Family and Plasminogen That Binds Three Members of the Factor H Surface-Exposed Immune Evasion Protein Is a Pseudomonas Aerugino

Home , Factor H

Dihydrolipoamide Dehydrogenase of Pseudomonas aeruginosa Is a Surface-Exposed Immune Evasion Protein That Binds Three Members of the Factor H This information is current as Family and Plasminogen of October 4, 2021. Teresia Hallström, Matthias Mörgelin, Diana Barthel, Marina Raguse, Anja Kunert, Ralf Hoffmann, Christine Skerka and Peter F. Zipfel

J Immunol 2012; 189:4939-4950; Prepublished online 15 Downloaded from October 2012; doi: 10.4049/jimmunol.1200386 http://www.jimmunol.org/content/189/10/4939 http://www.jimmunol.org/ Supplementary http://www.jimmunol.org/content/suppl/2012/10/15/jimmunol.120038 Material 6.DC1 References This article cites 54 articles, 19 of which you can access for free at: http://www.jimmunol.org/content/189/10/4939.full#ref-list-1

Why The JI? Submit online. by guest on October 4, 2021

• 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

*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 © 2012 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Dihydrolipoamide Dehydrogenase of Pseudomonas aeruginosa Is a Surface-Exposed Immune Evasion Protein That Binds Three Members of the Factor H Family and Plasminogen

Teresia Hallstro¨m,* Matthias Mo¨rgelin,† Diana Barthel,* Marina Raguse,* Anja Kunert,* Ralf Hoffmann,‡ Christine Skerka,* and Peter F. Zipfel*,x

The opportunistic human pathogen Pseudomonas aeruginosa causes a wide range of diseases. To cross host innate immune barriers, P. aeruginosa has developed efficient strategies to escape host complement attack. In this study, we identify the 57- kDa dihydrolipoamide dehydrogenase (Lpd) as a surface-exposed protein of P. aeruginosa that binds the four human plasma proteins, Factor H, Factor H-like protein-1 (FHL-1), complement Factor H-related protein 1 (CFHR1), and plasminogen. Factor H contacts Lpd via short consensus repeats 7 and 18–20. Factor H, FHL-1, and plasminogen when bound to Lpd were functionally active. Factor H and FHL-1 displayed complement-regulatory activity, and bound plasminogen, when converted to the active Downloaded from protease plasmin, cleaved the chromogenic substrate S-2251 and the natural substrate fibrinogen. The lpd of P. aeruginosa is a rather conserved gene; a total of 22 synonymous and 3 nonsynonymous mutations was identified in the lpd gene of the 5 laboratory strains and 13 clinical isolates. Lpd is surface exposed and contributes to survival of P. aeruginosa in human serum. Bacterial survival was reduced when Lpd was blocked on the surface prior to challenge with human serum. Similarly, bacterial survival was reduced up to 84% when the bacteria was challenged with complement active serum depleted of Factor H, FHL-1, and CFHR1, demonstrating a protective role of the attached human regulators from complement attack. In summary, Lpd is http://www.jimmunol.org/ a novel surface-exposed virulence factor of P. aeruginosa that binds Factor H, FHL-1, CFHR1, and plasminogen, and the Lpd- attached regulators are relevant for innate immune escape and most likely contribute to tissue invasion. The Journal of Immu- nology, 2012, 189: 4939–4950.

seudomonas aeruginosa is an opportunistic human path- Upon infection, P. aeruginosa is immediately confronted and ogen that can reside as a harmless commensal, but can also attacked by the complement system, which forms the first and P cause severe diseases, including pneumonia and septicemia central layer of host innate immune defense. Three major pathways (1). P. ae ru gi nos a is a major cause of hospital-acquired infections, activate the complement cascade, and each pathway initiates a series by guest on October 4, 2021 particularly in immunocompromised individuals. In cystic fibrosis of tightly regulated reactions that all merge on the level of C3 and patients, this Gram-negative bacterium is a major cause of chronic generate C3 convertases. Each C3 convertase cleaves C3 to the in- lung infections (2–4). P. aeruginosa infections are difficult to treat, flammatory anaphylatoxin and antimicrobial compound C3a and to and as antibiotic-resistant strains are developing worldwide, there the opsonin C3b (6). Further progression of the complement cascade is a growing need to understand the crosstalk of this human path- generates a C5 convertase that cleaves C5 into the anaphylatoxin ogenic bacterium with the host immune system in detail and to C5a and C5b. Surface-attached C5b initiates the terminal pathway develop efficient and suitable drugs to control severe P. aeruginosa and formation of the cell-damaging cytolytic terminal complement infections (5). complex (TCC) (6). When complement activation is not properly controlled, the reaction is amplified, inflammatory and antimicrobial products are generated, and TCC is formed. These toxic molecules *Department of Infection Biology, Leibniz Institute for Natural Product Research and attack, damage, and ultimately allow removal of an infectious mi- Infection Biology, 07745 Jena, Germany; †Department of Clinical Sciences, Lund University, SE-221 84 Lund, Sweden; ‡Center for Biotechnology and Biomedicine, crobe. However, at the same time, host cells, tissues, and bio- University of Leipzig, 04103 Leipzig, Germany; and xFaculty of Biology, Friedrich membranes can also be attacked, and therefore these self cells need Schiller University, 07743 Jena, Germany to inhibit the damaging effects of complement. To this end, host Received for publication February 3, 2012. Accepted for publication September 10, cells and tissues use a series of complement inhibitors or regulators 2012. that are distributed both on membranes and as soluble molecules and This work was supported by Grant 7480 from the Hans-Knoell Institute Grundlagen- that efficiently control and block complement activation. fond, Jena, and the Swedish Research Council. Factor H and Factor H-like protein 1 (FHL-1) are two central Address correspondence and reprint requests to Prof. Peter F. Zipfel, Leibniz Institute for Natural Product Research and Infection Biology (Hans-Knoell Institute), Beuten- fluid-phase regulators of the alternative complement pathway (7). bergstrasse 11a, 07745 Jena, Germany. E-mail address: [email protected] Factor H is a 150-kDa plasma protein composed of 20 short The online version of this article contains supplemental material. consensus repeats (SCRs). FHL-1, which is derived from the Abbreviations used in this article: εACA, ε-aminocaproic acid; ATCC, American Factor H gene by alternative splicing, is a 42-kDa plasma protein Type Culture Collection; CFHR1, complement Factor H-related protein 1; ECM, composed of 7 SCRs. The 7 SCRs of FHL-1 are identical to the N- extracellular matrix; FHL-1, Factor H-like protein-1; Lpd, dihydrolipoamide dehydrogenase; MS, mass spectrometry; NB, nutrient broth; NCTC, National Collec- terminal SCRs of Factor H, and, in addition, the FHL-1 protein tion of Type Cultures; NHS, normal human serum; OMP, outer membrane protein; has a C-terminal extension of four unique amino acids (7, 8). RT, room temperature; SCR, short consensus repeat; TCC, terminal complement Factor H and FHL-1 regulate the alternative pathway of comple- complex; uPA, urokinase-type plasminogen activator. ment by binding C3b, by accelerating the decay of the alternative Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 pathway C3-convertase (C3bBb), and by acting as a cofactor for www.jimmunol.org/cgi/doi/10.4049/jimmunol.1200386 4940 Lpd INTERACTS WITH COMPLEMENT REGULATORS AND PLASMINOGEN the Factor I-mediated cleavage of C3b (6–8). The complement Materials and Methods Factor H-related protein 1 (CFHR1) is a member of the Factor H Bacterial strains and culture conditions protein family, and the protein is transcribed from a separate gene. P. aeruginosa strains American Type Culture Collection (ATCC) 27853; The corresponding plasma protein is composed of 5 SCRs and National Collection of Type Cultures (NCTC) 10662, SG137, and PAO1; appears as two isoforms with either two, 42-kDa (CFHR1b), or and the PAO1 derivative AH377 (24) were routinely cultured in enriched one, 37-kDa (CFHR1a) attached carbohydrate chains (9). CFHR1 nutrient broth (NB; Serva, Amstetten, Austria) at 37˚C. Thirteen clinical inhibits complement at the level of the C5 convertases and blocks isolates were derived from patients with different diseases. Cultures were ∼ TCC assembly and formation (10). grown to an OD600 1.0. Transformed Escherichia coli M15-expressing Lpd were grown in Luria Bertani liquid broth supplemented with 25 mg/ml To survive and establish an infection, any pathogenic microbe kanamycin and 10 mg/ml carbenicillin. must control, inactivate, and evade the innate immune response of the host, including the complement attack. Acquisition of soluble Biotinylation and isolation of P. aeruginosa surface proteins host complement regulators is one common and important evasion Surface biotinylation of intact P. aeruginosa was performed, as described strategy used by many, virtually all pathogenic microbes. P. aeru- (25). Briefly, bacteria (2 3 1011) were washed in buffer A (PBS, 1 mM ginosa, similar to other pathogenic bacteria, uses multiple strategies CaCl2, 0.5 mM MgCl2) and resuspended in buffer B (buffer A supplemented with 1.6 mM D-biotin). Bacteria were pelleted, and surface pro- to actively evade and control host innate immunity and complement teins were labeled with biotin by incubation with 500 ml 400 mM EZLink attack. P. aeruginosa binds the complement regulators Factor H, Sulfo-NHS-LC-Biotin (Thermo Fisher Scientific, Bonn, Germany) for FHL-1, and CFHR1, and the attached regulators inhibit complement 30 min on ice. Cells were washed extensively in buffer C (50 mM Tris [pH activation at the bacterial surface, block the generation of the op- 7.4], 100 mM NaCl, 27 mM KCl, 1 mM CaCl2, 0.5 mM MgCl2) and sonin C3b, and increase serum resistance (11). Elongation factor resuspended in buffer C supplemented with protease inhibitors (Complete; Roche, Mannheim, Germany). Bacteria were lysed by sonication, and Downloaded from Tuf, a 43-kDa surface protein, was identified as the first Factor H, biotinylated surface proteins were purified by affinity chromatography FHL-1, and CFHR1-binding protein of this Gram-negative patho- using ImmunoPure Immobilized Monomeric Avidin (Pierce). Proteins gen (11). In addition, P. aeruginosa secretes the endogenous pro- were eluted with D-biotin (2 mM in PBS), according to the manufacturer’s teases elastase and alkaline protease, which degrade and inactivate recommendations. The elute fractions were separated by SDS-PAGE and transferred to a membrane, and biotinylated proteins were identified by human complement components, including C1q, C2, and C3b (12, HRP-conjugated avidin (Roche). 13). Degradation of the opsonin C3b reduces phagocytosis of the pathogen by human neutrophils, and degradation of C3 inhibits Isolation of Factor H ligands with magnetic beads http://www.jimmunol.org/ fibroblast cell growth (12, 14, 15). In addition, P. aeruginosa Factor H SCR 8–20 (40 mg) was covalently coupled to magnetic beads produces an alginate layer that forms a mechanical barrier and that (100 ml suspension), according to the manufacturer’s instructions (Invi- reduces the accessibility and the action of host complement pro- trogen, Karlsruhe, Germany). Then the beads were incubated with a prepa- teins and other plasma defense factors (16). During late infection, ration of purified biotinylated P. ae ru gi no sa surface proteins for 2 h at 37˚C. After washing three times with 50 mM HEPES (pH 7.5), 1% Nonidet P-40, 1 P. aeruginosa forms biofilms, which prevent complement-mediated mM dithioerythritol, 1 mM MgCl2,and1mMCaCl2, proteins were eluted attack and phagocytosis (16, 17). with 40 ml 1 M NaCl for 10 min at 37˚C, separated by SDS-PAGE, and Following interaction with the complement system, many visualized by silver staining. Individual bands were excised from the gel and pathogens leave the site of infection and disseminate into deeper analyzed by mass spectrometry. by guest on October 4, 2021 tissue layers. For this purpose, pathogenic microbes interact with Protein identification by peptide mass fingerprinting and use either pathogen-encoded or host-derived proteolytic enzymes to degrade the extracellular matrix (ECM). The 92-kDa Silver-stained bands were destained using the ProteoSilver Plus silver stain kit (Sigma-Aldrich, Steinheim, Germany), and Coomassie-stained bands were human serum protein plasminogen is a human proenzyme that is destained, according to the manufacturer’s protocol. Subsequently, proteins acquired by several pathogenic microbes (18). Plasminogen is were reduced and alkylated using 10 mM DTT and 100 mM iodoacetamide composed of five consecutive kringle domains (K1–K5) that are in 25 mM NH4HCO3, washed with 25 mM NH4HCO3, and dehydrated with linked to a protease domain (P) (19). When activated by endoge- acetonitrile. Digestion with trypsin, MALDI-TOF-mass spectrometry (MS), and protein identification by peptide mass fingerprinting were performed, as nous activators like tissue-type plasminogen activator or urokinase- described (25). type plasminogen activator (uPA), the protease plasmin degrades ECM components like fibrinogen, fibronectin, vitronectin, and Expression and purification of recombinant proteins laminin and regulates cell migration, coagulation, fibrinolysis, in- The lpd gene (GenBank accession no. Q9I3D1: BankIt1558120 Seq1 flammation, wound healing, and tissue remodeling (18, 19). Plas- JX475907; BankIt1558120 Seq2 JX475908; BankIt1558120 Seq3 minogen is acquired by several pathogens, including P. aeruginosa, JX475909; BankIt1558120 Seq4 JX475910; BankIt1558120 Seq5 Streptococcus pneumoniae, Borrelia burgdorferi, Staphylococcus JX475911; BankIt1558120 Seq6 JX475912; BankIt1558120 Seq7 JX475913; BankIt1558120 Seq8 JX475914; BankIt1558120 Seq9 aureus, Lactobacillus johnsonii, and Candida albicans for ECM JX475915; BankIt1558120 Seq10 JX475916; BankIt1558120 Seq11 interaction, and activated plasmin is then used for destruction of JX475917; BankIt1558120 Seq12 JX475918; BankIt1558120 Seq13 host basement membranes and ECM (11, 20–23). JX475919; BankIt1558120 Seq14 JX475920; BankIt1558120 Seq15 In the current study, we identify P. aeruginosa dihydrolipoamide JX475921; BankIt1558120 Seq16 JX475922) was amplified from genomic DNA of P. aeruginosa strain PAO1. The amplicon was cloned into ex- dehydrogenase (Lpd) as a novel bacterial host regulator-binding pression vector pQE (Qiagen, Hilden, Germany) and transformed into protein, and thus as a multifunctional bacterial protein. Lpd is ex- E. coli strain M15, and protein expression was induced by 1 mM isopropyl posed at the bacterial surface and binds the human plasma proteins b-D-thiogalactoside. Recombinant Lpd was sequenced by MALDI-TOF- Factor H, FHL-1, CFHR1, and also plasminogen. Bound to Lpd, the MS/MS. Recombinant fragments of Factor H (SCRs 1–4, 1–5, 1–6, 8–11, four human plasma proteins aid in complement evasion as well as 11–15, 15–18, 15–20, 18–20, and 19–20) and FHL-1 (SCRs 1–7) were expressed in the baculovirus system, as described (26). Recombinant ECM degradation. Taken together, Lpd is a novel virulence factor CFHR1 and fragments of CFHR1 (SCRs 1–2, 3–4) were expressed in that mediates complement evasion and may facilitate tissue invasion Pichia pastoris, as described (27). The plasminogen gene (NM_000301) of the Gram-negative bacterium P. ae ru gi nos a. The identification of was amplified from human liver cDNA (Invitrogen) by PCR using Phusion a new virulence factor and immune evasion protein of P. aeruginosa High Fidelity DNA Polymerase (Finnzymes, Espoo, Finland) or Hot- StarTaq DNA Polymerase (Qiagen) with specific primers (Table I) intro- shows the complexity of the bacterial host immune crosstalk and ducing the restriction enzyme sites KpnI and XbaI. The sequence for the offers new insights into the interaction of P. aeruginosa with the signal peptide was excluded. The PCR constructs were subcloned into human host. E. coli Top10 cloning vectors pCR4Blunt-TOPO or pCR2.1Blunt-TOPO The Journal of Immunology 4941

(Life Technologies, Darmstadt, Germany). The resulting fragments were FHL-1 (20 mg/ml). After extensive washing with PBS-Tw, C3b (10 mg/ml) cloned into the vector pPICZaB (Life Technologies) of P. pastoris strain and Factor I (1.4 mg/ml) were added, and the reactions were terminated by 333. To produce recombinant proteins, P. pastoris was grown and induced the addition of SDS-PAGE sample buffer (RotiLoad1; Carl Roth, Karlsruhe, with 1% methanol. After 2 d of expression, the culture supernatants were Germany). The samples were subjected to SDS-PAGE, transferred to a ni- harvested and recombinant proteins were purified. All recombinant pro- trocellulose membrane, and analyzed by Western blotting. C3b degradation teins were purified by nickel affinity chromatography using HisTrap che- fragments were visualized by goat anti-human C3 (1:5000) that was added lating columns in a A¨ kta fast protein liquid chromatography system (GE for 1 h, followed by HRP-conjugated rabbit anti-goat (1:2500) (Dako) added Healthcare, Freiburg, Germany). for 40 min. After additional washing, development was performed with ECL Western blotting detection reagents (Applichem). Abs Plasminogen activation assay Polyclonal Lpd antiserum was raised by immunizing rabbits i.m. with 200 mg purified recombinant Lpd emulsified in CFA (Difco and BD Biosciences, Lpd was immobilized onto a microtiter plate at 2.5, 5, 10, and 20 mg/ml Heidelberg, Germany) and boosted on days 18 and 36 with the same dose of overnight at 4˚C. The plates were washed with PBS-Tw and blocked for 6 h protein in IFA. Blood was drawn 3 wk later. Generation of the human Factor at 4˚C with PBS containing 0.2% gelatin. After washing, plasminogen (8 mg/ H-specific mAbs, mAb E22 against SCR 3, and mAb C18 against SCR 20 ml) was bound to immobilized Lpd overnight at 4˚C, washed again, and has been described (28). Alexa fluor 488–conjugated polyclonal goat anti- incubated with uPA (Chemicon, Schwalbach, Germany) (1.48 nM) and D- rabbit was purchased from Molecular Probes, and the HRP-conjugated rabbit Valyl-Leucyl-Lysyl-p-Dinitroanilid (S-2251; Fluka) (0.54 mM) in Tris-HCl anti-goat, HRP-conjugated swine anti-rabbit, and HRP-conjugated rabbit (0.32 M) and NaCl (pH 7.5, 1.77 M). Plasmin activity was measured in anti-mouse were obtained from Dako (Glostrup, Denmark). A mouse His intervals of 30 min at 37˚C by recording the absorbance at 405 nm (Spektra- mAb was purchased from Qiagen. Polyclonal goat anti-Factor H and poly- Max 190; Molecular Devices). clonal goat anti-human C3 were obtained from Complement Technology (Tyler, TE), and polyclonal goat anti-plasminogen from Acris Antibodies Fibrinogen degradation by Lpd-bound plasminogen

(Herford, Germany). CFHR1 was detected with the mAb JHD10 (10) or m m Downloaded from rabbit CFHR1 antiserum (29). Fibrinogen degradation was assayed by rabbit Lpd (5 g/ml) or BSA (5 g/ml) was immobilized onto microtiter plates anti-human fibrinogen (Calbiochem, Nottinghamn, U.K.). overnight at 4˚C. The plates were washed with PBS-Tw and blocked for 6 h at 4˚C with PBS containing 0.2% gelatin. After washing, the immobilized ELISA proteins were incubated with plasminogen (20 mg/ml) overnight at 4˚C, washed, and incubated with uPA (160 ng/ml) together with fibrinogen (10 Microtiter plates (F96 Maxisorb, Nunc-Immuno Module) were coated with mg/ml) at 37˚C. The reaction was terminated at various time points (0.5–6 h) Lpd (5 mg/ml) or gelatin (5 mg/ml) overnight at 4˚C or with live bacteria by addition of SDS-PAGE sample buffer (Rotiload1; Carl Roth). Samples (0.5 3 107/well) for 1 h at 37˚C. The plates were washed four times with were separated by SDS-PAGE, transferred to a nitrocellulose membrane, and http://www.jimmunol.org/ PBS containing 0.1% Tween 20 (PBS-Tw) and blocked for 1 h with PBS analyzed by Western blotting. Fibrinogen degradation was analyzed using supplemented with 2% BSA (PBS-BSA) or blocking buffer I (Applichem, a rabbit anti-human fibrinogen (1:1000) that was added for 1 h, followed by Darmstadt, Germany). After washing, the plates were incubated for 1 h at HRP-conjugated swine anti-rabbit (1:2500) added for 40 min. Development room temperature (RT) with Factor H (Complement Technology) (0.08– was performed with ECL Western blotting detection reagents (Applichem). 0.67 mM); FHL-1 (0.07–0.6 mM); equimolar amounts of Factor H, FHL-1, CFHR1, and plasminogen (0.13 mM); and equimolar amounts of the various Flow cytometry Factor H constructs (0.13 mM), CFHR1 (0.07–0.58 mM), the two CFHR1 constructs (0.27 mM), plasminogen (Chromogenix, Milano, Italy) (0.001– The surface expression of Lpd on P. aeruginosa was analyzed by flow ∼ m m cytometry. The P. aeruginosa strain SG137 was grown to an OD600 1.0 0.54 M), recombinant plasminogen fragments (10 g/ml), or Lpd antise- 8 rum (1:1000). Thereafter, the wells were washed, and goat anti-human Factor and washed once in PBS-BSA, and then bacteria (10 ) were incubated with

H (1:500 to detect SCRs 8–11 and 1:2000 for the other deletion mutants), rabbit Lpd antiserum (1:100) for 45 min at 4˚C. Bacteria were washed and by guest on October 4, 2021 rabbit CFHR1 antiserum (1:1000), goat anti-plasminogen (1:1000 to detect incubated for 30 min with Alexa fluor 488-conjugated goat anti-rabbit K1-5-P and K1-5, 1:750 to detect K4-5-P, and 1:500 to detect K5), or mouse (1:200), washed, and analyzed by flow cytometry (FACScan LRII; BD anti-CFHR1 mAb (JHD10) (1:500) (10) was added in PBS-BSA for 1 h at Biosciences, Mountain View, CA). All incubations were kept in PBS-BSA, RT. After additional washings, HRP-conjugated rabbit anti-goat (1:2500), and secondary pAb was added separately as a negative control. HRP-conjugated swine anti-rabbit (1:2500), or HRP-conjugated rabbit anti- Transmission electron microscopy mouse (1:2500) was added for 40 min at RT. The reaction was developed with 1,2-phenylenediamine dihydrochloride (Dako), and the absorbance was The expression of Lpd on the surface of P. aeruginosa and the binding of measured at 492 nm. In the competition assays, the effect of heparin (0.25–5 Factor H and CFHR1 were analyzed by negative staining and electron mi- mg/ml) on Factor H (5 mg/ml), FHL-1 (5 mg/ml), or CFHR1 (5 mg/ml) croscopy, as described (31). Rabbit Lpd antiserum, plasma-purified human binding; the effect of mAbs E22 (20 mg/ml) and C18 (20 mg/ml) on Fac- Factor H, or recombinant CFHR1 was labeled with 5 nm colloidal gold, tor H binding; or the effect of the lysine analog ε-aminocaproic acid (εACA) 15 nm colloidal gold, or 45 nm colloidal gold, respectively. Bacteria (SG137) (0.1–10 mM) on plasminogen (5 mg/ml) binding to immobilized Lpd was were mixed with the Abs or Factor H and/or CFHR1, and 5 ml aliquots were assayed. adsorbed onto carbon-coated grids for 1 min, washed with two drops of water, and stained on two drops of 0.75% uranyl formate. The grids were Cofactor assay rendered hydrophilic by glow discharge at low pressure in air. Specimens Cofactor activity of Lpd-bound Factor H or FHL-1 was performed, as de- were observed in a Jeol JEM 1230 electron microscope (JEOL, Tokyo, scribed (30). Briefly, microtiter plates were coated with Lpd (10 mg/ml) or Japan) operated at 60 kV accelerating voltage. Images were recorded with HSA (10 mg/ml) overnight at 4˚C. The plates were washed four times with a Multiscan 791 CCD camera (Gatan, Pleasanton, CA). PBS-Tw and blocked for 1 h with blocking buffer I (Applichem). After washing, the plates were incubated for 1 h at RT with Factor H (20 mg/ml) or Isolation of P. aeruginosa outer membrane proteins Bacteria grown to exponential phase were washed with 50 mM Tris-HCl (pH 8.0). The pellet was resuspended in 50 mM Tris-HCl (pH 8.0) con- Table I. Primer used for the plasminogen fragments taining 3% Empigen (Calbiochem) and protease inhibitors (Complete; Roche). Outer membrane proteins (OMPs) were extracted by rotating the mixture at 37˚C for 2 h. The bacterial cells, stripped of their outer mem- Plasminogen Fragment Primer Sequence (59 to 39) branes, were centrifuged, and the supernatants were collected. Thereafter, the supernatants were subjected to SDS-PAGE, transferred to a membrane, K1-K5-P Forward GGTACCACGAGCCTCTGGATGAC and analyzed by Western blotting using a rabbit Lpd antiserum (1:500) for K1-K5-P Reverse GCGGCCGCATTATTTCTCATCACTC 1 h, followed by HRP-conjugated swine anti-rabbit (1:2500) for 40 min. K4-K5-P Forward GGTACCTAGCACCACCTGAGCTA Development was performed with ECL Western blotting detection reagents K4-K5-P Reverse GCGGCCGCATTATTTCTCATCACTC (Applichem). K5-P Forward GGTACCTAGACTGTATGTTTGGG K5-P Reverse GCGGCCGCATTATTTCTCATCACTC Sequencing of the lpd gene K1-K5 Forward GGTACCTAGTGTATCTCTCAGAG The lpd gene of the 5 P. aeruginosa strains (SG137, ATCC 27853, NCTC K1-K5 Reverse TCTAGAGTACACTGAGGGACATC 10662, PAO1, AH377) and of 13 clinical isolates was amplified by PCR 4942 Lpd INTERACTS WITH COMPLEMENT REGULATORS AND PLASMINOGEN using HotStarTaq (Qiagen) using the primers 59-TAGCGTTTCTCTT- CACGGCG-39 and 59-CCGCGTCGCTGTGACGCCCC-39. The sequence of the amplified products was determined using an ABI 3100 sequence analyzer (Applied Biosystems). The sequences were compared with that of the lpd gene of P. aeruginosa PAO1 (GenBank accession no. Q9I3D1). Serum bactericidal assay The five P. aeruginosa strains (SG137, ATCC 27853, NCTC 10662, PAO1, AH377) were grown to an OD600 ∼1.0 and diluted in Mg-EGTA buffer. Bacteria (104 CFU) were incubated in normal human serum (NHS), used at increasing concentrations (0–50%), NHS or complement active HSDFH/ FHL-1/CFHR1 used at 10–50% depending on the strain (SG137, 10%; ATCC 27853, 20%; NCTC 10662, 15%; PAO1, 50%; AH377, 50%), or depleted serum reconstituted with purified Factor H (100 mg/ml) in a final volume of 100 ml at 37˚C. After 0 and 60 min, 10 ml aliquots were re- FIGURE 1. Recombinant expression of P. aeruginosa Lpd and genera- moved and spread onto NB plates. After 18 h of incubation at 37˚C, CFU tion of a Lpd-specific antiserum. (A) Recombinant Lpd was expressed as were determined. In the inhibition experiments, bacteria (strain SG137) a N-terminal His-tag protein. The Lpd-encoding cDNA was amplified from were preincubated with Lpd antiserum (1:20), Factor H deletion mutants genomic DNA derived from P. aeruginosa strain PAO1 by PCR and cloned SCRs 18–20 (25 mg/ml), or SCRs 8–11 (25 mg/ml) for 10 min at RT. After into expression vector pQE, and the corresponding protein was recombi- washing, bacteria were challenged with 10% NHS in a final volume of 100 ml at 37˚C. In addition, bacteria were incubated with an Efb antiserum nantly expressed in E. coli with a N-terminal His-tag. Recombinant Lpd was (1:20), directed against the S. aureus protein Efb, which is absent on the purified by affinity chromatography. The E. coli lysate, flow through (FT), m wash (w), and elution (e) fractions were separated by SDS-PAGE and an- surface of P. aeruginosa,or5 l rabbit IgGs. Downloaded from alyzed by silver staining (A) or by Western blotting (B). The His antiserum Statistics detected a 57-kDa band, and the Lpd antiserum detected a 57- and a 50-kDa Results were analyzed by Student t test for paired data. The p value band. The mobility of the size markers is indicated on the left in kDa. #0.05 was considered to be statistically significant. *p # 0.05, **p # # 0.01, ***p 0.001. was dose dependent (Fig. 2B). Factor H contains two heparin binding domains located in SCR 7 and SCRs 19–20, and FHL-1 Results contains one heparin binding domain in SCR 7 (33). Therefore, we http://www.jimmunol.org/ Identification of a second Factor H-binding protein of asked whether heparin affects the Lpd/Factor H or Lpd/FHL-1 P. aeruginosa interaction. Heparin inhibited binding of both human regulators, To identify additional Factor H-binding proteins of P. aeruginosa, and the effect was dose dependent (Fig. 2C, 2D). Heparin used at surface proteins of P. aeruginosa strain PAO1 were biotinylated, 1 mg/ml inhibited Factor H binding to Lpd by 73% and FHL-1 extracted, and absorbed to a Factor H/SCR 8–20 matrix. Bound binding by 66%. proteins were eluted, separated by SDS-PAGE, and visualized by To localize the region(s) of Factor H and FHL-1 that contacts silver staining. This approach identified two major bands with Lpd, a series of Factor H deletion mutants were tested for binding to mobilities of ∼57 and 43 kDa (data not shown). The 57-kDa band immobilized Lpd. The N-terminal SCRs 1–7 (FHL-1) and the C- by guest on October 4, 2021 was identified as Lpd, and the 43 kDa band as Tuf, the known terminal constructs SCRs 15–20 and SCRs 18–20 bound to Lpd Factor H-binding protein (11). Lpd of P. aeruginosa is a 478-aa (Fig. 2E). Deletion fragments SCRs 1–4, SCRs 1–5, SCRs 1–6, flavoprotein with a predicted molecular mass of 50.2 kDa (32). SCRs 8–11, SCRs 11–15, SCRs 15–18, and SCRs 19–20 bound either with reduced intensity or did not bind. Thus, the human Recombinant expression of Lpd and generation of polyclonal complement regulator Factor H binds to Lpd via two interaction Lpd antiserum domains that are included within SCR 7 and SCRs 18–20, and P. aeruginosa lpd gene was amplified, the corresponding PCR FHL-1 binds via one domain, that is, SCR 7. To confirm that the product was cloned into expression vector pQE, recombinantly C terminus of Factor H binds to bacterial Lpd, epitope-specific expressed in E. coli as a N-terminal His-tagged protein, and pu- mAbs were used as blocking agents. A mAb that binds to SCR3 rified by nickel chelate chromatography. The bacterial lysate, (E22) and a mAb that binds to SCR 20 (C18) of Factor H were flow-through (FT), wash (w), and elute (e) fractions were sep- used. The mAb C18, which binds to the C terminus, blocked arated by SDS-PAGE, and proteins were visualized by silver Factor H binding to immobilized Lpd by .60% (Fig. 2F). In staining. This approach identified in the elute fraction two bands, contrast, mAb E22, which reacts with the N terminus, did not one of 57 and one of 50 kDa (Fig. 1A, lane 4). In addition, a His affect Factor H binding to Lpd. Thus, this demonstrates that in- antiserum identified the 57-, but not the 50-kDa band (Fig. 1B, hibition is not caused by steric hindrance and that the surface lane 1). Purified recombinant Lpd was used to immunize rabbits, recognition region of Factor H (SCRs 18–20) is important for and the corresponding polyclonal Lpd antiserum identified both binding of Lpd. the 57- and 50-kDa bands (Fig. 1B, lane 3). The two bands were Factor H and FHL-1 bound to Lpd exhibit cofactor activity excised from the gel and analyzed by mass spectrometry, and both the upper and the lower band were identified as Lpd. The reactivity Factor H and FHL-1 control alternative pathway-mediated com- with the His- and the Lpd-specific antiserum, together with the plement activation at the level of the C3 convertase by acting as mass spectometry data, suggests that the 57-kDa band represents cofactors for the serine protease Factor I. To demonstrate that the the full-length His-tagged Lpd protein, and the 50-kDa band most regulators bound to Lpd are functionally active, we assayed cofactor likely is a degradation product that has the His tag cleaved off. activity of Lpd-bound Factor H or FHL-1 for C3b degradation. C3b degradation was revealed as the appearance of the degradation Recombinant Lpd binds Factor H and FHL-1 fragments a943 and a941 kDa. Bound to Lpd, Factor H and FHL-1 To confirm that Lpd of P. aeruginosa is a Factor H-binding pro- were functionally active, acting as cofactors in the degradation of tein, binding of Factor H to Lpd was assayed. Factor H bound to C3b (Fig. 3A, 3B, lanes 3). Thus, Factor H and FHL-1 bound to immobilized Lpd, and binding was dose dependent (Fig. 2A). In P. aeruginosa Lpd maintain their central regulatory functions and addition, FHL-1 also bound to immobilized Lpd, and again binding control complement activation at the level of the C3 convertase. The Journal of Immunology 4943

FIGURE 3. Lpd-bound Factor H and FHL-1 are functionally active and maintain complement-regulatory activity. Factor H (A) and FHL-1 (B) when bound to immobilized Lpd displayed cofactor activity for Factor I in C3b inactivation. Factor H (A) or FHL-1 (B) was bound to Lpd and, after extensive washing, C3b and Factor I were added. After incubation, the supernatants were separated by SDS-PAGE, and C3b cleavage products were analyzed by Western blotting using goat anti-C3 and HRP-conjugated anti-goat. Cofactor activity of Factor H or FHL-1 bound to Lpd is visualized by the appearance of the cleavage fragments a943 and a941 (A, B, lane 3). The C3b cleavage pattern obtained with immobilized Factor H (A) or FHL-1 (B) incubated with C3b and Factor I was used as a positive control to identify the degradation products (A, B, lane 2). No cleavage

occurred when HSA was used as a matrix (A, B, lane 4). A representative Downloaded from experiment of three is demonstrated.

(Fig. 4B). Thus, heparin inﬂuences the binding of the complement regulator CFHR1 to P. aeruginosa Lpd. In addition, the interacting domains of CFHR1 that are relevant for contact of Lpd were localized by assaying binding of equimolar http://www.jimmunol.org/ amounts of CFHR1 deletion mutants. Both deletion mutants bound to immobilized Lpd, but CFHR1/SCRs 3–5 bound with ∼4 times higher intensity to Lpd compared with CFHR1/SCRs 1–2 (Fig. 4C). Thus, the C-terminal CFHR1/SCRs 3–5 represent the major binding region of Lpd. by guest on October 4, 2021

FIGURE 2. Lpd binds the alternative pathway regulators Factor H and FHL-1. Factor H (A) and FHL-1 (B) bound to immobilized Lpd, and binding was dose dependent. Binding of Factor H (0.08–0.67 mM) or FHL- 1 (0.07–0.6 mM) to immobilized Lpd was assayed by ELISA. Bound Factor H or FHL-1 was detected with polyclonal Factor H antiserum, followed by HRP-conjugated anti-goat. (C and D) Heparin inhibits the Lpd/Factor H and Lpd/FHL-1 interaction in a dose-dependent manner. The effect of increasing concentrations of heparin (0.25–5 mg/ml) on binding of Factor H or FHL-1 to immobilized Lpd was assayed. (E) Localization of Lpd binding domains within Factor H and FHL-1. Binding of recombinant Factor H and FHL-1 fragments (0.13 mM) to immobilized Lpd was assayed by ELISA. Factor H uses two contact regions, SCR 7 and SCRs 18–20 to contact Lpd. (F) A mAb against SCR 20 (C18) of Factor H inhibits the Lpd/Factor H interaction. The effect of E22 (SCR 3) and C18 (SCR 20) on Factor H binding of Factor H to immobilized Lpd was assayed. The mean values of three independent experiments and SDs are presented. Statistical signiﬁcance of differences was estimated using Stu- dent t test. **p # 0.01, ***p # 0.001. FIGURE 4. Lpd binds the terminal pathway regulator CFHR1. (A)CFHR1 binds to immobilized Lpd, and binding was dose dependent. Binding of Lpd binds the terminal pathway regulator CFHR1 CFHR1 (0.07–0.58 mM) to immobilized Lpd was assayed by ELISA. Bound The C-terminal surface attachment region SCRs 18–20 of Factor CFHR1 was detected with polyclonal CFHR1 antiserum, followed by HRP- H forms one binding site for Lpd. Given the extensive sequence conjugated anti-rabbit. (B) Heparin inhibits the Lpd/CFHR1 interaction. The similarity of the C termini of Factor H (SCRs 18–20) and CFHR1 effect of heparin used at increasing concentrations (0.25–5 mg/ml) on CFHR1 binding to immobilized Lpd was assayed. (C) Localization of Lpd binding (SCRs 3–5) (100, 100, and 97%) (10), we asked whether CFHR1 domains within CFHR1. Binding of recombinant CFHR1 fragments (0.27 also binds to Lpd. CFHR1 bound to immobilized Lpd, and binding mM) SCRs 1–2 and SCRs 3–5 to immobilized Lpd was assayed by ELISA. was dose dependent (Fig. 4A). CFHR1 is a heparin-binding pro- The major region of CFHR1 that contacts Lpd is SCRs 3–5. The mean tein, and, therefore, we also assayed the effect of heparin for the values of three independent experiments and SD are presented. Statistical CFHR1/Lpd interaction (10). Heparin affected CFHR1 binding signiﬁcance of differences was determined by the Student t test. **p # 0.01, to Lpd, and at 5 mg/ml heparin inhibited the interaction by 56% ***p # 0.001. 4944 Lpd INTERACTS WITH COMPLEMENT REGULATORS AND PLASMINOGEN

Lpd binds human plasminogen deletion mutants were generated, and these proteins were tested for Several microbial Factor H-binding proteins also bind plasminogen Lpd binding. Recombinant full-length plasminogen (K1-5-P), the (34–37). We therefore asked whether P. aeruginosa Lpdalsobinds deletion mutant that has kringle domains 4 and 5 linked to the protease domains (i.e., K4-K5-P) and a fragment that includes all ﬁve plasminogen. Plasminogen bound to Lpd, and binding was dose kringle domains, but lacks the protease domain (K1-5), bound to dependent and saturated at 0.27 mM plasminogen (Fig. 5A). As ly- immobilized Lpd (Fig. 5C). The fragment that includes kringle sine residues are often relevant for plasminogen binding, we assayed 5 and the protease domain bound to immobilized Lpd with a re- ε whether the lysine analog ACA inﬂuences the plasminogen/Lpd duced intensity. Thus, this indicates that kringle domain 4 of plas- ε interaction. ACA used at 1.0 mM inhibited plasminogen binding minogen contains the major Lpd binding site. to immobilized Lpd by 50%, and at 10 mM εACA blocked the interaction completely (Fig. 5B). Thus, lysine residues are relevant for Lpd-bound plasminogen is accessible for uPA and is the Lpd/plasminogen contact. proteolytically active To localize the domain(s) of plasminogen that contacts Lpd, Plasminogen is a zymogen that can be converted to active plasmin recombinant plasminogen and a series of recombinant plasminogen (19). To determine whether Lpd-bound plasminogen is accessible for Downloaded from http://www.jimmunol.org/ by guest on October 4, 2021

FIGURE 5. Plasminogen binds to Lpd, and bound to Lpd it is functionally active. (A) Plasminogen binding to Lpd was assayed by ELISA. Plasminogen used at 0.001–0.54 mM bound to immobilized Lpd dose dependently. Bound plasminogen was detected with polyclonal plasminogen antiserum, followed by HRP-conjugated anti-goat. (B) Lysine residues are responsible for the interaction between Lpd and plasminogen. The effect of increasing concentrations (0.1–10 mM) of the lysine analog εACA on plasminogen binding to immobilized Lpd was assayed. (C) Localization of the Lpd binding site within plasminogen. The major Lpd binding site is located within kringle domain 4 of plasminogen. Recombinant deletion mutants of plasminogen, as indicatedin the left part of the figure, were generated in P. pastoris. Binding of purified deletion mutants (10 mg/ml) to immobilized Lpd was assayed. Bound fragments were detected with polyclonal plasminogen antiserum and HRP-conjugated rabbit anti-goat. The mean values of three independent experiments and SD are presented. Statistical significance of differences was estimated using Student t test. *p # 0.05, **p # 0.01, ***p # 0.001. (D) Plasminogen bound to Lpd was treated with the activator uPA to generate plasmin, and degradation of the synthetic chromogenic substrate S-2251 was assayed photometrically. Lpd (2.5–20 mg/ml) was immobilized onto microtiter plates, and then plasminogen (8 mg/ml) was added. After extensive washing, the activator uPA was added together with the chromogenic substrate S-2251, and conversion of the chromogenic substrate was determined by measuring the absorbance at 405 nm. (E) Degradation of the natural substrate fibrinogen. The fibrinolytic activity of uPA-activated plasminogen bound to Lpd was evaluated. At the indicated time points (15, 30, 45, 60, and 120 min), the supernatants were separated by SDS-PAGE, proteins were transferred to a membrane, and degradation of fibrinogen was assayed by Western blotting using a rabbit fibrinogen antiserum and HRP-conjugated anti-rabbit IgG. When BSA, which does not bind plasminogen, was used as a matrix, fibrinogen remained intact and was not cleaved. Data shown are representative of three independent experiments. The Journal of Immunology 4945 and converted by the physiological activator uPA to plasmin, plas- proteins compete for binding. Plasminogen bound dose depen- minogen bound to Lpd was treated with uPA, and the activity of dently to immobilized Lpd and did not affect Factor H binding (Fig. plasmin was assayed. Lpd-bound plasminogen was converted to 6C). Similarly, Factor H when used at increasing concentrations plasmin, which cleaved the chromogenic substrate S-2251 (Fig. 5D). bound dose dependently to Lpd, and plasminogen binding was not In addition, we tested whether plasmin bound to Lpd also affected (Fig. 6D). Thus, the two human plasma proteins plas- cleaved the physiological substrate fibrinogen. Fibrinogen was minogen and Factor H bind independently and most likely to degraded by newly formed plasmin, as revealed by the decrease in different regions of the Lpd protein. At physiological plasma intensity of the two fibrinogen bands (64 and 58 kDa). Already after levels (plasminogen:Factor H molar ratio of 1:1.5, indicated by the 30 min fibrinogen was degraded (Fig. 5E, lane 3, upper panel), arrow), both Factor H and plasminogen bound to Lpd. and, after 1 h, fibrinogen was no longer detectable (Fig. 5E, lane 5, Lpd binds equimolar amounts of Factor H, FHL-1, CFHR1, upper panel). Thus, Lpd-bound plasminogen is accessible for the and plasminogen activator uPA and is activated to plasmin, which cleaves both a synthetic and a natural substrate. The four human regulators, Factor H, FHL-1, CFHR1, and plasminogen, bind to Lpd. To compare the intensity of binding, each Competition of CFHR1 and Factor H for binding to Lpd human plasma protein was bound at equimolar amounts to immo- Both CFHR1 and Factor H bind Lpd via their conserved C ter- bilized Lpd. Factor H, FHL-1, and CFHR1 bound to Lpd with minus. We therefore asked whether the two proteins bind simul- comparable intensity and plasminogen with a stronger intensity (Fig. taneously to different sites of Lpd or whether they compete for 6D). As Factor H, FHL-1, and CFHR1 are detected with the same binding. CFHR1 when used at increasing concentrations slightly antiserum, it is concluded that the three human complement pro- reduced Factor H binding to Lpd at a molar ratio of 0.1:1–1:1 and teins bind with similar intensity to Lpd. In this setup, the intensity of Downloaded from was further reduced by 53% at a ratio of 4:1 (Fig. 6A). Similarly, plasminogen binding cannot be directly compared with the other in a reverse setting, Factor H slightly reduced CFHR1 binding to plasma proteins as plasminogen was detected with a different an- Lpd. At a 4-fold excess of Factor H, CFHR1 binding was reduced tiserum. by 50% (Fig. 6B). The latter ratios exceed the physiological level Lpd is a bacterial surface protein of the two regulators in human plasma, where the molar ratio is

0.35:1 (CFHR1:Factor H). These results show that CFHR1 and Surface biotinylation and also binding of human regulators sug- http://www.jimmunol.org/ Factor H bind simultaneously to bacterial Lpd at physiological gested surface expression of P. aeruginosa Lpd. Therefore, we plasma concentrations. analyzed Lpd expression at the bacterial surface. First, Lpd was detected at the bacterial surface by ﬂow cytometry (Fig. 7A). In ad- Factor H and plasminogen bind simultaneously to Lpd dition, when Lpd was analyzed on the surface of intact bacteria by As plasminogen and Factor H also bind to Lpd, we asked whether electron microscopy, the protein was evenly distributed over the sur- the two proteins bind simultaneously to Lpd and whether the two face (Fig. 7B). Next, expression levels of Lpd were compared for the by guest on October 4, 2021

FIGURE 6. Factor H and CFHR1 and Factor H and plasminogen bind to different sites in the Lpd protein. (A and B) Binding of CFHR1 and Factor H to immobilized Lpd was analyzed by ELISA. Bound CFHR1 was detected with anti-CFHR1 mAb (JHD10) (white columns), and bound Factor H was detected with Factor H-specific antiserum (black columns). The CFHR1:Factor H ratio of 0.35:1 represents the physiological plasma levels and is indicated by the arrow. (C) The effect of increasing levels of plasminogen on Factor H binding was assayed by ELISA. Plasminogen bound to Lpd was detected with specific antiserum (white columns), and bound Factor H was visualized with the Factor H antiserum (black columns). Plasminogen used at increasing levels (molar ratios are shown) did not influence Factor H binding to Lpd. (D) Similarly, Factor H used at increasing levels did not influence plasminogen binding to Lpd. The plasminogen: Factor H ratio of 1:1.5 represents the serum levels and is indicated by the arrow. (E) Binding of equimolar concentrations of Factor H, FHL-1, CFHR1, and plasminogen (0.13 mM) to immobilized Lpd was tested by ELISA. Bound Factor H, FHL-1, and CFHR1 were detected with the polyclonal Factor H antiserum, and bound plasminogen was detected with a plasminogen-specific antiserum. The mean values of three independent experiments and SD are presented. Statistical significance of differences was estimated using Student t test. *p # 0.05. 4946 Lpd INTERACTS WITH COMPLEMENT REGULATORS AND PLASMINOGEN

cell ELISA, and OMP preparations, identified Lpd as a surface protein of P. ae ru gi nos a. P. aeruginosa has at least two Factor H, FHL-1– and CFHR1- binding proteins. Therefore, we wanted to demonstrate the role of Lpd for serum resistance. To this end, we blocked Lpd on the surface of strain SG137, which had the highest Lpd surface level using the specific Lpd antiserum. Then bacteria were challenged with NHS, and, after incubation, bacterial survival was assayed. Blockade of Lpd reduced the bacterial survival by 56% compared with the survival of SG137 in NHS (Fig. 7E). The effect was specific, as neither an unrelated antiserum nor purified rabbit IgGs affected bacterial survival. Thus, surface-exposed Lpd contributes to serum resistance of P. aeruginosa.

Lpd sequence variation in clinical P. aeruginosa isolates To assay whether the sequence of the lpd gene in the different isolates is conserved or polymorphic, the lpd gene of all 13 clinical P. aeruginosa isolates and of the 5 laboratory strains was amplified and the sequence was determined. Sequence analysis revealed a total of 25 nucleotide exchanges among the 18 strains. Twenty-two Downloaded from nucleotide exchanges represented synonymous exchanges (Fig. 8A), and three nucleotide exchanges represented nonsynonymous exchanges. The nonsynonymous exchange, which causes an exchange of Asn97 to Ser, was identified in the laboratory strain ATCC 27853 and in two clinical isolates (i.e., 9 and 11) (Fig. 8C). An Ala179 to Val exchange was identified in the laboratory strain SG137, http://www.jimmunol.org/ and a His306 to Tyr in the clinical isolate 1 (Fig. 8C). Thus, the lpd gene of P. aeruginosa is rather conserved. FIGURE 7. Lpd is a surface-exposed protein of P. aeruginosa and con- To assay whether the clinical isolates of P. aeruginosa express tributes to bacterial serum resistance. (A) Surface expression of Lpd was Lpd in the outer membrane, the 13 clinical Pseudomonas isolates identified by flow cytometry. Bacteria (P. aeruginosa strain SG137) were were tested for the presence of Lpd in the OMP fractions. Lpd was stained with Lpd antiserum, and bound Abs were detected with Alexa 488- detected in the OMP fraction of all 13 P. aeruginosa clinical isolates labeled rabbit antiserum by flow cytometry (mean fluorescence intensity, as a 57-kDa band (Fig. 8B). The secondary Ab alone did not cross- 6 6 1015 330, compared with background mean fluorescence intensity, 408 react with Lpd (Supplemental Fig. 1). However, when assaying Lpd 82). (B) Lpd surface expression was confirmed by transmission electron by guest on October 4, 2021 levels by whole-cell ELISA, 11 clinical isolates expressed Lpd; microscopy. The P. aeruginosa strain SG137 was incubated with a gold- labeled Lpd antiserum and revealed an even distribution of Lpd on the however, the level of expression varied (Fig. 8C). Lpd expression bacterial surface. (C) Five laboratory strains of P. aeruginosa were analyzed was highest for isolate 1, followed by the clinical isolates 2, 5, and for Lpd surface expression using a whole-cell ELISA. Intact bacteria were 4. Compared with isolate 1, Lpd levels among the 10 other clinical immobilized onto microtiter plates, and Lpd surface expression was isolates ranged from 38 to 75%. Clinical isolates 10 and 13 did not detected with Lpd antiserum and HRP-conjugated goat anti-rabbit pAb. The express Lpd when analyzed by whole-cell ELISA. Thus, Lpd is mean values of three independent experiments and SD are presented. (D) expressed in the outer membrane of all tested P. aeruginosa strains, Lpd is present in the outer membrane fractions of the five laboratory strains but surface expression varies between strains. of P. ae ru gi nos a strains as a 57-kDa band. A membrane fraction enriched for outer membrane proteins was prepared from the indicated P. ae ru gi nos a Distribution of Factor H and CFHR1 on the bacterial surface strains, separated by SDS-PAGE, transferred to a membrane, and analyzed Binding and surface distribution of the two human complement by Western blotting using the Lpd antiserum. A typical result of three independent experiments is shown. (E) Lpd contributes to survival of regulators Factor H and CFHR1 were evaluated by electron P. aeruginosa. Lpd on the surface of P. aeruginosa strain SG137 was microscopy. Used as single proteins, Factor H (Fig. 9AI) and also blocked with Lpd antiserum. Then bacteria were challenged with NHS CFHR1 (Fig. 9AII) bound to the surface of P. ae ru gi nos a.Both (10%), and, after incubation, bacteria were plated on NB agar plates and the proteins were distributed over the whole bacterial surface. When number (CFU) of surviving bacteria was determined. Efb antiserum added together, the two proteins, labeled with either gold particles detecting an unrelated S. aureus protein and a purified rabbit IgG prepa- of 15 nm (Factor H) or 45 nm (CFHR1) bound in close vicinity to ration were used as controls. The mean values of three independent each other (Fig. 9AIII, black arrows), but also to distinct sites (Fig. experiments and SD are presented. Statistical significance of differences 9AIII, white arrows). Thus, the two human proteins colocalize at the # # was estimated using Student t test. *p 0.05, ***p 0.001. surface of P. aeruginosa and apparently bind to the same bacterial proteins. five P. aeruginosa strains using a whole-cell ELISA. The Lpd levels were highest on the surface of strain SG137. Lpd levels for the strains Bound to the surface of P. aeruginosa, the complement ATCC 27853, NCTC 10662, PAO1, and AH377 ranged from 50 to regulators Factor H, FHL-1, and CFHR1 protect against the 80% in comparison with strain SG137 (Fig. 7C). In addition, the complement attack presence of Lpd was analyzed in the OMP fractions of the five strains. We were interested to determine the contribution of all three human Lpd was detected in the OMP preparations of all five strains regulators for serum resistance. To assay whether the five laboratory (i.e., SG137, ATCC 27853, NCTC 10662, PAO1, and AH377) as a 57- P. aeruginosa strains varied in serum resistance and to determine kDa band (Fig. 7D, lanes 1–5). Thus, four different and indepen- the optimal serum concentration for bacterial survival in comple- dent methods, that is, flow cytometry, electron microscopy, whole- ment regulator-depleted serum, the five laboratory P. ae ru gi nos a The Journal of Immunology 4947 Downloaded from http://www.jimmunol.org/

FIGURE 8. Clinical isolates of P. aeruginosa express Lpd. (A) Sequence variation of the lpd gene of P. aeruginosa was analyzed in the 5 P. aeruginosa strains and 13 clinical isolates. A total of 25 nucleotide exchanges was detected in the 1437-bp–long lpd gene. Twenty-two nucleotide exchanges and three nonsynonymous nucleotide exchanges were identified. The three nonsynonymous exchanges resulted in substitution of Asn97 to Ser, which was detected in the laboratory strain ATCC 27853 and in two clinical isolates (9 and 11); the Ala179 to Val exchange was identified in the laboratory strain SG137; and the His306 to Tyr exchange in the clinical isolate 1. (B) Lpd is detected in the outer membrane fractions of all clinical isolates as a 57-kDa band. A fraction containing outer membrane proteins was prepared from the 13 clinical P. aeruginosa isolates, separated by SDS-PAGE, transferred to a membrane, and analyzed by Western blotting using the Lpd antiserum. A typical experiment of three is demonstrated. (C) Lpd expression was tested on the surface of 13 by guest on October 4, 2021 clinical P. aeruginosa isolates using a whole-cell ELISA. Intact bacteria were immobilized onto microtiter plates, and Lpd surface expression was detected with the Lpd antiserum and HRP-conjugated goat anti-rabbit pAb. The mean values of three independent experiments and SD are presented. Statistical significance of differences was determined by using Student t test. *p # 0.05, **p # 0.01, ***p # 0.001. strains were challenged with increasing concentrations of NHS. The SCRs 8–11, which do not bind to P. aeruginosa, did not affect the P. aeruginosa strains showed differences in serum resistance (Fig. bacterial survival. Thus, this demonstrates that the human com- 9B). Survival of the strains SG137, ATCC 27853, and NCTC 10662 plement regulators when bound to the bacterial surface protect the was reduced dose dependently when the NHS concentration was bacterium from the damaging effects of the activated complement increased. In contrast, PAO1 and AH377 survived in all tested NHS system. concentrations. Therefore, the various P. ae ru gi nos a strains were cultivated in different concentrations of regulator-depleted human Discussion serum to obtain optimal results. To this end, all five laboratory In this study, we identify Lpd as a novel surface-exposed com- strains were challenged with complement active human serum in plement regulator-binding protein of the Gram-negative bacterium which all three regulators were depleted, that is, HSDFH/FHL-1/ P. aeruginosa. Lpd binds Factor H and three additional human CFHR1 (Supplemental Fig. 2). After incubation, bacteria were re- plasma proteins, that is, FHL-1, CFHR1, and plasminogen. The covered and survival was evaluated. In this case, survival of strain surface-bound complement inhibitors contribute to serum resis- SG137 was reduced by 84% (Fig. 9C). Survival of the other strains tance, and bacterial survival was reduced when P. aeruginosa was also reduced, but to a different extent. Survival of strain AH strain SG137 was challenged with complement active human se- 377 was reduced by 70%, strain ATCC 27853 by 62%, strain PAO1 rum depleted of Factor H/FHL-1/CFHR1. Lpd is a bacterial sur- by 37%, and strain NCTC 10662 by 34% (Fig. 9C). Addition of face protein involved in serum resistance of P. aeruginosa. Thus, Factor H to the depleted serum reversed the effect, and survival of Lpd is a new virulence factor of P. aeruginosa that binds Factor H, all three strains was increased (Fig. 9D). To confirm the protective FHL-1, CFHR1, and plasminogen. effect of Factor H for the survival of P. aeruginosa, one of the Lpd Lpd is the second Factor H, FHL-1, CFHR1, and plasminogen- binding sites within Factor H was blocked by preincubating strain binding immune evasion protein identified from P. aeruginosa. SG137 with the C-terminal surface binding region SCRs 18–20 of Lpd, which has a predicted mass of 51.6 kDa, was initially iden- Factor H. Then bacteria were challenged with NHS, and, after in- tified as a cytoplasmic protein and as a component of the enzymatic cubation, bacterial survival was assayed. Blockade of Factor H pyruvate dehydrogenase complex, which catalyzes the electron binding to P. ae ru gi nos a by the C-terminal SCRs 18–20 reduced the transfer between pyridine nucleotides and disulfide components (32, bacterial survival by 62% (Fig. 9E). The effect was specific, as 38, 39). Sequence analysis of the 5 laboratory strains and 13 clinical 4948 Lpd INTERACTS WITH COMPLEMENT REGULATORS AND PLASMINOGEN

isolates revealed 22 synonymous and 3 nonsynonymous exchanges. Thus, lpd is a highly conserved bacterial gene. In this study, we characterize Lpd as a Pseudomonas surface protein, which is present in the outer membrane of all strains tested. Lpd was also identified at the surface of the 5 analyzed laboratory strains and on 80% of the clinical P. aeruginosa isolates when analyzed by whole- cell ELISA. Similar to P. aeruginosa Lpd, Lpd derived from the pathogenic microbes Trypanosoma brucei and C. albicans is also surface localized (40, 41). Lpd is a moonlighting protein that lacks a typical secretion signal (42). Therefore, it is unclear how Lpd and also the other microbial moonlighting proteins are exported to and anchored into the microbial membrane. In addition to Lpd, several other moonlighting proteins, including Tuf from P. aeruginosa, Gpm1 from C. albicans,andTuffromMycoplasma pneumoniae, bind complement regulators and other human plasma proteins (11, 23, 43). When Lpd was blocked at the surface of strain SG137 with specific antiserum and the bacteria were challenged with NHS, bacterial survival was reduced by 56%. Thus, Lpd contributes to serum resistance. The human complement regulator Factor H binds Pseudomonas Downloaded from Lpd via two contact domains, that is, SCR 7 and SCRs 18–20, and FHL-1 binds via one region, that is, SCR 7. Consequently, when bound to Lpd, the complement-regulatory region (i.e., SCRs 1–4) of both human proteins is exposed to the outside and able to bind C3b and to regulate complement. Factor H and FHL-1 bound to Lpd maintained their complement-regulatory activity as cofactors http://www.jimmunol.org/ for Factor I and contributed to C3b inactivation. This degradation of C3b can block C3b opsonization on the bacterial surface and consequently interfere with uptake and phagocytosis of the bacteria. In addition, Lpd-bound Factor H and FHL-1, by inactivating C3b, would impede formation of the C3 convertases, potentially blocking progression of the complement cascade and generation of the inflammatory anaphylatoxins C3a and C5a. Acquisition of human regulators from serum is a common evasion strategy used by guest on October 4, 2021 by many pathogenic microbes. Factor H and FHL-1 are recruited by bacterial and fungal complement regulator-acquiring surface proteins and are attached via the same domains, that is, SCRs 6–7 and SCRs 18–20. Microbial proteins that bind Factor H and FHL- 1 via these domains include Rck from Salmonella enterica and Gpm1 and Pra1 from C. albicans (23, 35, 44, 45). A second group of microbial surface proteins binds the two human regulators either via SCRs 6–7 (e.g., M proteins and Fba from Streptococcus pyogenes, NspA from Neisseria meningitidis, CRASP-1 from FIGURE 9. Factor H and CFHR1 colocalize at the surface of P. ae ru- B. burgdorferi) or via SCRs 18–20 (e.g., Scl1 from S. pyogenes ginosa, and the attached complement regulators protect the bacteria from and Por1B from Neisseria gonorrheae) of Factor H (30, 46–48). In complement damage. (A) Factor H and CFHR1 surface staining was eval- addition, pathogenic microbes also bind the human terminal uated by electron microscopy. Factor H was directly labeled with gold complement pathway regulator CFHR1. The emerging list of ∅ ∅ particles of 15 nm, and CFHR1 with particles of 45 nm. The two proteins microbial CFHR1-binding proteins include CRASP-3, CRASP-4, P. aeruginosa were bound to strain SG137 either as individual proteins or in and CRASP-5 of B. burgdorferi; Sbi of S. aereus; and Scl1 of combination, and binding was determined by transmission electron microscopy. Factor H (I)andCFHR1(II) bind to the surface of P. ae ru gi nos a and are distributed over the surface. When used together (III), both proteins bind to the same spots (black arrows) and also to unique likely independent bacteria obtained after cultivation in NHS was set to 100%. (D) P. aeru- sites (white arrows). Data shown are representative of three independent ginosa strains SG137, ATCC 27853, and NCTC 10662 were incubated experiments. (B) Serum resistance of the five laboratory P. ae ru gi nos a strains with NHS, HSDFH/FHL-1/CFHR1,orHSDFH/FHL-1/CFHR1 supple- was analyzed by challenging intact bacteria with NHS, used at increasing mented with purified Factor H (100 mg/ml). After incubation, the cells concentrations (0–50%) diluted in Mg-EGTA buffer. After incubation, the were plated on NB agar plates, and the number (CFU) of surviving bacteria cells were plated on NB agar plates, and the number (CFU) of surviving was determined. Survival after 1 h is shown. (E) Preincubation of SCRs bacteria was determined. (C) The relevance of surface-bound Factor H, FHL- 18–20 of Factor H with SG137 reduced bacterial survival in NHS. Bacteria 1, and CFHR1 was analyzed by challenging intact bacteria with complement were incubated with Factor H deletion mutants SCRs 8–11 or SCRs 18–20 active human serum in which the three human regulators were depleted. P. and were thereafter challenged with NHS (10%). After incubation, bacteria aeruginosa strains were incubated in HSDFH/FHL-1/CFHR1 diluted in Mg- were plated on NB agar plates, and the number (CFU) of surviving bacteria EGTA buffer. After incubation, the cells were plated on NB agar plates, and was determined. The mean values of three independent experiments and the number (CFU) of surviving bacteria was determined. To allow a direct SD are presented. Statistical significance of differences was estimated comparison of the survival rates of all five bacterial strains, the fraction of using Student t test. *p # 0.05, **p # 0.01, ***p # 0.001. The Journal of Immunology 4949

S. pyogenes (30, 49, 50). In this study, we show that Lpd binds References human CFHR1 and the major binding region is located within the 1. Lyczak, J. B., C. L. Cannon, and G. B. Pier. 2000. Establishment of Pseudo- C-terminal surface binding region SCRs 3–5. This region shows monas aeruginosa infection: lessons from a versatile opportunist. Microbes Infect. 2: 1051–1060. high level of sequence identity with SCRs 18–20 of Factor H (100, 2. Hassett, D. J., T. R. Korfhagen, R. T. Irvin, M. J. Schurr, K. Sauer, G. W. Lau, 100, and 97%) (10), thus explaining cooperative binding of the M. D. Sutton, H. Yu, and N. Hoiby. 2010. Pseudomonas aeruginosa biofilm two human regulators to Lpd. infections in cystic fibrosis: insights into pathogenic processes and treatment strategies. Expert Opin. Ther. Targets 14: 117–130. Lpd is a multifunctional bacterial protein that, in addition to Factor 3. Sadikot, R. T., T. S. Blackwell, J. W. Christman, and A. S. Prince. 2005. Pathogen- H, FHL-1, and CFHR1, also binds plasminogen. Recruitment of host interactions in Pseudomonas aeruginosa pneumonia. Am. J. Respir. Crit. Care Med. 171: 1209–1223. plasmin(ogen) is used by many pathogens and is a strategy for tissue 4. Trautmann, M., P. M. Lepper, and M. Haller. 2005. Ecology of Pseudomonas penetration. Several microbial proteins bind Factor H, FHL-1, and aeruginosa in the intensive care unit and the evolving role of water outlets as also plasminogen. These include Tuf from P. aeruginosa,CRASPs a reservoir of the organism. Am. J. Infect. Control 33: S41–S49. 5. Nordmann, P., T. Naas, N. Fortineau, and L. Poirel. 2007. Superbugs in the from B. burgdorferi, BpcA from Borrelia parkeri, and M protein coming new decade; multidrug resistance and prospects for treatment of from S. pyogenes (11, 22, 36, 51, 52). Plasminogen is composed of Staphylococcus aureus, Enterococcus spp. and Pseudomonas aeruginosa in five kringle domains each of ∼80 aa in size linked to a protease 2010. Curr. Opin. Microbiol. 10: 436–440. 6. Zipfel, P. F., and C. Skerka. 2009. Complement regulators and inhibitory pro- domain. This human plasma protein also binds fibrinogen and to teins. Nat. Rev. Immunol. 9: 729–740. other ECM proteins. Apparently, plasminogen is recruited by mi- 7. Zipfel, P. F., C. Skerka, J. Hellwage, S. T. Jokiranta, S. Meri, V. Brade, P. Kraiczy, M. Noris, and G. Remuzzi. 2002. Factor H family proteins: on crobial proteins via different domains. Lpd attaches plasminogen complement, microbes and human diseases. Biochem. Soc. Trans. 30: 971–978. via a major binding site localized in kringle domain 4. Streptococcal 8. Rodrı´guez de Co´rdoba, S., J. Esparza-Gordillo, E. Goicoechea de Jorge, surface protein PAM binds human plasminogen via kringle domain M. Lopez-Trascasa, and P. Sańchez-Corral. 2004. The human complement factor H: functional roles, genetic variations and disease associations. Mol. Immunol. Downloaded from 2, and the Streptococcus canis protein SCM binds plasminogen via 41: 355–367. kringle domain 5 (53, 54). Thus, the three bacterial proteins bind to 9. Skerka, C., R. D. Horstmann, and P. F. Zipfel. 1991. Molecular cloning of different kringle domains of plasminogen. Plasminogen bound to a human serum protein structurally related to complement factor H. J. Biol. Chem. 266: 12015–12020. Lpd is accessible for the activator uPA, and, when converted to the 10. Heinen, S., A. Hartmann, N. Lauer, U. Wiehl, H. M. Dahse, S. Schirmer, active protease plasmin, cleaved the synthetic chromogenic sub- K. Gropp, T. Enghardt, R. Wallich, S. Ha¨lbich, et al. 2009. Factor H-related strate S-2251 and also the natural substrate fibrinogen. As plasmin protein 1 (CFHR-1) inhibits complement C5 convertase activity and terminal complex formation. Blood 114: 2439–2447. http://www.jimmunol.org/ degrades ECM components such as fibrinogen, laminin, fibronectin, 11. Kunert, A., J. Losse, C. Gruszin, M. Hu¨hn, K. Kaendler, S. Mikkat, D. Volke, and vitronectin (18), acquisition of plasminogen to the bacterial R. Hoffmann, T. S. Jokiranta, H. Seeberger, et al. 2007. Immune evasion of the human pathogen Pseudomonas aeruginosa: elongation factor Tuf is a factor H surface may facilitate tissue invasion and allow dissemination into and plasminogen binding protein. J. Immunol. 179: 2979–2988. deeper tissue layers. 12. Hong, Y. Q., and B. Ghebrehiwet. 1992. Effect of Pseudomonas aeruginosa P. aeruginosa attaches the human regulators Factor H and CFHR1 elastase and alkaline protease on serum complement and isolated components C1q and C3. Clin. Immunol. Immunopathol. 62: 133–138. simultaneously, and the two human proteins colocalize at the surface 13. Laarman, A. J., B. W. Bardoel, M. Ruyken, J. Fernie, F. J. Milder, J. A. van of intact bacteria. This binding profile suggests that, in addition to Strijp, and S. H. Rooijakkers. 2012. Pseudomonas aeruginosa alkaline protease Lpd, P. aeruginosa uses additional surface proteins to bind Factor H blocks complement activation via the classical and lectin pathways. J. Immunol. 188: 386–393. andCFHR1.Atpresent,thetwoPseudomonas proteins Lpd and Tuf 14. Kharazmi, A., H. O. Eriksen, G. Doring, W. Goldstein, and N. Høiby. 1986. ¨ by guest on October 4, 2021 have been identified to bind Factor H, FHL-1, and CFHR1 (11). The Effect of Pseudomonas aeruginosa proteases on human leukocyte phagocytosis and bactericidal activity. Acta Pathol. Microbiol. Immunol. Scand. [C] 94: 175– relevance of surface-attached human complement regulators was 179. determined by challenging intact bacteria with complement active, 15. Schmidtchen, A., E. Holst, H. Tapper, and L. Bjo¨rck. 2003. Elastase-producing Factor H-, FHL-1–, and CFHR1-depleted human serum, and then Pseudomonas aeruginosa degrade plasma proteins and extracellular products of human skin and fibroblasts, and inhibit fibroblast growth. Microb. Pathog. 34: survival of bacteria was evaluated. In this case, bacterial survival was 47–55. reduced by up to 84%. Addition of Factor H to the depleted serum 16. Kharazmi, A. 1991. Mechanisms involved in the evasion of the host defence by resultedinanincreaseinsurvivalofP. aeruginosa. Thus, Factor H Pseudomonas aeruginosa. Immunol. Lett. 30: 201–205. 17. Marrie, T. J., G. K. Harding, A. R. Ronald, J. Dikkema, J. Lam, S. Hoban, and when attached to the surface protects the pathogenic bacterium from J. W. Costerton. 1979. Influence of mucoidy on antibody coating of Pseudo- complement damage and contributes to serum resistance. The monas aeruginosa. J. Infect. Dis. 139: 357–361. 18. Bergmann, S., and S. Hammerschmidt. 2007. Fibrinolysis and host response in damaging effect of the depleted human serum was strain dependent bacterial infections. Thromb. Haemost. 98: 512–520. and ranged from 84 to 34%. However, the survival rate did not di- 19. Castellino, F. J., and V. A. Ploplis. 2005. Structure and function of the rectly correlate with Lpd surface levels. This suggests that P. aer- plasminogen/plasmin system. Thromb. Haemost. 93: 647–654. 20. Antikainen, J., V. Kuparinen, K. La¨hteenma¨ki, and T. K. Korhonen. 2007. uginosa uses additional bacterial complement regulator-binding Enolases from Gram-positive bacterial pathogens and commensal lactobacilli proteins. Thus, Lpd of P. aeruginosa is a novel bacterial viru- share functional similarity in virulence-associated traits. FEMS Immunol. Med. lence factor expressed at the bacterial surface, which acquires Microbiol. 51: 526–534. 21. Attali, C., C. Frolet, C. Durmort, J. Offant, T. Vernet, and A. M. Di Guilmi. 2008. human complement regulators and consequently protects the Streptococcus pneumoniae choline-binding protein E interaction with bacterium from complement attack. The elucidation of the mech- plasminogen/plasmin stimulates migration across the extracellular matrix. Infect. anisms that P. aeruginosa uses to interact with the human host and Immun. 76: 466–476. 22. Hallstro¨m, T., K. Haupt, P. Kraiczy, P. Hortschansky, R. Wallich, C. Skerka, and identification of a novel P. aeruginosa virulence and immune evasion P. F. Zipfel. 2010. Complement regulator-acquiring surface protein 1 of Borrelia protein identify novel candidates for vaccine development. burgdorferi binds to human bone morphogenic protein 2, several extracellular matrix proteins, and plasminogen. J. Infect. Dis. 202: 490–498. 23. Poltermann, S., A. Kunert, M. von der Heide, R. Eck, A. Hartmann, and P. F. Zipfel. 2007. Gpm1p is a factor H-, FHL-1-, and plasminogen-binding Acknowledgments surface protein of Candida albicans. J. Biol. Chem. 282: 37537–37544. We thank Maria Baumgarten for skillful and expert technical assistance, 24. Klausen, M., A. Heydorn, P. Ragas, L. Lambertsen, A. Aaes-Jørgensen, Rita Walleń for help with electron microscopy, Maria Poetsch for se- S. Molin, and T. Tolker-Nielsen. 2003. Biofilm formation by Pseudomonas aeruginosa wild type, flagella and type IV pili mutants. Mol. Microbiol. 48: quencing Lpd by MALDI-TOF-MS/MS, and Monika von der Heide for 1511–1524. sequencing the lpd gene. 25. Sabarth, N., S. Lamer, U. Zimny-Arndt, P. R. Jungblut, T. F. Meyer, and D. Bumann. 2002. Identification of surface proteins of Helicobacter pylori by selective biotinylation, affinity purification, and two-dimensional gel electro- phoresis. J. Biol. Chem. 277: 27896–27902. Disclosures 26. Ku¨hn, S., and P. F. Zipfel. 1995. The baculovirus expression vector pBSV-8His The authors have no financial conflicts of interest. directs secretion of histidine-tagged proteins. Gene 162: 225–229. 4950 Lpd INTERACTS WITH COMPLEMENT REGULATORS AND PLASMINOGEN

27. Heinen, S., P. Sanchez-Corral, M. S. Jackson, L. Strain, J. A. Goodship, 41. Ebanks, R. O., K. Chisholm, S. McKinnon, M. Whiteway, and D. M. Pinto. 2006. E. J. Kemp, C. Skerka, T. S. Jokiranta, K. Meyers, E. Wagner, et al. 2006. De Proteomic analysis of Candida albicans yeast and hyphal cell wall and associ- novo gene conversion in the RCA gene cluster (1q32) causes mutations in ated proteins. Proteomics 6: 2147–2156. complement factor H associated with atypical hemolytic uremic syndrome. Hum. 42. Henderson, B., and A. Martin. 2011. Bacterial virulence in the moonlight: Mutat. 27: 292–293. multitasking bacterial moonlighting proteins are virulence determinants in in- 28. Oppermann, M., T. Manuelian, M. Jo´zsi, E. Brandt, T. S. Jokiranta, S. Heinen, fectious disease. Infect. Immun. 79: 3476–3491. S. Meri, C. Skerka, O. Go¨tze, and P. F. Zipfel. 2006. The C-terminus of com- 43. Dallo, S. F., T. R. Kannan, M. W. Blaylock, and J. B. Baseman. 2002. Elongation plement regulator factor H mediates target recognition: evidence for a compact factor Tu and E1 beta subunit of pyruvate dehydrogenase complex act as fi- conformation of the native protein. Clin. Exp. Immunol. 144: 342–352. bronectin binding proteins in Mycoplasma pneumoniae. Mol. Microbiol. 46: 29. Skerka, C., and P. F. Zipfel. 2008. Complement factor H related proteins in 1041–1051. immune diseases. Vaccine 26(Suppl. 8): I9–I14. 44. Ho, D. K., H. Jarva, and S. Meri. 2010. Human complement factor H binds to 30. Reuter, M., C. C. Caswell, S. Lukomski, and P. F. Zipfel. 2010. Binding of the outer membrane protein Rck of Salmonella. J. Immunol. 185: 1763–1769. human complement regulators CFHR1 and factor H by streptococcal collagen- 45. Kraiczy, P., C. Skerka, M. Kirschfink, V. Brade, and P. F. Zipfel. 2001. Immune like protein 1 (Scl1) via their conserved C termini allows control of the com- evasion of Borrelia burgdorferi by acquisition of human complement regulators plement cascade at multiple levels. J. Biol. Chem. 285: 38473–38485. FHL-1/reconectin and factor H. Eur. J. Immunol. 31: 1674–1684. 31. Bengtson, S. H., C. Sandeń, M. Mo¨rgelin, P. F. Marx, A. I. Olin, L. M. Leeb- 46. Blom, A. M., T. Hallstro¨m, and K. Riesbeck. 2009. Complement evasion strat- Lundberg, J. C. Meijers, and H. Herwald. 2009. Activation of TAFI on the egies of pathogens: acquisition of inhibitors and beyond. Mol. Immunol. 46: surface of Streptococcus pyogenes evokes inflammatory reactions by modulating 2808–2817. the kallikrein/kinin system. J. Innate Immun. 1: 18–28. 47. Lewis, L. A., J. Ngampasutadol, R. Wallace, J. E. Reid, U. Vogel, and S. Ram. 32. Stover, C. K., X. Q. Pham, A. L. Erwin, S. D. Mizoguchi, P. Warrener, 2010. The meningococcal vaccine candidate neisserial surface protein A (NspA) M. J. Hickey, F. S. Brinkman, W. O. Hufnagle, D. J. Kowalik, M. Lagrou, et al. binds to factor H and enhances meningococcal resistance to complement. PLoS 2000. Complete genome sequence of Pseudomonas aeruginosa PAO1, an op- Pathog. 6: e1001027. portunistic pathogen. Nature 406: 959–964. 48. Shaughnessy, J., L. A. Lewis, H. Jarva, and S. Ram. 2009. Functional compar- 33. Blackmore, T. K., J. Hellwage, T. A. Sadlon, N. Higgs, P. F. Zipfel, H. M. Ward, ison of the binding of factor H short consensus repeat 6 (SCR 6) to factor H and D. L. Gordon. 1998. Identification of the second heparin-binding domain in binding protein from Neisseria meningitidis and the binding of factor H SCR 18 human complement factor H. J. Immunol. 160: 3342–3348. to 20 to Neisseria gonorrhoeae porin. Infect. Immun. 77: 2094–2103. 34. Grosskinsky, S., M. Schott, C. Brenner, S. J. Cutler, P. Kraiczy, P. F. Zipfel, 49. Haupt, K., P. Kraiczy, R. Wallich, V. Brade, C. Skerka, and P. F. Zipfel. 2007. Downloaded from M. M. Simon, and R. Wallich. 2009. Borrelia recurrentis employs a novel Binding of human factor H-related protein 1 to serum-resistant Borrelia burg- multifunctional surface protein with anti-complement, anti-opsonic and invasive dorferi is mediated by borrelial complement regulator-acquiring surface pro- potential to escape innate immunity. PLoS One 4: e4858. teins. J. Infect. Dis. 196: 124–133. 35. Luo, S., S. Poltermann, A. Kunert, S. Rupp, and P. F. Zipfel. 2009. Immune 50. Haupt, K., M. Reuter, J. van den Elsen, J. Burman, S. Ha¨lbich, J. Richter, evasion of the human pathogenic yeast Candida albicans: Pra1 is a factor H, C. Skerka, and P. F. Zipfel. 2008. The Staphylococcus aureus protein Sbi acts as FHL-1 and plasminogen binding surface protein. Mol. Immunol. 47: 541–550. a complement inhibitor and forms a tripartite complex with host complement 36. Schott, M., S. Grosskinsky, C. Brenner, P. Kraiczy, and R. Wallich. 2010. Mo- factor H and C3b. PLoS Pathog. 4: e1000250.

lecular characterization of the interaction of Borrelia parkeri and Borrelia 51. Brissette, C. A., K. Haupt, D. Barthel, A. E. Cooley, A. Bowman, C. Skerka, http://www.jimmunol.org/ turicatae with human complement regulators. Infect. Immun. 78: 2199–2208. R. Wallich, P. F. Zipfel, P. Kraiczy, and B. Stevenson. 2009. Borrelia burgdorferi 37. Seling, A., C. Siegel, V. Fingerle, B. L. Jutras, C. A. Brissette, C. Skerka, infection-associated surface proteins ErpP, ErpA, and ErpC bind human plas- R. Wallich, P. F. Zipfel, B. Stevenson, and P. Kraiczy. 2010. Functional char- minogen. Infect. Immun. 77: 300–306. acterization of Borrelia spielmanii outer surface proteins that interact with dis- 52. Sanderson-Smith, M. L., K. Dinkla, J. N. Cole, A. J. Cork, P. G. Maamary, tinct members of the human factor H protein family and with plasminogen. J. D. McArthur, G. S. Chhatwal, and M. J. Walker. 2008. M protein-mediated Infect. Immun. 78: 39–48. plasminogen binding is essential for the virulence of an invasive Streptococcus 38. de Kok, A., A. F. Hengeveld, A. Martin, and A. H. Westphal. 1998. The pyruvate pyogenes isolate. FASEB J. 22: 2715–2722. dehydrogenase multi-enzyme complex from Gram-negative bacteria. Biochim. 53. Fulde, M., M. Rohde, A. Hitzmann, K. T. Preissner, D. P. Nitsche-Schmitz, Biophys. Acta 1385: 353–366. A. Nerlich, G. S. Chhatwal, and S. Bergmann. 2011. SCM, a novel M-like 39. McCully, V., G. Burns, and J. R. Sokatch. 1986. Resolution of branched-chain protein from Streptococcus canis, binds (mini)-plasminogen with high afﬁnity oxo acid dehydrogenase complex of Pseudomonas aeruginosa PAO. Biochem. J. and facilitates bacterial transmigration. Biochem. J. 434: 523–535.

233: 737–742. 54. Wistedt, A. C., H. Kotarsky, D. Marti, U. Ringdahl, F. J. Castellino, J. Schaller, by guest on October 4, 2021 40. Danson, M. J., K. Conroy, A. McQuattie, and K. J. Stevenson. 1987. Dihy- and U. Sjo¨bring. 1998. Kringle 2 mediates high afﬁnity binding of plasminogen drolipoamide dehydrogenase from Trypanosoma brucei: characterization and to an internal sequence in streptococcal surface protein PAM. J. Biol. Chem. 273: cellular location. Biochem. J. 243: 661–665. 24420–24424. Supplementary Fig. 1: Lpd is present in the outer membrane protein (OMP) preparation of all 13 clinical P. aeruginosa isolates. Lpd is detected in the outer membrane fractions of all tested clinical isolates as a single band of 57 kDa. A fraction containing outer membrane proteins was prepared from the thirteen clinical P. aeruginosa isolates, separated by SDS-PAGE, transferred to a membrane and analyzed by Western blotting using the Lpd antiserum. The secondary antibody used alone did not react with Lpd or any other proteins in the OMP preparation (antibody control, bottom panel).

Preparation of OMPs of P. aeruginosa

P. aeruginosa clinical isolates were grown to exponential phase, then bacteria were washed in

50 mM Tris-HCl (pH 8.0) and pelleted by centrifugaiton. The bacterial pellet was resuspended in 50 mM Tris-HCl (pH 8.0) containing 3 % Empigen (Calbiochem) and protease inhibitors (Complete, Roche). OMPs were extracted by rotating the mixture at 37°C for 2 h. Thereafter, the bacterial cells, stripped of their outer membranes, were centrifuged and the OMP containing supernatants were subjected to SDS-PAGE, transferred to a membrane and Lpd expression was analyzed by Western blotting using a rabbit Lpd antiserum

(1:500) and HRP-conjugated swine anti-rabbit (1:2500). The membrane was developed with

ECL Western blotting detection reagents (Applichem). The secondary antibody alone was used as a negtive control.

Lpd is present in the OMP fraction Lpd was identified in the OMP fraction of all thirteen P. aeruginosa clinical isolates as a singel band of 57 kDa (Fig. 1, upper panel). The secondary antibody did not cross react with

Lpd (Fig. 1, lower panel).

Supplementary Fig. 2: Human serum depleted of Factor H, FHL-1 and CFHR1 is complement active. A, Alternative pathway mediated hemolysis of rabbit erythrocytes with 10

% NHS or 10 % depleted serum. A414 nm, absorbance at 414 nm. The mean values out of three separate experiments are shown and error bars correspond to SD. B, C5b-9 deposition via the alternative (20 %), lectin (1 %) and classical (1 %) pathways. Activation was recorded by following TCC assembly using an anti-C5b-9 mAb and HRP-conjugated rabbit anti-mouse pAb.

Complement activation assays

Normal human serum (NHS) obtained from five healthy volunteers upon informed consent was pooled. Human serum was depleted from Factor H/FHL-1/CFHR1 by affinity chromatography using Factor H antiserum coupled to a mix of protein A-Sepharose and protein G-Sepharose (GE Healthcare) overnight at 4°C as described previously (1). To demonstrate that this depleted serum i.e. HSΔFH/FHL-1/CFHR1 was complement active, complement activity was assayed in a hemolytic assay using rabbit erythrocytes (RabbitER)

(Rockland, BioTrend Chemikalien GmbH, Köln, Germany). RabbitER (108/ml) were incubated with HSΔFactor H/FHL-1/CFHR1 (10 %) diluted in MgEGTA buffer (20 mM

HEPES, 144 mM NaCl, 7 mM MgCl2, and 10 mM EGTA, pH 7.4). All incubations were performed in a final volume of 100 μl. After incubation for 15 min at 37 °C, intact erythrocytes were sedimented by centrifugation and erythrocytes lysis was determined by measuring the release of hemoglobin at 414 nm.

To demonstrate that all three complement pathways were active after depletion of the regulators an ELISA based activation assay was used. The alternative, lectin or classical pathways were activated separately by specific activators and following incubation TCC

(C5b-9) deposition was assayed (2). Microtiter plates were coated with LPS (10 μg/ml)

(activation of the alternative pathway), mannan (100 μg/ml) (activation of the lectin pathway) or IgM (2 μg/ml) (activation of the classical pathway) over night at 4°C, washed four times with PBS-Tween and blocked for 1 h at 37°C with PBS containing 2 % BSA. NHS was diluted in MgEGTA buffer (alternative pathway) or GVB++ buffer (Complement

Technology) (classical/lectin pathways), added to the microtiter plates with activators and incubated for 1 h at 37°C. Heat-inactivated NHS (HiNHS) (56°C, 30 min) was used as a negative control. After washings, complement activation was assayed by determining TCC deposition with a mouse anti-human C5b-9 mAb (Dako) and HRP-conjugated rabbit anti- mouse pAbs. The reaction was developed with 1,2-phenylenediamine dihydrochloride (OPD;

Dako). The absorbance was measured at 492 nm.

Factor H/FHL-1/CFHR1-depleted human serum is complement active

In order to determine that the depleted serum maintains complement activity, this activity was determined by two independent and separate approaches. First the hemolytic activity of the depleted serum was tested using rabbit erythrocytes (RabbitER). RabbitER are lysed when incubated in NHS (Fig. 2A, column 2). Similar, RabbitER incubated in HSΔFactor H/FHL-

1/CFHR1 were lysed, thus demonstrating that the serum complement was still active (Fig. 2A, column 3). However, the lytic activity of the depleted serum is lower as compared to the NHS

(∼40 %). B, In addition, complement activity of the depleted serum was assayed for all three complement pathways using a ELISA based activation assay. Each activation pathway, i.e. the alternative, the lectin or the classical complement pathway was specifically activated and after incubation TCC (C5b-9) deposition was assayed. All three pathways were still active in the depleted serum. However, when activated via the alternative pathway the activity of the serum was 30 % reduced compared to NHS (Fig. 2B). Similarly, the classical pathway activity in HSΔFactor H/FHL-1/CFHR1 was reduced by 24 % and the lectin pathway was not affected.