The secreted Candida albicans protein Pra1 disrupts host defense by broadly targeting and blocking complement C3 and C3 activation fragments

Molecular Immunology xxx (xxxx) xxx–xxx

Contents lists available at ScienceDirect

Molecular Immunology journal homepage: www.elsevier.com/locate/molimm

Full Length Article

The secreted Candida albicans protein Pra1 disrupts host defense by broadly targeting and blocking complement C3 and C3 activation fragments Shanshan Luoa,1,2, Prasad Dasaria,1, Nadine Reihera,b, Andrea Hartmanna, Susanne Jackscha, Elisabeth Wendec, Dagmar Barzd, Maria Joanna Niemiece, Ilse Jacobsene, Niklas Beyersdorff, Thomas Hünigf, Andreas Klosc, Christine Skerkaa, Peter F. Zipfela,b,g,⁎ a

Department of Infection Biology, Leibniz Institute for Natural Product Research and Infection Biology, Hans Knöll Institute, Jena, Germany, Center for Sepsis Control and Care, University Hospital Jena, Germany, c Institute of Medical Microbiology, Medical School Hannover, Germany, d Institute for Transfusion Medicine, Medical Faculty, Friedrich Schiller University Jena Germany, e Microbial Immunology Group, Leibniz Institute for Natural Product Research and Infection Biology, Hans Knöll Institute, Jena, Germany, f University of Würzburg, Institute for Virology and Immunobiology, Würzburg, Germany, g Friedrich Schiller University, Jena, Germany, b

A R T I C L E I N F O

A B S T R A C T

Keywords: Complement evasion Candida albicans C3 proeteolytic cleavage Immune control

Candida albicans the most frequently isolated clinical fungal pathogen can cause local as well as systemic and life-threatening infections particularly in immune-compromised individuals. A better and more detailed understanding how C. albicans evades human immune attack is therefore needed for identifying fungal immuneevasive proteins and develop new therapies. Here, we identified Pra1, the pH-regulated C. albicans antigen as a hierarchical complement inhibitor that targets C3, the central human complement component. Pra1 cleaved C3 at a unique site and further inhibited effector function of the activation fragments. The newly formed C3a-like peptide lacked the C-terminal arginine residue needed for C3a-receptor binding and activation. Moreover, Pra1 also blocked C3a-like antifungal activity as shown in survival assays, and the C3b-like molecule formed by Pra1 was degraded by the host protease Factor I. Pra1 also bound to C3a and C3b generated by human convertases and blocked their effector functions, like C3a antifungal activity shown by fungal survival, blocked C3a binding to human C3a receptor-expressing HEK cells, activation of Fura2-AM loaded cells, intracellular Ca2+ signaling, IL-8 release, C3b deposition, as well as opsonophagocytosis and killing by human neutrophils. Thus, upon infection C. albicans uses Pra1 to destroy C3 and to disrupt host complement attack. In conclusion, candida Pra1 represents the first fungal C3-cleaving protease identified and functions as a fungal master regulator of innate immunity and as a central fungal immune-escape protein.

1. Introduction The frequency of opportunistic fungal infections increases at an alarming rate and is a cause of serious health problems. Candida albicans the most frequently clinical isolated human fungal pathogen can cause local as, well as systemic, life-threatening infections particularly in immune-compromised individuals (Alves et al., 2017; Jong et al.,

2001; Olczak-Kowalczyk et al., 2016). Despite currently available antifungal therapies, C. albicans induced mortality and morbidity remain high (Alonso-Valle et al., 2003; Gudlaugsson et al., 2003; Pappas et al., 2003). Over 75% of patients with systemic candidemia die. Simultaneously, treatment-resistant C. albicans strains are increasing, while vaccine development has remained challenging (Iannitti et al., 2012). A better and more detailed understanding how C. albicans evades human

Abbreviations: Pra1, pH-regulated antigen 1; TCC, terminal complement complex; FHL-1, Factor H like protein 1; C4BP, C4b binding protein; DPBS, Dulbecco's Phosphate Buffered Saline; CR3, complement receptor 3; NHS, normal human serum; C3aL, C3a like protein; C3bL, C3b like protein; PMSF, phenylmethylsulfonyl fluoride; CHAPS, 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate; GFP, green fluorescence protein; ABESF, 4-(2-Aminoethyl)benzenesulfonyl fluoride hydrochloride; Gpm1, phosphoglycerate mutase 1; GPD2, Glycerol-3-phosphate dehydrogenase 2 ⁎ Corresponding author at: Department of Infection Biology, Leibniz Institute for Natural Products Research and Infection Biology, Hans-Knöll Institute, Beutenberg Str. 11a, 07745, Jena, Germany. E-mail address: [email protected] (P.F. Zipfel). 1 Equal contribution of the two first authors. 2 Current address, State Key Laboratory of Molecular Vaccinology and Molecular Diagnostics, School of Public Health, Xiamen University, P.R. China. http://dx.doi.org/10.1016/j.molimm.2017.07.010 Received 18 May 2017; Received in revised form 12 July 2017; Accepted 18 July 2017 0161-5890/ © 2017 Elsevier Ltd. All rights reserved.

Please cite this article as: Luo, S., Molecular Immunology (2017), http://dx.doi.org/10.1016/j.molimm.2017.07.010

Molecular Immunology xxx (xxxx) xxx–xxx

S. Luo et al.

Secreted Pra1 also sequesters host zinc for endothelial cell invasion (Citiulo et al., 2012). Previously we showed that Pra1 binds and complexes human C3 and blocks C3 conversion by the host C3 convertases. Here we identify Pra1 as a C3 cleaving protease that blocks the effector functions of the Pra1 generated C3 activation fragments. Moreover, Pra1 also blocks C3a and C3b action when the effector fragments are generated by the human C3 convertase.

immune attack is of interest and provides a basis to identify new fungal immune evasive proteins to develop new therapies. The human host uses an efficient and sophisticated immune system to protect itself from infectious microbes. The human complement system, as a central component of innate immunity, controls immune homeostasis and recognizes and eliminates infectious agents. Complement, which forms the first defense line of innate immunity, is conserved among many species (Walport, 2001; Zipfel and Skerka, 2009). Complement is relevant for controlling fungal infections and complement deficient animals present with a higher fungal burden (Luo et al., 2009; Mullick et al., 2004; Tsoni et al., 2009). Complement is activated immediately upon infection (Zipfel, 2009; Zipfel and Skerka, 2009) via the alternative (AP), classical (CP), and lectin (LP) pathways. C3 convertases are formed which cleave the central complement protein C3 generate the anaphylatoxin C3a and the opsonin C3b. C3a is a chemoattractant with antifungal and antimicrobial activity (Sonesson et al., 2007). C3a binds to the C3a receptor (C3aR) on immune cells (Heeger and Kemper, 2012; Klos et al., 2013) and initiates inflammatory responses like granule release, generates oxygen radicals, chemotaxis, regulates histamine release, smooth muscle contraction, and increases vascular permeability (Dutow et al., 2014; Klos et al., 2013). C3b, the second product generated by the host C3 convertase, acts as an opsonin and induces opsonophagocytosis (Walport, 2001; Zipfel and Skerka, 2009). In addition, C3b attached to a microbial surface initiates the amplification loop of complement and generates additional C3 convertases. This allows C5 convertase generation and initiates the terminal complement complex (TCC) formation, which results in cell lysis and inflammation (Zipfel and Skerka, 2009). Candida albicans activates all three complement pathways (Cheng et al., 2012; Luo et al., 2009; Zhang and Kozel, 1998). Mice deficient for the central complement components, i.e. C3 or C5, show an increased mortality and infection associated pathology up on candida infection (Mullick et al., 2004; Tsoni et al., 2009). The activated complement system directs host antifungal defense, but the fungus itself controls host complement attack (Luo et al., 2013b). Candida has established multiple evasion mechanisms to escape host complement attack. Candida acquires host complement regulators, such as Factor H, FHL-1, C4BP, CFHR1, plasminogen and vitronectin from human plasma to its surface (Behnsen et al., 2008; Meri et al., 2004; Meri et al., 2002; Zipfel et al., 2008; Zipfel et al., 1999; Zipfel et al., 2011; Zipfel et al., 2007). The human inhibitors attached to the fungal surface retain regulatory functions and assist in immune evasion (Limper and Standing, 1994; Spreghini et al., 1999; Zipfel and Skerka, 2009). Four candida Factor H and FHL1 binding proteins are identified: pH-regulated antigen 1 (Pra1), glycerol-3-phosphate dehydrogenase 2 (Gpd2), high-affinity glucose transporter 1 (Hgt1) and phosphoglycerate mutase1 (Gpm1) (Lesiak-Markowicz et al., 2011; Luo et al., 2013b; Poltermann et al., 2007; Zipfel et al., 2011). Pra1, Gpm1, and Gpd2 also bind plasminogen and allow conversion to plasmin, which degrades C3 and C3b, thereby, blocking host complement (Luo et al., 2013a). Using a proteome approach eight additional candida plasminogen-binding proteins were identified (Crowe et al., 2003). Candida Pra1 also binds the central complement protein C3 and blocks complement activation (Luo et al., 2010). In addition, secreted C. albicans aspartyl proteases (Sap1, Sap2 and Sap3) by degrading and inactivating C3b, C4b and C5 block complement action and effector functions (Gropp et al., 2009). Candida Pra1 is a 299 amino acid multifaceted fungal protein, which is expressed in alkaline pH and which binds human fibrinogen (Casanova et al., 1992). Pra1 is located in the cytoplasm, on the surface of candida yeast and hyphae, and is secreted. Pra1 expression is upregulated upon hyphae induction and Pra1 is concentrated at the tip of hyphae (Luo et al., 2009; Soloviev et al., 2007). Pra1 also binds to human integrin receptors αMβ2 (CR3) and to αXβ2 (CR4) and blocks CR3 or CR4 mediated recognition and signaling (Soloviev et al., 2007; Jawhara et al., 2012). Also, Pra1 binds mouse CD4+ and blocks proinflammatory cytokine INF-γ and TNF-α (Bergfeld et al., 2017).

2. Material and methods 2.1. Proteins, serum and antibodies Human C3, C3b, iC3b, C3c, C3d, C3a, C3a-desArg, C4a, C5a, Factor H, and Factor I were purchased from Complement Technology, Inc. Texas, USA. Polyclonal goat anti-human Factor H, polyclonal goat antihuman C3, and polyclonal rabbit anti-human C3a were purchased from Complement Technology, Inc. Texas, USA. Horseradish peroxidase (HRP)-conjugated rabbit anti-goat and HRP–conjugated goat antirabbit were obtained from Dako Deutschland GmbH, Hamburg. Human serum was collected from five healthy donors, pooled together, and stored at −80 °C until use. The C3a synthetic peptides were purchased from JPT Peptides, Berlin, Germany. Recombinant Pra1, Gpd2, and Gpm1 were expressed in Pichia pastoris strains (Luo et al., 2013a; Luo et al., 2009; Poltermann et al., 2007). 2.2. C. albicans strain, human cell lines and growth conditions The Candida albicans wild type SC5314 (Gillum et al., 1984), Pra1 knockout, and Pra1 overexpression (Citiulo et al., 2012; Luo et al., 2011) strains were cultivated in YPD medium (2% (w/v) glucose, 2% (w/v) peptone, 1% (w/v) yeast extract) at 30 °C. Yeast cells were collected by centrifugation and counted with a hemocytometer (FeinOptik, Bad Blankenburg). 2.3. Pra1 binding to C3 and C3 activation fragments C3 and C3 activation fragments C3b, iC3b, C3c, C3d, and C3a (1 μM, 100 μl/well) were immobilized onto the 96 well microtiter plate (MaxiSorb, Nunc) overnight at 4 °C. After washing, the nonspecific binding sites were blocked with gelatin (0.2% in Dulbecco’s Phosphate Buffered Saline (DPBS)) for 2 h at room temperature (RT). Following washing, 180 nM of Pra1 was added to the immobilized C3 fragments and the mixture was incubated for 1.5 h at RT. Unbound Pra1 was removed by washing with DPBS-T buffer (DPBS containing 0.05% Tween 20). Then polyclonal rabbit Pra1 anti-serum was added and following incubation for 1 h at RT, HRP-conjugated secondary goat rabbit antibody was added for 1 h at RT. In other orientation, Pra1 (180 nM, 100 μl/well) was immobilized onto a microtiter plate (MaxiSorb, Nunc) overnight at 4 °C. After washing, the nonspecific binding sites were blocked as above, and the wells were incubated with 1 μM of each of C3 and C3 activation fragments C3b, iC3b, C3c, C3d, and C3a for 1.5 h at RT. After washing, the polyclonal goat anti-human C3 serum followed by HRP-conjugated secondary rabbit anti-goat antibody was added for 1 h at RT. After addition of 3,3′,5,5′ Tetramethylbenzidine (TMB, eBioscience, Frankfurt, Germany), the reaction was stopped by addition of 2 M H2SO4. Absorbance was measured at 450 nm (SpektraMax 190, Molecular Devices). 2.4. Biolayer interferometry The binding affinity of Pra1 to C3, C3a or C3b was evaluated by biolayer interferometry in a single channel BLItz system, (Forte Bio, Menlo Park, CA). Ni(II)-NTA biosensors were hydrated for at least 10 min in DPBS with gelatin (0.01%) and loaded with recombinant (His)6-tagged Pra1. After washing the tip briefly (30 s) to remove nonspecifically 2

Molecular Immunology xxx (xxxx) xxx–xxx

S. Luo et al.

2.7. Detection of Interleikin 8 (IL-8)

bound Pra1, C3, C3a or C3b were added as analytes and binding was followed for 200 s. C3 was used at concentrations ranging from 150, 312, 625, 1250, and 2500 nM, C3a at concentrations 150, 312, 625, 1250 or 2500 nM and C3b at concentrations 150, 300, 625 1250 and 2500 nM. For each concentration complex formation as well as dissociation was followed for 200 s. For C3a interaction with Pra1, C4a and C5a were used as non-binding controls. In case of C3 or C3b interaction, C3 or C3b were heat inactivated at 95 °C for 10 min and denatured C3 or C3b were used as non-binding controls. The values were subtracted from buffer blank and the interaction affinity KD values were determined by fitting the data to a 1:1 model algorithm using BLITZ software. The graphs were plotted using Excel MicrosotOffice2010.

2.5.

125

Isolated 2 × 106 neutrophils were resuspended in RPMI 1640 in 24 well plate (nunc). The neutrophils were stimulated with either C3a with increasing concentration Pra1 (22.5, 45, 90, and180 nM) or C5a in 1mland incubated at 37 °C in 5% CO2 incubator for 10 h. The neutrophils were centrifuged (200g for 5 min) and the supernatants were analyzed for secreted IL-8 levels using IL-8 detection kit according to the manufactureŕs instructions (Immotool).

2.8. Neutrophil migration assays Purified neutrophils were stained with Calcein AM (Invitrogen) following the manufactureŕs instructions and resuspended in RPMI 16040 without phenol red supplemented with 10% FCS. Transwell inserts (3 μm, polycarbonate membrane, Corning Lise Science) were filled with neutrophil suspension, and the inserts were placed into receiver plate filled with DPBS containing C3a complexed with increasing concentration of Pra1 (13, 26, 52, 104, 208, and 416 nM). After 1 h, the inserts were removed from receiver plate and the number of migrated neutrophils were quantified by measuring Calcein AM fluorescence signals (Klos et al., 1992).

I-hC3a binding on HEK cells stably expressing the human C3aR

To determine whether candida Pra1 as a C3a ligand interferes with C3a binding to the human C3a receptor, “competitive” binding studies were performed on the human embryonic kidney cell line HEK293 (ATCC CRL 1573) stably transfected with the human C3a receptor (C3aR; HEK293pQCXINhC3aR, clone #12576), as described (Johswich et al., 2006). The cells were cultured in Dulbecco’s modified Eagle’s medium/Ham’s F-12 medium containing 10% fetal calf serum (and G418 for permanent selection). Binding assays were carried out in HAGCM (20 mM HEPES, pH 7.4, 125 mM NaCl, 5 mM KCl, 1 mM CaCl2, 1 mM MgCl2, 0.25% BSA, 0.5 mM glucose) in a total volume of 50 μl per well of a microtiter plate (Greiner, Essen, Germany). Tracer concentrations (0.12 nM) of 125I-labeled 125I-C3a (PerkinElmer Life Sciences) were pre-incubated for 30 min at RT with increasing concentrations (0.18, 0.18, 1.8, 18, 180, 1800 nM) of Pra1, or Gpm1 as negative control. To allow specific C3a-binding, preformed 125IC3a:Pra1 complexes and controls were incubated for 90 min at 4 °C with 2 × 106 of the C3aR expressing HEK cells. Forty μl of the mixture was filtered using Multiscreen-HTSTM filter plates (Millipore, Billerica, MA) and washed twice with 100 μl of HAG-CM. The filter plates were pretreated with 2% protamine sulfate overnight at 4 °C and extensively washed with HAG-CM before use. After filtration the plates were dried, and 50 μl of MicroScintO (PerkinElmer Life Sciences) per well were added. Radioactivity bound to the cells on the filter was determined using a TopCount NXT (Canberra-Packard, Dreieich, Germany).

2.9. Candida Pra1 is a C3 cleaving protease C3 (108 nM) was incubated with recombinant candida Pra1 (90, 180, and 360 nM) in DPBS (50 μl reaction) or with Gpm1 (500 nM), which was used as a control, for 1 h at 37 °C. The reaction was stopped either by addition of reducing SDS loading buffer (Roti®-Load 1) for detection of C3bL molecule, or by addition of non-reducing SDS loading buffer (Roti®-Load 2) for detection of C3aL molecule. Then the reaction mixture was boiled for 10 min at 95 °C and subjected to SDS-PAGE (10%). The separated proteins were transferred to a nitrocellulose membrane and the membrane was probed with polyclonal goat antihuman C3 or polyclonal rabbit anti-human C3a serum, followed by HRP-conjugated rabbit anti-goat or HRP-conjugated goat anti-rabbit IgGs. C3 (180 nM) was incubated with equal concentration of Pra1 (180 nM) (30 μl reaction) in DPBS and at the indicated time points, the reaction was stopped by addition of 10 μl Roti®-Load 1. For dose-dependent cleavage, C3 was incubated with Pra1 at increasing concentrations (22.5, 45, 90, 180 and 360 nM) in DPBS for 1 h at 37 °C. Afterwards the reaction mixture was separated by SDS-PAGE, proteins were visualized either by Silver- or Coomassie staining. For comparing the mobilities of the C3a-like peptide (C3aL) generated by Pra1 and C3a generated by the endogenous C3 convertase the fragments were separated by SDS-PAGE (15%) to allow direct comparison of the mobilities of the two fragments. Both C3aL and C3a were identified with polyclonal rabbit C3a anti-serum.

2.6. Fura2-AM assay for the determination of the intracellular Ca2+ response in RBL2H3-cells stably expressing the human C3aR To determine whether candida Pra1 blocks C3aR-dependent cellular responses, the effect of Pra1 on cytosolic Ca2+ was monitored in the rat basophilic leukocyte cell line RBL-2H3 (CRL 1593; ATCC, Manassas, VA; a gift from Dr. M. Oppermann, Göttingen, Germany) stably transduced by us with the pEFBOShC3aR vector (clone #12078). The C3aRexpressing RBL cells (Settmacher et al., 2003) were cultured in Dulbecco’s modified Eagle’s medium containing 10% fetal calf serum (and G418 for permanent selection). Human C3a (Advanced Research Technologies, San Diego, CA) was preincubated with Pra1 (3.6 nM), or Gpm1 (5.5 nM) as negative control for 30 min at RT. Then, RBL cells (2 × 105 cells/ml) were challenged at 35 °C with the preformed C3a:Pra1 complexes or with Gpm1 preincubated with Pra1, used as a negative control, and then C3aR-dependent Ca2+ release was continuously recorded for 1 min. The Ca2+ measurement was performed as described (Klos et al., 1992) using Fura2/AM (Calbiochem, Bad Soden/Taunus, Germany) as indicator and the Luminescence Spectrometer LS 50 B (Perkin Elmer, Beaconsfield, UK). Fura-2/AM–loading and Ca2+ measurement (calculating and depicting the maximal increase after stimulation) were essentially performed as described earlier (Klos et al., 1992).

2.10. Inhibition of Pra1 mediated cleavage of C3 by protease inhibitors To define which type of protease, Pra1 mediated cleavage of C3 was analyzed in the presence of various protease inhibitors. Recombinant Pra1 (180 nM) was combined with C3 (54 nM) for 1 h at 37 °C in the presence of various proteases inhibitors. The protease inhibitors tested include phenylmethylsulfonyl fluoride (PMSF, Roth), E64 (Sigma), 4(2-Aminoethyl)benzenesulfonyl fluoride hydrochloride (AEBSF, Sigma), Phenanthroline (AppliChem), EDTA (10 mM) or protease inhibitor Complete (EDTA-free, Roche, Mannheim, Germany). After incubation, the reaction was stopped by addition of Roti®-Load 1, then the proteins were separated by SDS-PAGE, transferred to a membrane, and Pra1 generated C3 fragments were evaluated by Western blotting using polyclonal rabbit C3d antiserum. 3

Molecular Immunology xxx (xxxx) xxx–xxx

S. Luo et al.

These mixtures (10 μl) were added 1 × 105C. albicans resuspended in 40 μl of Tris (10 mM) and the cells were incubated for 2 h at 30 °C (Sonesson et al., 2007). After incubation, the candida yeast cells were harvested, diluted in Tris-HCl (1:100), and then 10 μl of the diluted sample was spread on YPD-agar plate and incubated for 2 days at 30 °C. Then C. albicans colonies were counted. To analyze whether Pra1 protects candida and blocks the antifungal activities, candida yeast cells (1 × 105 in 50 μl) were treated with C3a, C3a-desArg or C4a (each at 2.5 μM) and Pra1 used at concentration ranging from 0.6 to 2.6 μM then the cells were incubated for 30 °C for 2 h. After incubation, dead candida yeast cells were stained with propidium iodide and the cells were immediately analyzed by flow cytometry (LSR II BD). For laser scanning microscopy, after incubation, the cells were stained with calcofluor and propidum iodide for 10 min and then layered on poly-L-lysine coated coverslips. The coverslips were mounted on glass slide and the images were obtained with LSM 710, (Zeiss, Germany) using ZEN 2009 software.

2.11. N-terminal sequencing of Pra1 cleaved C3 To localize the cleavage site of Pra1 in the C3 protein, Pra1 (66) was added to C3 (1800 nM) in DPBS (30 μl). Following incubation for 1 h at 37 °C, the reaction was stopped by addition of Roti-Load 1 (10 μl), then the mixture was boiled for 10 min at 95 °C, and subjected to SDS-PAGE (7%). Upon transfer to a PVDF membrane, the membrane was stained with coomassie blue R250 (1% freshly prepared in 40% methanol/1% acidic acid) for 2 h at room temperature to identify the position of the bands. Then the membrane was destained with methanol (50%) and the membrane was washed three times with degassed water. Then the 110 kDa band was cut out and subjected to N- terminal protein sequence analyses (Alphalyse, Denmark). 2.12. Binding assay C3 (55 nM, 50 μl/well) was immobilized onto the microtiter plate (MaxiSorb, Nunc) overnight at 4 °C. After washing, nonspecific binding sites were blocked with gelatin (0.2% in DPBS) for 2 h at RT. After blocking, Factor H (32 nM) together with increasing amount of Pra1 (90, 180, 360 nM) was added to immobilized C3 and incubated for 1.5 h at RT. Wells were washed with DPBS-T buffer (DPBS containing 0.05% Tween 20) and polyclonal goat anti-Factor H was added and incubated for 1 h at RT. After washing with DPBS-T, HRP-conjugated rabbit anti-goat as a secondary antibody was added and incubated for 1 h at RT. After addition of TMB, the reaction was stopped by 2 M H2SO4. Absorbance signals were measured at 450 nm in a microtiter plate reader (SpektraMax 190, Molecular Devices).

2.16. C3 cleavage by C. albicans The C. albicans wild type WT SC5314 was cultivated overnight in YPD medium (10 ml), and the candida cells were pelleted by centrifugation and washed three times with DPBS. 20 × 106 candida cells were incubated with C3 (135 nM) of in 40 μl at 37 °C with continuous shaking at 400 rpm. After incubation, the cells were again pelleted by centrifugation for 1 min at 16,000g, the supernatant was recovered, and Roti®-Load 1 buffer was added. C3 cleavage was detected by Western blot analysis using polyclonal goat anti-human C3 as above.

2.13. Cofactor activity

2.17. Human neutrophil isolation

To analyze whether candida Pra1 generated C3b-like molecule (C3bL) is cleaved by Factor I in presence of Factor H, C3 (108 nM) was incubated with Pra1 (45, 90, and 180 nM) in the presence of Factor H (64 nM) and Factor I (0.05 μg) in DPBS (50 μl reaction volume) for 1 h at 37 °C. In addition, C3 (108) was incubated with candida Pra1 alone (180 μg) or candida Gpm1 (270 nM) in DPBS (50 μl reaction system) for 1 h at 37 °C. After incubation, the reaction was stopped by addition of Roti®-Load 1 (10 μl), boiled for 10 min at 95 °C and then separated by SDS-PAGE. The cleavage products were analyzed by Western blotting using polyclonal goat C3 anti-serum.

Human polymorphonuclear lymphocytes (PMN) were isolated from heparinized blood of healthy volunteers as previously described (Dasari et al., 2011). In brief, 1 vol of dextran T 500 (Carl Roth GmbH, Germany; 3%, in isotonic salt solution, pH 7.4) was added to 5 vols whole blood, and cells were allowed to sediment in tilted 15 ml centrifugation tubes for 20 min at RT. The erythrocyte-depleted supernatant was centrifuged for 20 min at 200g, and the pellet was resuspended in Hank’s balanced salt solution (10 ml; HBSS, Lonza Cologne, GmbH, Germany). The neutrophils were layered on top of 2 ml Ficoll-paque™ PLUS gradients (GE Healthcare, Germany), then centrifuged for 20 min at 400g, at 20 °C. The residual erythrocytes contaminating cell pellets were lysed in a buffer containing NH4Cl (150 mM), KHCO3 (10 mM), and EDTA (10 mM), pH 7.4. Cells were again centrifuged (200g, 10 min), washed, suspended in PRMI 1640, and kept on ice. The purity of the human neutrophils was determined by flow cytometry based on CD16 and CD66b surface expression. Neutrophils used in all our experiments were more than 97% pure.

2.14. Cleavage of C3aL by candida Pra1 To analyze whether Pra1 processes C3aL, C3 (108 nM) was incubated with candida Pra1 (used at 45, 90, 180, 360, 900 and 1800 nM) for 1 h at 37 °C. Similarly, Pra1 used at different concentrations was added to C3a (555 nM) and then the mixture was incubated for 1 h at 37 °C. Upon addition of non-reducing loading dye, the mixture was separated by SDS-PAGE (15%) and then transferred to nitrocellulose membrane. As a control, C3a was incubated with candida Gpm1 (1350 nM). The blot was probed with polyclonal rabbit C3a anti-serum followed by corresponding secondary antibody.

2.18. FITC labeling of C. albicans The candida Pra1 WT, Pra1 ko, and Pra1 overexpression strain (Pra1+++) strains were grown overnight in YPD medium (10 ml), the cells were harvested and washed with DPBS. Hundred million candida cells were stained with 1 ml of 3 mM Fluoresceine iosthiocynate (FITC) for 20 min at 37 °C while shaking at 400 rpm. The cells were thoroughly washed with DPBS and kept on ice until use.

2.15. C. albicans survival analyses The Candida albicans wild-type strain, SC5314, was cultivated overnight in 10 ml yeast extract-peptone-dextrose (YPD), and 200 μl of an overnight culture was transferred to fresh YPD medium (10 ml) and further cultivated to mid-logarithmic phase at 30 °C. Candida yeast cells were harvested and counted using a hemocytometer (Fein-Optik, Bad Blankenburg). The candida yeast cells were washed three times and resuspended in 10 mM Tris, pH 7.4. To analyze C3aL mediated antifungal activity, C3 (108 nM) was first incubated with increasing amounts of candida Pra1 (45, 90, 180, 360, 900, and 1800 nM or candida Gpm1 (1350) in DPBS (50 μl reaction volume) for 1 h at 37 °C.

2.19. Live cell imaging The candida strain expressing GFP was grown overnight in YPD medium (10 ml), the candida cells were pelleted by centrifugation and washed with DPBS. NHS (5%, 200 μl) was preincubated with Pra1 (900 nM) for 1 h at 37 °C. After incubation, human neutrophils and GFP expressing candida cells were mixed 1:3 ratio with NHS preincubated for 1 h with Pra1. Live cell imaging was performed by LSM 710 (Zeiss, 4

Molecular Immunology xxx (xxxx) xxx–xxx

S. Luo et al.

Germany) using ZEN 2009 software.

3.2. Pra1 is a C3 cleaving protease

2.20. C3 deposition, phagocytosis and candida killing

To analyze if Pra1 cleaves C3, Pra1 was incubated with C3, and then proteins were separated by SDS-PAGE and transferred to a membrane. In this case, a new ca. 110 kDa band, α’L appeared (Fig. 2A, lane 2), indicating that Pra1 cleaves the α chain of C3. This cleavage was specific for Pra1, as Gpm1, a related candida immune evasion protein also expressed in Pichia cells, did not generate this 110 kDa band (Fig. 2A, lane 3). Pra1 mediated C3 cleavage was time- and dose dependent ((Fig. 2B, lanes 1–6, and Supplementary Fig. 1B, lanes 1–6; Fig. 2C, lanes 1–6, and Supplementary Fig. 1B, lanes 1–5). The shorter cleavage fragment of ca. 9.0 kDa was identified when cleaved C3 was separated by high percentage SDS-PAGE (15%) (Fig. 2C, lanes 2–4), and the cleavage was dose dependent. Thus Pra1 is a C3 cleaving protease. Cleavage was specific for Pra1 as Gpm1, a related candida protein also expressed by Pichia cells, did not cleave C3. Pra1 did cleave neither C3b nor iC3b nor C3c (Supplementary Fig. 2A, 2B, and 2C).

The C. albicans wild type WT SC5314, Pra1 knock out (ko) and overexpression strain (Pra1+++) were grown overnight in YPD medium (10 ml), candida cells were pelleted by centrifugation and washed with DPBS. Pra1 or BSA (450 nM) was incubated with increasing percentage of NHS increasing concentration of Pra1 or BSA were added to NHS (5%, 200 μl) and pretreated for 1 h at 37 °C. Then, 5 × 106 candida cells were added to Pra1–NHS and the cells were kept for 20 min at 37 °C while shaking at 400 rpm C. albicans WT, ko, and (Pra1+++) cells (5 × 106) were incubated with 10% NHS for 15 min. After incubation, the cells were washed three times with DPBS, and C3 deposited on the candida surface was detected with a polyclonal goat anti-human C3 serum (Comptech) followed by Alexa Fluor 647-conjugated rabbit anti-goat IgG. Fluorescence of 10,000 candida cells were measured with flow cytometry (BD LSR II) and median fluorescence intensity of each candida population was calculated using Flowjow software. For phagocytosis by PMN, Pra1 or BSA was incubated with in HNS as above and 2 × 106 freshly isolated human PMN and 6 × 106 FITC-labeled candida cells were mixed with either Pra1 treated NHS or BSA treated NHS and incubated for 20 min at 37 °C while with agitation at 400 rpm. PMN and C. albicans WT, ko, and (Pra1+++) cells (5 × 106) were incubated with 5% NHS for 15 min. The reaction was stopped by addition of ice-cold paraformaldehyde (4%, 100 μl). The percentage of phagocytosis was analyzed by flow cytometry (BD LSR II) measuring fluorescence of 10,000 gated neutrophils. For candida killing assay, the assay was adopted to candida as previously described (Dasari et al., 2011). 3 × 106 freshly isolated PMN and 1 × 106 candida cells were mixed with Pra1 treated NHS and the cells were incubated for 5 h min at 37 °C with agitation 400 rpm. After incubation, 100 μl of sample was withdrawn and mixed with 3-[(3-cholamidopropyl) dimethylammonio]-1-propanesulfonate (CHAPS) (100 μl, 1% solution). After 5 min incubation at 37 °C, 800 μl of water was added for 3 min at 37 °C to lyse human neutrophils and to release candida cells. Finally, candida cell were plated on YPD-agar plate and the plates were incubated for 3 days at 30 °C. Colonies representing surviving, or recovered C. albicans cells were counted.

3.3. Pra1 cleaves human C3 at a unique site Pra1 treated C3 was sequenced (Supplementary Fig. 2D) and the Nterminal sequence of the α’L band was xLGxAx which match with the sequence H743LGLAR748 in the α chain of C3 (bold letters matching residues) (Fig. 2D, upper panel). Thus, Pra1 cleaved C3 between Ser742 and H743 and the cleavage site was positioned six residues upstream of the site used by the endogenous C3 convertase (Fig. 2D, lower panel). In analogy to C3b, the 110 kDa fragment of the α chain of C3 with the six residue extension was termed α’L, the shorter fragment C3aL. C3aL generated by Pra1 has a higher mobility than C3a (Supplementary Fig. 2E, lane 1 vs lane 3). 3.4. PMSF inhibited Pra1 mediated C3 cleavage Pra1 has a metalloprotease like motif and we asked whether protease inhibitors inhibit Pra1 cleavage of C3. PMSF, as well as a protease inhibitor cocktail, blocked C3 cleavage by Pra1. PMSF inhibited Pra1 cleavage dose dependently and at 50 μM blocked C3 cleavage completely (Fig. 2E, lanes 3–6, and Supplementary Fig. 3A). In contrast, neither AEBSF, E64, EDTA, nor Phenanthroline blocked Pra1 mediated C3 cleavage (Supplementary Fig. 3B, 3C, 3D, and 3E). Based on the effects that neither the metalloprotease inhibitor EDTA nor Phenanthroline blocked Pra1 cleavage of C3, thus, we conclude that Pra1 is no typical metalloprotease.

3. Results To investigate whether C. albicans directly inactivates the central human complement component C3, C3 was added to candida yeast cells. Following incubation, cleavage products were identified by Western blotting and a new 110 kDa band, α’ Like (α’L) appeared (Fig. 1A). The intensity of this band increased with time (Fig. 1A, lane 2–6). To identify the fungal C3 cleaving protease, we searched for protease like motifs in known candida proteins. The multifunctional fungal immune evasion protein Pra1 comprised the sequence H178RFWH182 which resembled a metalloprotease motif HExxH (Citiulo et al., 2012; Rawlings and Barrett, 1995). Pra1 binds to C3 and C3b (Luo et al., 2010) and when tested for binding to C3 activation fragments by ELISA, recombinantly expressed Pra1 bound to immobilized C3, C3a, C3b, iC3b and C3d. (Fig. 1B). Binding was confirmed in the other orientation with Pra1 immobilized (Supplementary Fig. 1 A).

3.5. C3aL has antifungal activity and is processed by Pra1 To analyze the antifungal activity of Pra1 generated C3aL, candida yeast cells were challenged with C3 treated with Pra1. The mixture with C3aL displayed antifungal activity, and 60% of fungi were damaged when Pra1 was used at 180 nM (Fig. 3A). Pra1 used at 360 nM or higher inactivated C3aL and blocked this antifungal activity (Fig. 3A). To directly demonstrate that Pra1 processes C3aL and to exclude that Pra1 blocks the antifungal activity of C3aL by complex formation, C3 treated with Pra1 was separated at high percentage SDS-PAGE and C3aL generation was followed. In this case, C3aL generation first increased with increasing levels of Pra1 (Fig. 3B, lanes 2–4), and at higher Pra1 levels C3aL intensity decreased (Fig. 3B, lanes 5–7). C3aL was absent when Pra1 was used at 1800 nM (Fig. 3B, lane 7). Thus Pra1 processed and inactivated C3aL but not C3a (Fig. 3C, lanes 2–5). To localize the Pra1 binding regions in C3aL/C3a linear peptides were synthesized, spotted on the membrane and tested for Pra1 binding. Pra1 bound to two regions, the first binding region i.e. S706CQRRTRF713 was represented by three peptides and the second T732ELRRQH738 by four peptides (Fig. 3D). Each of the two linear binding regions is contained in the C3aL and in C3a (Fig. 3E).

3.1. Pra1 binds to C3 with high affinity The binding intensities of human C3, C3a or C3b to Pra1 were determined by biolayer interferometry. C3 bound with an affinity of 361 nM, C3a with 292 nM, and C3b bound with 2.81 μM (Fig. 1C, D, and E). C3 and C3a showed comparable affinities but displayed different rates of association and dissociation. C3a:Pra1 complexes formed rapidly (Ka 4.7 × 104 1/M sec) and dissociated with a faster rate constant (kd 7.9 × 10−4/sec). 5

Molecular Immunology xxx (xxxx) xxx–xxx

S. Luo et al.

Fig. 1. Candida cleaves C3 and Pra1 is a fungal C3 and C3a binding protein. (A) Intact candida yeast cells cleave C3. C. albicans were incubated with C3, and C3 cleavage fragments were identified by Western blotting. A new 110 kDa α’-L band was detected and appearance was time dependent (lanes 2–5). One representative figure of three independent experiments is shown. (B) Candida Pra1 binds to C3 and to C3 activation fragments. C3, C3a, C3b, iC3b, C3c or C3d were immobilized onto a microtiter plate. Pra1 was added, and bound Pra1 was detected with rabbit Pra1 anti-serum. (C) C3 binds to Pra1 with high affinity. C3 (150, 312, 625, 1250 and 2500 nM) binding to NTA coupled Pra1 was evaluated by biolayer interferometry. C3 bound to Pra1 with an affinity of 361 nM. Heat inactivated C3 (95 °C for 5 min) did not bind to Pra1 (bottom line). (D) Binding of C3a (150, 312, 625, 1250 and 2500 nM) to Pra1 was evaluated by the same approach. C3a bound to Pra1 with an affinity of 270 nM. C4a and C5a did not bind to Pra1 (bottom lines). (E) Binding of C3b (150, 312, 625, 1250 and 2500 nM) to immobilized Pra1 had an affinity of 2810 nM. Heat inactivated C3b, (95 °C, 5 min) did not bind to Pra1 (bottom line). Panel B represent mean values ± SD of three separate experiments.

Fig. 2. Candida Pra1 is a C3 cleaving protease. (A) Candida Pra1 cleaves human C3. C3 was incubated with Pra1, and the proteins were separated by SDS-PAGE; under reducing conditions, C3 cleavage fragments were identified by Western blotting. Intact C3 was represented by the α (117 kDa) and the β chain (75 kDa) (lane 1). An additional α’L band of ca. 110 kDa appeared when C3 was incubated with Pra1 (lane 2). C3 incubated with Gpm1, a second candida immune evasion protein, remained intact (lane 3). (B) Time-dependent Cleavage of C3 by Pra1. C3 (180 nM) was incubated with Pra1 (180 nM),for different time and C3 cleavage products were analyzed by silver staining. Pra1 cleaved C3, and α’L band of 110 kDa appeared in a time dependent manner (lanes 2–6). (C) Dose dependent cleavage of C3 by Pra1. C3 (180 nM) was incubated with Pra1 (22.5, 45, 90, 180 and 360 nM) in DPBS for 1 h at 37 °C, and C3 cleavage products were analyzed by silver staining. Pra1 cleaved C3, and the intensity of the new 110 kDa α’L band correlated with Pra1 levels (lanes 2–6). (D) Pra1 cleaves C3 and generates a shorter C3a-like peptide. C3 (108 nM) was incubated with Pra1 (90, 180, and 360 nM). The reaction mixtures were separated by SDS-PAGE (15%) under non-reducing conditions, and the cleavage products were analyzed by Western blotting. The C3a specific antibody identified C3aL as a band of high mobility of (arrow head). (E) Pra1 cleaves C3 at a unique site. Pra1 cleaves the α chain of C3 between residues Ser742 and His743. The Pra1 cleavage site is shown by black vertical arrow. This cleavage site is distinct from that of the endogenous C3 convertase, which cleaves between Arg748 and Ser749 and shown by the grey arrow. PMSF, the protease inhibitor blocked Pra1 cleavage of C3. C3 (108 nM) was incubated with Pra1 (180) in the presence of PMSF. C3 cleavage products were detected with goat anti-C3 serum. The position of the α’L band generated by Pra1 is indicated by the arrowhead (lane 2). In the presence of PMSF the α’L band disappeared and the effect was dose dependent (lanes 3–6). One representative figure of three independent experiments is shown in panel A, B, C, and E.

6

Molecular Immunology xxx (xxxx) xxx–xxx

S. Luo et al.

Fig. 3. Pra1 binds to C3aL and blocks the effector functions of the short activation fragments. (A) Pra1 generated C3aL has antifungal activity. C3 (108 nM) was incubated with Pra1 (45, 90, 180, 360, 900, and 1800 nM) for 1 h at 37 °C. Then the reaction mixtures (10 μl) were diluted in Tris-HCl (10 mM, pH, 8.0, 40 μl) and added to C. albicans SC5314 (5 × 104 cells). At low Pra1 concentrations, i.e. 45, 90, and 180 nM fungal survival was decreased, i.e. Pra1 generated C3aL had antifungal activity. However, fungal survival increased at Pra1 levels of 180 nM and higher showing that Pra1 blocked C3aL mediated antifungal activity. Gpm1, a related candida immune evasion protein did not block C3a induced antifungal activity. Antifungal C3a was used as a control (B) Candida Pra1 processes and degrades C3aL. C3 (108 μg) was incubated with Pra1 (45, 90,180, 360, 900, and 1800 nM) for 1 h at 37 °C. Then C3 cleavage products were analyzed by Western blotting using rabbit antiC3a serum. At low Pra1 levels i.e. 45, 90, and 180 nM C3aL was generated (lanes 2–4) and the intensity of the C3aL band decreased with increasing Pra1 levels (lanes 5–7). No C3aL band was detected at Pra1 levels of 1800 nM (lane 7). In the presence of Gpm1 no C3aL band appeared (lane 8). (C) Candida Pra1 does not cleave C3a, generated by the endogenous C3 convertase. C3a (555 nM) was incubated with Pra1 for 1 h at 37 °C. Again the reaction mixtures were separated by SDS-PAGE under non-reducing conditions, transferred to a membrane and developed with a polyclonal rabbit anti-C3a serum. Pra1 (used at 180, 360, 900, and 1800 nM) did not cleave human C3a (lanes 2–5). (D) Localization of the linear Pra1 interacting residues in C3aL and C3a. Linear peptides, with a length of 13 residues and an overlap of 10 residues, which cover the complete C3a sequence, were synthesized and spotted onto a membrane. Then the membrane was probed with Pra1. Following washing, bound Pra1 was detected with HRP-coupled penta-His antibody. Pra1 bound to three consecutive peptides, which represent the region S706CQRRTF7713 (numbering according to the C3 sequence). Pra1 bound to a second stretch of four consecutive peptides, representing the T732ELRRQH738 (E) Sequence of C3a peptide and the linear Pra1 binding motifs are highlighted in blue. The C-terminal six residues which are contained in the C3a peptide and which are absent in the C3aL peptide are underlined. The data shown represent the mean ± standard deviation of three separate, independent experiments. One representative experiment of three independent experiments is shown in panel B, C, and D. Figure A represents mean ± SD of three independent experiments.

cleavage. C3 and Pra1 were combined with Factor I and Factor H, and following incubation, proteins were separated by SDS-PAGE. The 110 kDa α’L band disappeared, and three bands of 68, 46 and 43 kDa were appeared (Fig. 4B, lane 3–5). Thus Pra1 generated C3bL was processed by host regulators Factor I and Factor H.

3.6. Pra1 generated C3bL is processed and inactivated by human regulators We next asked whether Pra1 generated C3bL is accessible and inactivated by host regulators similar to the inactivation of host C3 convertase generated C3b by Factor H and Factor I. First, binding of Factor H to immobilized C3bL was tested in the presence of Pra1. Factor H did not bind to C3 but bound to newly generated C3bL. Factor H binding increased with increasing Pra1 levels (Fig. 4A). Next we tested if Factor H bound to C3bL acts as a cofactor and assists in C3bL

3.7. Pra1 inhibits C3a effector functions Pra1 also binds to C3a and to C3b, the fragments generated by the Fig. 4. Pra1 generated C3bL is processed and inactivated by host complement inhibitors. (A) Factor H binds to C3 in the presence of Pra1. C3 (54 nM) was immobilized onto the microtiter plate and after blocking Factor H (32 nM) together with Pra1 (90, 180 and 360 nM) was added. After incubation, bound Factor H was detected with a goat Factor H anti-serum. In the presence of Pra1, Factor H bound to C3bL, and this binding was dose dependent. Factor H did not bind to immobilized C3 (column 4) and Factor H antiserum did not cross react with C3 (column 5). Panel A represents the mean ± SD of three separate experiments. (B) C3bL generated by Pra1 is processed by human Factor I. Purified C3, Factor H and Factor I were incubated with Pra1 for 1 h at 37 °C. Then the mixtures were separated by SDS-PAGE, transferred to a membrane and C3 cleavage products were identified by Western blotting. Pra1 cleaved C3 and generated C3bL (lane 2). In presence of Factor H and Factor I, Pra1 generated C3bL was further processed and shorter cleavage fragments of 68, 46 and 43 kDa were identified (lanes 3–5). Figure B shows a representative blot of three independent experiments.

7

Molecular Immunology xxx (xxxx) xxx–xxx

S. Luo et al.

Fig. 5. Pra1 blocks C3a mediated antifungal activity. (A) Pra1 blocks C3a antifungal activity. C3a (2.5 μM) was added to candida cells, and after incubation, candida yeast cells were stained with calcofluor (blue fluorescence) and propidium iodide (red fluorescence). Damaged yeast cells were identified as propidium iodide (red fluorescence) positive cells by confocal microscopy (upper panels). Pra1 used at 2.5 μM blocked C3a mediated antifungal activity and in this case, propidium iodide negative cells (red fluorescence) were identified (lower panels). (B) Pra1 inhibits C3a antifungal activity. Candida yeast cells were incubated with C3a (2.5 μM) and with Pra1 at the indicated concentrations (0.6–2.5 μM). Following incubation for 2 h at 30 °C, candida cells were stained with propidium iodide and analyzed by flow cytometry. C3a had antifungal activity; in the presence of C3a 91% of yeast cells were identified in quadrant II (panel II) as granular, propidium iodide positive cells. Pra1 blocked C3a antifungal activity in a dose dependent manner (C) Statistical analysis of three independent experiments shown in panel 5B. (D) C3a-desArg mediates antifungal activity and Pra1 blocks C3a-desArg mediated cell damage. The data in panel D represent mean values and ± SD of three separate experiments. (E) Endogenous Pra1 blocks C3a antifungal activity and Pra1 levels influence the degree of damage. C3a (0.5 μM) was added to wild type candida wild-type, a Pra1 knock out (Pra1ko)and a Pra1 overexpression strain (Pra1+++). Following incubation and staining with propidium iodide, the percentage of damaged candida yeast cells was evaluated by flow cytometry. Endogenous Pra1 blocked C3a mediated antifungal activity. For the wild type candida 63% of the cells were damaged by C3a (left panel). For the knock out strain 93% of the cells were damaged (middle panel), and for the overexpression strain (Pra1+++) only 39% of the cells appeared in QII (right panel). Representative data of three independent experiments are shown.

right quadrant). Pra1 at 2.5 μM decreased the fraction of damaged candida yeast cell by >90% and the effect was dose dependent (Fig. 5B, panels III–V and 5C). Furthermore, Pra1 also blocked the antifungal activity of C3a-desArg, again in a dose dependent manner (Fig. 5D). Surprisingly, also C4a has antifungal activity. Pra1, which does not bind to C4a (data not shown) did not block C4a mediated antifungal activity (Supplementary Fig. 4A). C5a had no antifungal activity (Supplementary Fig. 4B). Thus Pra1 efficiently blocks the antifungal activity of both C3a and C3adesArg, but not of C4a. Next, we asked if candida derived, surface exposed Pra1 blocks C3a antifungal activity. When wild-type candida cells were challenged with C3a, 63% of the cells were damaged. But 93% of the Pra1 ko strain were damage by C3a. In contrast, the Pra1 overexpression strain was more resistant to the antifungal C3a, and only 39% of the cells were damaged (Fig. 5E). Thus Pra1 expressed at the fungal surface blocks the antifungal activity of C3a.

human C3 convertases. We, therefore, analyzed if Pra1 blocked the effector function of both C3a and C3b. C3a has antifungal activity and is an anaphylatoxin (Klos et al., 2013; Sonesson et al., 2007). Therefore, C3a was added to candida yeast cells. After incubation, candida yeast cells with stained with calcofluor and propidium iodide, and by confocal microscopy, damaged cells were identified as calcofluor (blue fluorescence) and propidium iodide (with red fluorescence) positive cells (Fig. 5A, upper panel). When challenged C3a together with Pra1, the candida cells remained intact and were identified as calcofluor positive, propidium iodide negative cells (Fig. 5A, lower panel). Thus Pra1 blocked the direct antifungal effect of C3a. Similarly, the antifungal activity of C3a was evaluated by flow cytometry. Intact candida yeast cells were identified as granular, propidium iodide-negative cells in QI (Fig. 5B, panel I, upper left quadrant) and damaged cells as granular, propidium iodide positive cells in QII. Upon C3a challenge, 91% of the candida cells were damaged (Fig. 5B; panel II, upper 8

Molecular Immunology xxx (xxxx) xxx–xxx

S. Luo et al.

Fig. 6. Pra1 blocks C3a binding to the human C3a receptor and C3a receptor mediated signaling. (A) Pra1 inhibits binding of C3a to the C3a receptor. HEK293 cells which stably express the C3a receptor were incubated either with 125I-C3a alone, or 125I-C3a together with Pra1 used at different concentrations. Pra1 inhibited C3a binding to the human C3a receptor and the effect was dose-dependent. In addition Gpd2, an unrelated fungal immune evasion protein, was added to C3a. Gpd2 did not influence on C3a binding. (B) Pra1 blocked C3a receptor triggered calcium release. Fura-AM loaded C3a receptor transfected RBL cells as well as non-transfected wild type cells were stimulated with C3a (0.3 nM) in the presence of Pra1 (3.6 nM) or Gpd2 used as a negative control (5.2 nM). Pra1, but not Gpd2 blocked C3a induced calcium signaling. (C) Pra1 blocked C3a-activated chemotaxis. Migration of human PMNs through transwells towards a C3a gradient is shown and the effect of Pra1 used at (13, 26, 52, 104, 208, and 416 nM) was assessed after 60 min. Pra1 inhibited chemotaxis of human PMN in response to C3a. Gpd2, a second fungal immune evasion protein had no effect. Pra1, Gpd2 and bovine serum albumin (150 nM) alone lacked chemo attractive activity, but fMLP, used as a positive control (11 nM). (D) Pra1 blocked C3a induced IL-8 secretion. Isolated human neutrophils (2 × 106) were suspended in RPMI 1640 in a 24 well plate and stimulated either with C3a or C5a, or with the anaphylatoxins in combination with Pra1 (22.5, 45, 90, and180 nM) for 10 h. After incubation, IL-8 levels were measured in the culture supernatants. Pra1 blocked C3a mediated IL-8 secretion and the effect was dose dependent. Pra1 had no effect on C5a induced IL-8 secretion. Error bars represent mean values ± SEM of three independent experiments.

dependently (Fig. 7A). Pra1 used at 225 nM reduced C3b deposition by 75% (Fig. 7B). To determine whether endogenous candida derived Pra1 also restricts C3b deposition on the fungal cell surface, wild-type cells, the Pra1 ko, and Pra1+++ strains were challenged with NHS (10%) and after 15 min C3b deposition was evaluated. The Pra1 ko strain had two times more C3b deposited on the surface as compared to the wild type strain or the Pra1 overexpression strain (Fig. 7C). Next we evaluated whether Pra1 by blocking C3b mediated opsonization affects phagocytosis of candida yeast cells by human neutrophils. To this end, FITC labeled candida cells were challenged with NHS in the presence of Pra1. Then candida cells were added to human neutrophils, and after 20 min neutrophils having candida phagocytosed were identified as granular, FITC positive cells (Supplementary Fig. 5). Pra1 reduced phagocytosis of opsonized candida wild type cells in a dose dependent manner (Fig. 7D). Pra1 at 450 nM reduced phagocytosis by 54% (Fig. 7E). Similarly phagocytosis of the Pra1 knock out strain was more efficient and increased by 30%. However the Pra1 overexpressing strain was more resistant and phagocytosis was reduced by 60% (Fig. 7F). The Pra1 blocking effect on phagocytosis of candida cells by neutrophils was followed by live cell imaging. Candida cells challenged with complement active NHS (5%) were efficiently phagocytosed. Phagocytosed candida cells were detected already after 2 min and after 25 min most candida cells were phagocytosed by human neutrophils (Movie S1, Supplementary Information). Pra1 when added to complement active NHS blocked phagocytosis and even after 25 min most yeast cells were identified as individual cells and not associated with neutrophils (Movie S2, Supplementary Information). Thus Pra1 blocks opsonophagocytosis (Fig. 8). Pra1 protects candida from complement and killing by neutrophils. To confirm the protective effect, candida yeast cells were challenged with complement active NHS (5%), then human neutrophils were added. After 5 h surviving candida cells were determined by plating and only 12% of the yeast cells survived this treatment (Fig. 7G). However, Pra1 when added to NHS blocked neutrophil killing and candida survival increased by 60% (Fig. 7G).

3.8. Pra1 blocks C3a-C3a receptor binding and C3R mediated effector signaling We next asked if Pra1 influences C3a receptor (C3aR) binding and signaling. To this end, Pra1 was bound to I125-C3a, and the complexes were added to C3aR expressing HEK cells. Pra1 blocked I125-C3a binding to the C3aR (Fig. 6A). The effect was dose dependent. Pra1 >18 nM blocked C3a binding completely. In contrast, Gpd2, a related candida immune evasion protein also expressed in Pichia cells did not affect C3aR binding (Fig. 6A). C3a induces intracellular Ca2+ release. Pra1 when added to C3a inhibited C3a mediated calcium release of Fura2-AM loaded C3aR expressing rat basophilic cells (Fig. 6B). Neither Pra1, nor Gpd2 alone caused Ca2+-release in these cells (data not shown). Thus, Pra1 by complexing C3a inhibits C3aR binding and C3aR mediated calcium signaling. C3a is chemotactic and attracts human PMNs (Klos et al., 2013). Pra1 blocking effect on C3a mediated chemoattraction was assayed in a transmigration assay. Pra1 at 104 nM blocked C3a mediated cell migration of PMN (Fig. 6C). Pra1 specifically inhibited C3a chemotaxis, but did not influence chemotaxis by C5a (data not shown). The anaphylatoxins C3a and C5a induce IL-8 secretion by human PMNs (Dutow et al., 2014). Therefore we tested whether Pra1 affects C3a mediated IL8 release. PMNs were stimulated with the Pra1:C3a complex and after 10 h IL-8 was quantified in the supernatant (Fig. 6D). Pra1 suppressed C3a mediated, but not C5a induced IL-8 secretion. Again the effect was dose dependent (Fig. 6D). Thus Pra1 blocks C3a mediated effector functions. 3.9. Pra1 blocks C3b deposition and opsonophagocytosis of C. albicans by human neutrophils To analyze if Pra1 protects candida from opsonophagocytosis (Luo et al., 2010), candida yeast cells were challenged with complement active human serum (NHS) for 20 min, and then surface deposited C3b was analyzed by flow cytometry. Pra1 blocked C3b opsonization dose 9

Molecular Immunology xxx (xxxx) xxx–xxx

S. Luo et al.

Fig. 7. Pra1 blocks C3b deposition at the fungal surface and C3b mediated killing by human neutrophil. Pra1 blocks C3b deposition on the candida surface. (A) Pra1 (1800 nM) was added to NHS (200 μl) at the indicated concentrations for 1 h and then the mixtures were added to candida yeast cells and after incubation C3b deposition was followed by flow cytometry. Heat inactivated human serum served as a control. (B) Pra1 (from 225 to 1800 nM) was added to NHS (5%, 200 μl) after 1 h the mixture was added to candida yeast cells and following 20 min incubation at 37 °C, C3b deposited on the candida surface was analyzed by flow cytometry. (C) Candida wildtype cells (WT), the Pra1 knock out (ko) and the overexpression strain (Pra1+++) were challenged with NHS (10%) for 15 min. Then C3b deposited on the fungal surface was measured by flow cytometry. (D) Pra1 blocked phagocytosis of C. albicans by human PMN. Pra1 (1800 nM) was added to NHS (200 μl), after 1 h the mixture was added to human neutrophils together with FITC-labeled candida cells, at a 1:3 ratio. After incubation phagocytosis was analyzed by flow cytometry. Heat inactivated serum served as a control. (E) Pra1 was added to NHS (5%, 200 μl), after 1 h, the mixture was added to human neutrophils and to FITC-labeled candida cells (ratio 1:3). Following incubation the rate of phagocytosis was measured by flow cytometry. (F) Human neutrophils and FITC-labeled candida cells, i.e. wild-type (WT), the Pra1 knock out (ko) and the overexpression strain (Pra1+++) were incubated together (1:3) for 15 min in NHS (5%). Then phagocytosis was determined. (G) Pra1 blocked killing of C. albicans by human neutrophils. Pra1 (1800 nM) was added to NHS (5%, 200 μl), then the mixture was added to human neutrophils and candida cells (3:1) and the cells were incubated for 5 h. Afterwards candida cells were plated on YPD-agar plates and surviving yeast cells were counted. All figures show mean ± SEM of three independent experiments.

4. Discussion Control and escape from host complement attack are central for C. albicans to survive in the human host. Here we identify candida Pra1 as a fungal master complement inhibitor that controls in a hierarchical manner human complement attack. Pra1 cleaves human C3 and generates C3aL and C3bL fragments. Pra1 cleaves and degrades C3aL, and C3bL is processed by host regulators. Furthermore, Pra1 binds to activation fragments C3a and C3b, which are generated by the human C3 convertase. Pra1 blocks C3aL meditated antifungal activity, C3a receptor binding and C3aR signaling. Furthermore, Pra1 blocks C3bL opsonophagocytosis by human neutrophils. Thus Pra1 inactivates C3 directly and furthermore blocks the effector functions of endogenously generated C3 activation fragments. By cleaving C3 and by blocking the action of C3 effector fragments, Pra1 directly destroys central immune weapons of the human host and forms a micro-environment which is beneficial for fungal survival and for infection. Pra1 is a fungal C3 cleaving protease. Pra1 mediated cleavage of C3 was inhibited by PMSF and by a proteinase inhibitor cocktail, but neither by phenanthrolin, EDTA, AEBSF, nor E64. Pra1 cleaves C3 between Ser742 and His743 and this site is positioned six residues upstream of the cleavage site used by the endogenous human C3 convertase. Pra1 binds to two linear regions, i.e. S706CQRRTRF713 and T732ELRRQH738, which are contained in the α3 and α4 helical regions of C3aL, C3adesArg and C3a (Bajic et al., 2013). Thereby Pra1 blocks the antifungal action of both C3aL and C3a. Pra1 by binding to C3a blocks C3a

Fig. 8. The secreted C. albicans protein Pra1 disrupts complement by targeting C3 in multiple manners. 1). Secreted Pra1 directly cleaves C3 and generates C3aL and C3bL fragments left panel. 2) Pra1 further degrades the C3aL activation fragment, 3) C3bL is degraded by human protease Factor I assisted by the cofactor Factor H. In addition, 4) Pra1 binds to the activation fragments C3a and 5) C3b which are generated by the human convertase and blocks antifungal activity of C3a and inhibits C3b opsonization.

10

Molecular Immunology xxx (xxxx) xxx–xxx

S. Luo et al.

Funding

receptor interaction. Pra1 further degrades C3aL, but not C3a. C3aL lacks the six C terminal residues, including Arg748 which is central for C3aR binding. Thus C3aL, similar to C3a-desArg, does not bind to the C3aR and lacks pro-inflammatory activity. Protease activity of Pra1 is specific for C3 since Pra1 does not cleave the C3 activation fragments, C3b, iC3b, and C3c. Further, Pra1 shows no substrate specificity for other structurally related human complement proteins C4 and C5. Cleavage of C3 by candida Pra1 generates active effector fragments and apparently causes an infection paradox. However, the fungus itself generates and furthermore inactivates antifungal C3aL. This is a specific effect as antimicrobial C3a which is generated by the host is resistant to Pra1 inactivation. In addition C3bL is efficiently processed and inactivated by the human regulators Factors H and Factor I. Inactivation of Pra1 generated C3bL by human regulators has also been described for the microbial pathogens, Staphylococcus aureus, Neisseria meningitides, Escherichia coli and for Enterococcus faecalis (Abreu et al., 2015). In addition NalP, from Neisseria meningitidis cleaves the α-chain of C3. The cleavage site is positioned four residues upstream from the C3 convertase site (Del Tordello et al., 2014), aureolysin, the metalloprotease from Staphylococcus aureus (Laarman et al., 2011) and similarly gelatinase (GelE) from Enterococcus faecalis cleave the α chain of C3 two residues downstream of the host convertase (Park et al., 2008). Thus, several microbial pathogens cleave and inactivate C3 at sites distinct from that of the endogenous C3 convertase. Pra1 protects candida by blocking all three complement pathways. Pra1 recruits human C4BP to inactivate the classical and the lectin pathways. Pra1 also recruits Factor H, FHL-1 to block the alternative pathway. Pra1 blocks formation of C3 convertase by complexing, further cleaving human C3, thereby directly inactivating this central complement component C3. Furthermore Pra1 also binds to the cleavage fragments C3a and C3b generated by the human C3 convertases and directly inhibits C3a, as well as C3b effector functions. Pra1 inhibits C3a antifungal activity, blocks C3a binding to C3aR, subsequent receptor signaling, as well as C3aR induced cellular responses. Pra1, upon binding to C3b blocks fungal opsonohagocytosis by human immune cells. Moreover, Pra1 binds to mouse CD4+ T cells and inhibits cytokine expression (Bergfeld et al., 2017). Given the role of C3a in adaptive immunity, for induction of human Th1 response and Th17 differentiation, it will be interesting to analyze whether Pra1 also modulates thisC3a activity in adaptive immunity (Asgari et al., 2013; Ghannam et al., 2014). C3b deposited on the surface of the fungal pathogen is recognized by the human complement receptors CR3 (αMβ2) and CR4 (αXβ2), which are expressed by human neutrophils and monocytes/macrophages (Zipfel and Skerka, 2009). Pra1 also binds to CR3 and CR4 and blocks CR3/CR4 mediated uptake of candida and addition of soluble Pra1 to the co-culture of candida cells and neutrophils enhances fungal survival (Jawhara et al., 2012; Soloviev et al., 2007). Thereby Pra1 modulates human immunity and generates an immune benefit for the pathogen. C. albicans uses Pra1 to block human immune attack at multiple levels. Here we identify Pra1 as a hierarchical C3 inhibitor. Timing of expression and also the site of inhibitor action are relevant for fungal immune escape. Pra1 is expressed on the surface of candida yeast cells and is also secreted into the surrounding medium. Pra1 expression is induced in hyphae. Secreted Pra1 reduces C3 in vicinity of the fungus and in consequence less C3 is available for attack. Pra1 by restricting the action of C3 and also of that of the C3 effector fragments dampens the hostile immune response, thereby generating a beneficial immune environment. In conclusion, candida Pra1 is a fungal master regulator of innate immunity. Pra1 inhibits complement action at multiple sites, inactivates native C3 and blocks the effector function of the host generated effector components C3a and C3b.

The work was funded by the Collaborative Research Center / Transregio 124 FungiNet of the Deutsche Forschungsgemeinschaft (DFG) projects C6 (PFZ & NB), C4 (CS) and C5 (IJ). NR acknowledges a fellowship from the Center for Sepsis Control and Care of the University Hospital of the Friedrich Schiller University, Jena Germany. Acknowledgements We thank Dr. Beate Fehlhaber and Claudia Rheinheimer for their assistance with the 125I-C3a binding and Ca2+ Fura-2AM studies. References Abreu, A.G., Fraga, T.R., Granados Martinez, A.P., Kondo, M.Y., Juliano, M.A., Juliano, L., Navarro-Garcia, F., Isaac, L., Barbosa, A.S., Elias, W.P., 2015. The Serine Protease Pic from Enteroaggregative Escherichia coli mediates immune evasion by the direct cleavage of complement proteins. J. Infect. Dis. 212, 106–115. Alonso-Valle, H., Acha, O., Garcia-Palomo, J.D., Farinas-Alvarez, C., FernandezMazarrasa, C., Farinas, M.C., 2003. Candidemia in a tertiary care hospital: epidemiology and factors influencing mortality. Eur. J. Clin. Microbiol. Infect. Dis. 22, 254–257. Alves, I.A., Camargo, F.P., Goulart, L.S., 2017. [Identification by PCR and antifungal susceptibility of vaginal clinical Candida sp. isolates]. Rev. Soc. Bras. Med. Trop. 43, 575–579. Asgari, E., Le Friec, G., Yamamoto, H., Perucha, E., Sacks, S.S., Kohl, J., Cook, H.T., Kemper, C., 2013. C3a modulates IL-1beta secretion in human monocytes by regulating ATP efflux and subsequent NLRP3 inflammasome activation. Blood 122, 3473–3481. Bajic, G., Yatime, L., Klos, A., Andersen, G.R., 2013. Human C3a and C3a desArg anaphylatoxins have conserved structures, in contrast to C5a and C5a desArg. Protein Sci. 22, 204–212. Behnsen, J., Hartmann, A., Schmaler, J., Gehrke, A., Brakhage, A.A., Zipfel, P.F., 2008. The opportunistic human pathogenic fungus Aspergillus fumigatus evades the host complement system. Infect. Immun. 76, 820–827. Bergfeld, A., Dasari, P., Werner, S., Hughes, T.R., Song, W.C., Hortschansky, P., Brakhage, A.A., Hunig, T., Zipfel, P.F., Beyersdorf, N., 2017. Direct binding of the pH-regulated protein 1 (Pra1) from Candida albicans inhibits cytokine secretion by mouse CD4+ T cells. Front. Microbiol. 8, 844. Casanova, M., Lopez-Ribot, J.L., Monteagudo, C., Llombart-Bosch, A., Sentandreu, R., Martinez, J.P., 1992. Identification of a 58-kilodalton cell surface fibrinogen-binding mannoprotein from Candida albicans. Infect. Immun. 60, 4221–4229. Cheng, S.C., Sprong, T., Joosten, L.A., van der Meer, J.W., Kullberg, B.J., Hube, B., Schejbel, L., Garred, P., van Deuren, M., Netea, M.G., 2012. Complement plays a central role in Candida albicans-induced cytokine production by human PBMCs. Eur. J. Immunol. 42, 993–1004. Citiulo, F., Jacobsen, I.D., Miramon, P., Schild, L., Brunke, S., Zipfel, P., Brock, M., Hube, B., Wilson, D., 2012. Candida albicans scavenges host zinc via Pra1 during endothelial invasion. PLoS Pathog. 8, e1002777. Crowe, J.D., Sievwright, I.K., Auld, G.C., Moore, N.R., Gow, N.A., Booth, N.A., 2003. Candida albicans binds human plasminogen: identification of eight plasminogenbinding proteins. Mol. Microbiol. 47, 1637–1651. Dasari, P., Reiss, K., Lingelbach, K., Baumeister, S., Lucius, R., Udomsangpetch, R., Bhakdi, S.C., Bhakdi, S., 2011. Digestive vacuoles of Plasmodium falciparum are selectively phagocytosed by and impair killing function of polymorphonuclear leukocytes. Blood 118, 4946–4956. Del Tordello, E., Vacca, I., Ram, S., Rappuoli, R., Serruto, D., 2014. Neisseria meningitidis NalP cleaves human complement C3, facilitating degradation of C3b and survival in human serum. Proc. Natl. Acad. Sci. U. S. A. 111, 427–432. Dutow, P., Fehlhaber, B., Bode, J., Laudeley, R., Rheinheimer, C., Glage, S., Wetsel, R.A., Pabst, O., Klos, A., 2014. The complement C3a receptor is critical in defense against Chlamydia psittaci in mouse lung infection and required for antibody and optimal T cell response. J. Infect. Dis. 209, 1269–1278. Ghannam, A., Fauquert, J.L., Thomas, C., Kemper, C., Drouet, C., 2014. Human complement C3 deficiency: Th1 induction requires T cell-derived complement C3a and CD46 activation. Mol. Immunol. 58, 98–107. Gillum, A.M., Tsay, E.Y., Kirsch, D.R., 1984. Isolation of the Candida albicans gene for orotidine-5'-phosphate decarboxylase by complementation of S. cerevisiae ura3 and E. coli pyrF mutations. Mol. Gen. Genet. 198, 179–182. Gropp, K., Schild, L., Schindler, S., Hube, B., Zipfel, P.F., Skerka, C., 2009. The yeast Candida albicans evades human complement attack by secretion of aspartic proteases. Mol. Immunol. 47, 465–475. Gudlaugsson, O., Gillespie, S., Lee, K., Vande Berg, J., Hu, J., Messer, S., Herwaldt, L., Pfaller, M., Diekema, D., 2003. Attributable mortality of nosocomial candidemia, revisited. Clin. Infect. Dis. 37, 1172–1177. Heeger, P.S., Kemper, C., 2012. Novel roles of complement in T effector cell regulation. Immunobiology 217, 216–224. Iannitti, R.G., Carvalho, A., Romani, L., 2012. From memory to antifungal vaccine design. Trends. Immunol. 33, 467–474. Jawhara, S., Pluskota, E., Verbovetskiy, D., Skomorovska-Prokvolit, O., Plow, E.F.,

Conflict of interest statement The authors declare no conflict of interest in this study. 11

Molecular Immunology xxx (xxxx) xxx–xxx

S. Luo et al. Soloviev, D.A., 2012. Integrin αXβ₂ is a leukocyte receptor for Candida albicans and is essential for protection against fungal infections. J. Immunol. 189, 2468–2477. Johswich, K., Martin, M., Thalmann, J., Rheinheimer, C., Monk, P.N., Klos, A., 2006. Ligand specificity of the anaphylatoxin C5L2 receptor and its regulation on myeloid and epithelial cell lines. J. Biol. Chem. 281, 39088–39095. Jong, A.Y., Stins, M.F., Huang, S.H., Chen, S.H., Kim, K.S., 2001. Traversal of Candida albicans across human blood-brain barrier in vitro. Infect. Immun. 69, 4536–4544. Klos, A., Bank, S., Gietz, C., Bautsch, W., Kohl, J., Burg, M., Kretzschmar, T., 1992. C3a receptor on dibutyryl-cAMP-differentiated U937 cells and human neutrophils: the human C3a receptor characterized by functional responses and 125I-C3a binding. Biochemistry 31, 11274–11282. Klos, A., Wende, E., Wareham, K.J., Monk, P.N., 2013. International union of basic and clinical pharmacology. [corrected]. LXXXVII. complement peptide C5a, C4a, and C3a receptors. Pharmacol. Rev. 65, 500–543. Laarman, A.J., Ruyken, M., Malone, C.L., van Strijp, J.A., Horswill, A.R., Rooijakkers, S.H., 2011. Staphylococcus aureus metalloprotease aureolysin cleaves complement C3 to mediate immune evasion. J. Immunol. 186, 6445–6453. Lesiak-Markowicz, I., Vogl, G., Schwarzmuller, T., Speth, C., Lass-Florl, C., Dierich, M.P., Kuchler, K., Wurzner, R., 2011. Candida albicans Hgt1p, a multifunctional evasion molecule: complement inhibitor, CR3 analogue, and human immunodeficiency virusbinding molecule. J. Infect. Dis. 204, 802–809. Limper, A.H., Standing, J.E., 1994. Vitronectin interacts with Candida albicans and augments organism attachment to the NR8383 macrophage cell line. Immunol. Lett. 42, 139–144. Luo, S., Poltermann, S., Kunert, A., Rupp, S., Zipfel, P.F., 2009. Immune evasion of the human pathogenic yeast Candida albicans: Pra1 is a Factor H, FHL-1 and plasminogen binding surface protein. Mol. Immunol. 47, 541–550. Luo, S., Hartmann, A., Dahse, H.M., Skerka, C., Zipfel, P.F., 2010. Secreted pH-regulated antigen 1 of Candida albicans blocks activation and conversion of complement C3. J. Immunol. 185, 2164–2173. Luo, S., Blom, A.M., Rupp, S., Hipler, U.C., Hube, B., Skerka, C., Zipfel, P.F., 2011. The pH-regulated antigen 1 of Candida albicans binds the human complement inhibitor C4b-binding protein and mediates fungal complement evasion. J. Biol. Chem. 286, 8021–8029. Luo, S., Hoffmann, R., Skerka, C., Zipfel, P.F., 2013a. Glycerol-3-phosphate dehydrogenase 2 is a novel factor H-, factor H-like protein 1-, and plasminogen-binding surface protein of Candida albicans. J. Infect. Dis. 207, 594–603. Luo, S., Skerka, C., Kurzai, O., Zipfel, P.F., 2013b. Complement and innate immune evasion strategies of the human pathogenic fungus Candida albicans. Mol. Immunol. 56, 161–169. Meri, T., Hartmann, A., Lenk, D., Eck, R., Wurzner, R., Hellwage, J., Meri, S., Zipfel, P.F., 2002. The yeast Candida albicans binds complement regulators factor H and FHL-1. Infect. Immun. 70, 5185–5192. Meri, T., Blom, A.M., Hartmann, A., Lenk, D., Meri, S., Zipfel, P.F., 2004. The hyphal and yeast forms of Candida albicans bind the complement regulator C4b-binding protein. Infect. Immun. 72, 6633–6641. Mullick, A., Elias, M., Picard, S., Bourget, L., Jovcevski, O., Gauthier, S., Tuite, A., Harakidas, P., Bihun, C., Massie, B., Gros, P., 2004. Dysregulated inflammatory response to Candida albicans in a C5-deficient mouse strain. Infect. Immun. 72, 5868–5876.

Olczak-Kowalczyk, D., Pawlowska, J., Garczewska, B., Smirska, E., Grenda, R., Syczewska, M., Kowalczyk, W., 2016. Oral candidiasis in immunosuppressed children and young adults after liver or kidney transplantation. Pediatr. Dent. 32, 189–194. Pappas, P.G., Rex, J.H., Lee, J., Hamill, R.J., Larsen, R.A., Powderly, W., Kauffman, C.A., Hyslop, N., Mangino, J.E., Chapman, S., Horowitz, H.W., Edwards, J.E., Dismukes, W.E., Group, N.M.S., 2003. A prospective observational study of candidemia: epidemiology, therapy, and influences on mortality in hospitalized adult and pediatric patients. Clin. Infect. Dis. 37, 634–643. Park, S.Y., Shin, Y.P., Kim, C.H., Park, H.J., Seong, Y.S., Kim, B.S., Seo, S.J., Lee, I.H., 2008. Immune evasion of Enterococcus faecalis by an extracellular gelatinase that cleaves C3 and iC3b. J. Immunol. 181, 6328–6336. Poltermann, S., Kunert, A., von der Heide, M., Eck, R., Hartmann, A., Zipfel, P.F., 2007. Gpm1p is a factor H-, FHL-1-, and plasminogen-binding surface protein of Candida albicans. J. Biol. Chem. 282, 37537–37544. Rawlings, N.D., Barrett, A.J., 1995. Evolutionary families of metallopeptidases. Methods Enzymol. 248, 183–228. Settmacher, B., Rheinheimer, C., Hamacher, H., Ames, R.S., Wise, A., Jenkinson, L., Bock, D., Schaefer, M., Kohl, J., Klos, A., 2003. Structure-function studies of the C3a-receptor: C-terminal serine and threonine residues which influence receptor internalization and signaling. Eur. J. Immunol. 33, 920–927. Soloviev, D.A., Fonzi, W.A., Sentandreu, R., Pluskota, E., Forsyth, C.B., Yadav, S., Plow, E.F., 2007. Identification of pH-regulated antigen 1 released from Candida albicans as the major ligand for leukocyte integrin alphaMbeta2. J. Immunol. 178, 2038–2046. Sonesson, A., Ringstad, L., Nordahl, E.A., Malmsten, M., Morgelin, M., Schmidtchen, A., 2007. Antifungal activity of C3a and C3a-derived peptides against Candida. Biochim. Biophys. Acta 1768, 346–353. Spreghini, E., Gismondi, A., Piccoli, M., Santoni, G., 1999. Evidence for alphavbeta3 and alphavbeta5 integrin-like vitronectin (VN) receptors in Candida albicans and their involvement in yeast cell adhesion to VN. J. Infect. Dis. 180, 156–166. Tsoni, S.V., Kerrigan, A.M., Marakalala, M.J., Srinivasan, N., Duffield, M., Taylor, P.R., Botto, M., Steele, C., Brown, G.D., 2009. Complement C3 plays an essential role in the control of opportunistic fungal infections. Infect. Immun. 77, 3679–3685. Walport, M.J., 2001. Complement. First of two parts. N. Engl. J. Med. 344, 1058–1066. Zhang, M.X., Kozel, T.R., 1998. Mannan-specific immunoglobulin G antibodies in normal human serum accelerate binding of C3 to Candida albicans via the alternative complement pathway. Infect. Immun. 66, 4845–4850. Zipfel, P.F., Skerka, C., 2009. Complement regulators and inhibitory proteins. Nat. Rev. Immunol. 9, 729–740. Zipfel, P.F., Jokiranta, T.S., Hellwage, J., Koistinen, V., Meri, S., 1999. The factor H protein family. Immunopharmacology 42, 53–60. Zipfel, P.F., Hallstrom, T., Hammerschmidt, S., Skerka, C., 2008. The complement fitness factor H: role in human diseases and for immune escape of pathogens, like pneumococci. Vaccine 26 (Suppl. 8), I67–74. Zipfel, P.F., Skerka, C., Kupka, D., Luo, S., 2011. Immune escape of the human facultative pathogenic yeast Candida albicans: the many faces of the Candida Pra1 protein. Int. J. Med. Microbiol. 301, 423–430. Zipfel, P.F., 2009. Complement and immune defense: from innate immunity to human diseases. Immunol. Lett. 126, 1–7. Zipfel, P.F., Wurzner, R., Skerka, C., 2007. Complement evasion of pathogens: common strategies are shared by diverse organisms. Mol. Immunol. 44, 3850–3857.

12