhibits C3 con- 1747–1750.

vertase activity, binds heparin and C-reactive protein, and associates with li- 32. Zeng, J., F. Teng, and B. E. Murray. 2005. Gelatinase is important for translo- by guest on October 1, 2021 poprotein. J. Immunol. 174: 6250–6256. cation of Enterococcus faecalis across polarized human enterocyte-like T84 cells. 5. van Lookeren Campagne, M., C. Wiesmann, and E. J. Brown. 2007. Infect. Immun. 73: 1606–1612. complement receptors and pathogen clearance. Cell. Microbiol. 9: 2095–2102. 33. Rozanov, D. V., A. Y. Savinov, V. S. Golubkov, T. I. Postnova, A. Remacle, 6. Nilsson, M., M. Weineisen, T. Andersson, L. Truedsson, and U. Sjo¨bring. 2005. S. Tomlinson, and A. Y. Strongin. 2004. Cellular membrane type-1 matrix met- Critical role for complement receptor 3 (CD11b/CD18), but not for Fc receptors, alloproteinase (MT1-MMP) cleaves C3b, an essential component of the comple- in killing of Streptococcus pyogenes by neutrophils in human immune serum. ment system. J. Biol. Chem. 279: 46551–46557. Eur. J. Immunol. 35: 1472–1481. 34. Jean, D., J. Hermann, F. Rodrigues-Lima, M. Barel, M. Balbo, and R. Frade. 7. Mo¨lka¨nen, T., J. Tyynela¨, J. Helin, N. Kalkkinen, and P. Kuusela. 2002. En- 1995. Identification on melanoma cells of p39, a cysteine proteinase that cleaves hanced activation of bound plasminogen on Staphylococcus aureus by staphy- C3, the third component of complement: amino-acid-sequence identities with lokinase. FEBS Lett. 24: 72–78. procathepsin L. Biochem. J. 312: 961–969. 8. Zipfel, P. F., R. Wu¨rzner, and C. Skerka. 2007. Complement evasion of patho- 35. Ollert, M. W., R. Frade, A. Fiandino, M. Panneerselvam, E. C. Petrella, M. Barel, gens: common strategies are shared by diverse organisms. Mol. Immunol. 44: M. K. Pangburn, R. Bredehorst, and C. W. Vogel. 1990. C3-cleaving membrane 3850–3857. proteinase. A new complement regulatory protein of human melanoma cells. 9. Silverman, G. J., C. S. Goodyear, and D. L. Siegel. 2005. On the mechanism of J. Immunol. 144: 3862–3867. staphylococcal protein A immunomodulation. Transfusion 45: 274–280. 36. Jean, D., M. Bar-Eli, S. Huang, K. Xie, F. Rodrigues-Lima, J. Hermann, and 10. Schenkein, H. A., H. M. Fletcher, M. Bodnar, and F. L. Macrina. 1995. Increased R. Frade. 1996. A cysteine proteinase, which cleaves human C3, the third com- opsonization of a prtH-defective mutant of Porphyromonas gingivalis W83 is ponent of complement, is involved in tumorigenicity and metastasis of human caused by reduced degradation of complement-derived . J. Immunol. 15: melanoma. Cancer Res. 56: 254–258. 5331–5337. 37. Hong, Y. Q., and B. Ghebrehiwet. 1992. Effect of Pseudomonas aeruginosa 11. Rooijakkers, S. H., M. Ruyken, J. van Roon, K. P. van Kessel, J. A. van Strijp, elastase and alkaline protease on serum complement and isolated components and W. J. van Wamel. 2006. Early expression of SCIN and CHIPS drives instant C1q and C3. Clin. Immunol. Immunopathol. 62: 133–138. immune evasion by Staphylococcus aureus. Cell. Microbiol. 8: 1282–1293. 38. Suter, S., U. B. Schaad, L. Roux, U. E. Nydegger, and F. A. Waldvogel. 1984. 12. Rooijakkers, S. H., K. P. van Kessel, and J. A. van Strijp. 2005. Staphylococcal Granulocyte neutral proteases and Pseudomonas elastase as possible causes of innate immune evasion. Trends Microbiol. 13: 596–601. airway damage in patients with cystic fibrosis. J. Infect. Dis. 149: 523–531. 13. Lee, L. Y., X. Liang, M. Ho¨o¨k, and E. L. Brown. 2004. Identification and char- 39. Janssen, B. J., and P. Gros. 2007. Structural insights into the central complement acterization of the C3 binding domain of the Staphylococcus aureus extracellular component C3. Mol. Immunol. 44: 3–10. fibrinogen-binding protein (Efb). J. Biol. Chem. 279: 50710–50716. 40. Vogel, C. W., D. C. Fritzinger, B. E. Hew, M. Thorne, and H. Bammert. 2004. 14. Langley, R., B. Wines, N. Willoughby, I. Basu, T. Proft, and J. D. Fraser. 2005. Recombinant cobra venom factor. Mol. Immunol. 41: 191–199. The staphylococcal -like protein 7 binds IgA and complement C5 41. Phipps, D., M. Menger, J. S. Chadwick, and W. P. Aston. 1987. A cobra venom and inhibits IgA-Fc ␣ RI binding and serum killing of bacteria. J. Immunol. 174: factor (CVF)-induced C3 convertase activity in the hemolymph of Galleria mel- 2926–2933. lonella. Dev. Comp. Immunol. 11: 37–46. 15. Fernie-King, B. A., D. J. Seilly, A. Davies, and P. J. Lachmann. 2002. Strepto- 42. Sutterwala, F. S., L. A. Rosenthal, and D. M. Mosser. 1996. Cooperation between coccal inhibitor of complement inhibits two additional components of the mu- CR1 (CD35) and CR3 (CD 11b/CD18) in the binding of complement-opsonized cosal innate immune system: secretory leukocyte proteinase inhibitor and ly- particles. J. Leukocyte Biol. 59: 883–890. sozyme. Infect. Immun. 70: 4908–4916. 43. Ehlers, M. R. 2000. CR3: a general purpose adhesion-recognition receptor es- 16. Dupont, H. 2007. The empiric treatment of nosocomial intra-abdominal infec- sential for innate . Microbes Infect. 2: 289–294. tions. Int. J. Infect. Dis. 11: S1–S6. 44. Towbin, H., T. Staehelin, and J. Gordon. 1992. Electrophoretic transfer of pro- 17. Karchmer, A. W. 2000. Nosocomial bloodstream infections: organisms, risk fac- teins from polyacrylamide gels to nitrocellulose sheets: procedure and some ap- tors, and implications. Clin. Infect. Dis. 31: S139–S143. plications. Biotechnology 24: 145–149.