p33 (gC1q Receptor) Prevents Cell Damage by Blocking the Cytolytic Activity of Antimicrobial Peptides

Johannes Westman, Finja C. Hansen, Anders I. Olin, Matthias Mörgelin, Artur Schmidtchen and Heiko Herwald
J Immunol December 1, 2013, 191 (11) 5714-5721; DOI: https://doi.org/10.4049/jimmunol.1300596

Abstract

The innate immune system is the first line of defense against invading microbes. Its specificity relies a great deal on host pattern recognition molecules that sense pathogen-associated molecular patterns of the invading pathogen. However, full protection is not always guaranteed, and some early defense mechanisms involved in bacterial killing, such as the complement system, can also exert cytolytic activity against host cells. Although these cascades are tightly regulated, the host has to take additional precautions to prevent its cell destruction. In this study, we describe that p33, a negatively charged surface protein found on endothelial cells also known as gC1q receptor, protects host cells from a cytolytic attack by antimicrobial peptides (AMPs), such as LL37 and β-defensin 3. To this end, we characterized the interaction of p33 with AMPs by biochemical and functional means. Our data show that p33 forms a doughnut-shaped trimer that can bind up to three AMPs, and we identified a segment in p33 forming a β-sheet that mediates the binding to all AMPs. Moreover, our results show that p33 abolishes the lytic activity of AMPs at an equimolar ratio, and it protects endothelial cells and erythrocytes from AMP-induced lysis. Taken together, our data suggest a novel protective mechanism of p33 in modulating innate immune response by neutralizing cytotoxic AMPs at the host cell surface.

Introduction

The early recognition and elimination of an invading pathogen are two key features of the innate immune system. To fulfill these tasks, the immune response has to distinguish between self and nonself because this will allow an attack of the foreign invader without causing harm to the host. Pattern recognition molecules play an important role in these processes because they target so-called pathogen-associated molecular patterns, which are exposed on the surface of the intruder. Considering the amplitude of toxicity of some innate defense mechanisms, their selectivity for pathogenic microorganisms is not always guaranteed; thus, there is a risk that some innate immune reactions may also have a deleterious impact on the host. The complement system belongs to this group of early host defense responses that can cause significant tissue injury. To protect themselves from complement-mediated cellular damage, eukaryotic cells express membrane complement regulatory proteins such as membrane cofactor protein, decay-accelerating factor, and membrane inhibitor of reactive lysis, also known as CD55, CD46, and CD59, respectively (for reviews, see Refs. 1, 2). These three proteins are able to interfere with different parts of the complement system and this protects the cells from a complement attack (3).

Antimicrobial peptides (AMPs) are another group of early defense mediators that can target and damage host cells. AMPs are usually short (<100 aa) and cationic molecules. They comprise helical or linear structures, or contain β-sheets stabilized by disulphide bridges. Most AMPs exert broad antibacterial activity against both Gram-positive and -negative bacteria (4) by forming pores in the bacterial cell wall that has a higher negative transmembrane potential than the plasma membrane of eukaryotic cells (5). Moreover, the lack of cholesterol makes the bacterial cell lipid layer more susceptible to an attack by AMPs (6). The specificity of AMPs to target bacterial membranes is concentration dependent, and already low AMP levels are sufficient to cause bacterial cell-wall destruction. However, at high concentrations, AMPs also explore cytolytic activity against eukaryotic cells (7–10). Notably, AMP levels can significantly elevate under inflammatory conditions such as in patients with severe sepsis, and this has been shown to correlate with circulatory derangement (11). It is therefore tempting to speculate that a massive release of AMPs into the bloodstream or at cell–cell interaction sites could contribute to tissue destruction under systemic induction of innate immune reactions, although not much research has been performed to address this issue.

p33 (also referred to as p32 or globular C1q receptor) is a multiligand binding protein that is expressed on the surface of various cell types such as endothelial, neutrophils, lymphocytes, and platelets (12, 13), but it can also be found intracellularly, for instance, in mitochondria or nucleus (14). The protein is synthesized as a precursor protein consisting of 282 aa. At the cell surface it is located in its mature form (aa 74–282) with a highly negative net charge (isoelectric point [pI] of 4.15). p33 was originally identified as a complement protein C1q binding protein that prevents complement activation (15). The binding of p33 to C1q is only of moderate nature (240 ± 10 nM) (12), and other p33-binding ligands with a higher affinity have been described. For instance, p33 interacts with plasma proteins such as high molecular weight kininogen (HK) and coagulation factor XII (12). An affinity in the lower nanomolar range (9 nM) was determined for this interaction, and the p33 binding site in HK was mapped to a highly cationic sequence located in domain 5 of HK (12). A synthetic 20mer peptide (HKH20) was found to resemble the p33 binding site in HK, and subsequent studies revealed that HKH20 uses antimicrobial activity against Streptococcus pyogenes and Escherichia coli (16). In a sepsis model, mice injected with S. pyogenes were treated with HKH20, and this resulted in a significant increase in survival (17).

The interaction between p33 and bacteria such as Staphylococcus aureus, Bacillus cereus, or Listeria monocytogenes, has been described previously (18–20). It has been speculated that these bacteria use p33 to adhere and eventually invade host cells. However, to our knowledge, it has not been reported that p33 is able to dampen a self-destructive host response to infection. In this article, we report a novel mechanism of p33 as modulator of the early innate immune response. We find that the binding of p33 to AMPs, including HKH20, β-defensin 3, and LL37, not only neutralizes their antimicrobial activity, but is also is a mechanism that prevents AMP-induced host cell damage.

Materials and Methods

Blood collection

Peripheral venous blood was collected from healthy human volunteers into 2.7 ml, 0.109 M buffered sodium citrate tubes (Becton Dickinson).

Bacterial strains

The E. coli ATCC 25922 was from the Department of Bacteriology, Lund University Hospital.

Purification of recombinant maltose-binding protein–p33

Recombinant maltose-binding protein (MBP)–p33 was expressed using the pMAL-c2 expression vector (New England Biolabs) and E. coli XL1-Blue strain (Stratagene, Heidelberg, Germany) as previously described (12). Isopropyl-β-d-thiogalactopyranoside was added to an exponential growing overnight culture of bacteria to a final concentration of 0.3 mM. Bacteria were harvested after 1 h (4000 × g, 10 min, 4°C), and the pellet was resuspended in 30 ml cold 20 mM Tris, 200 mM NaCl, 1 mM EDTA, 1 mM DTT, pH 7.4 (MBP-buffer). Bacteria were kept on ice and lysed by sonication (3 min in pulses of 15 s). The lysed cell wall was pelleted (15,000 × g, 20 min, 4°C), and the supernatant was removed and transferred (1 ml/min) to an amylose resin column pre-equilibrated with two column volumes of MBP buffer. After adding the bacterial supernatant, we rinsed the column with 12 column volumes MBP buffer before 4 column volumes 10 mM Tris-HCl, pH 7.4 (Tris-buffer). Bound MBP-p33 was eluted and fractionated with 10 mM Tris-HCl containing 10 mM maltose. The concentrations of MBP-p33 used in this study are calculated based on its monomeric form (75,500 Da).

Peptides and Abs

HKH20 and NAT26 were synthesized by Bachem, Feinchemikalien AG (Bubendorf, Switzerland). LL37 and the defensins were from Innovagen AB (Lund, Sweden), and GGL27 and GGL27(S) were from (Biopeptide, San Diego, CA). A list with all AMPs is shown in Table I. Overlapping peptides spanning the entire sequence of p33 (Table II) were generated by a peptide synthesis platform (Sigma-Aldrich PEPscreen Directory).

Polyclonal antisera to p33 were raised in rabbits (Innovagen AB). Ascites mouse anti-p33 IgG1 (60.11) and mouse anti-p33 IgG1 (74.5.2) were purchased from Nordic BioSite AB (Täby, Sweden). Peroxidase-conjugated goat anti-mouse IgG and peroxidase-conjugated goat anti-rabbit IgG were purchased from Bio-Rad Laboratories.

ELISA

Indirect ELISAs were performed by coating microtiter plates (Maxisorb, NUNC) overnight at 4°C with various AMPs (5 μM) in 15.9 mM Na2CO3, 30 mM NaHCO3, pH 9.6 (coating buffer). Plates were washed three times in deionized water, blocked in PBS containing 0.05% Tween 20 and 0.5% BSA (PBST) for 30 min at 37°C, and incubated with 2.5 μg/ml MBP-p33 (1 h, 37°C). Bound MBP-p33 was detected with an mAb against p33 (clone 60.11, MMS-606R; Nordic BioSite AB), diluted 1:3000 for 1 h at 37°C, and visualized by a peroxidase-conjugated secondary Ab against mouse IgG (1:2.500, 1 h, 37°C; Bio-Rad Laboratories). All incubations were followed by three washing steps in PBST. Competitive ELISAs were performed following the same protocol, except that 1 of the 20 aa overlapping p33 peptides (10 μM) was mixed with 2.5 μg/ml p33-MBP before incubation on the microtiter plates.

Cleavage and purification of native p33

MBP-p33 was treated with bovine factor Xa (50:1; New England Biolabs) overnight at room temperature (RT). Cleavage was visualized by SDS-PAGE, and a HiLoad 16/600 Superdex 75-pg gel filtration column (GE Healthcare) was used to separate p33 from MBP and factor Xa as described by the manufacturers. Notably, MBP-p33 and cleaved p33 form homotrimers, and have the same function and yield the same results in binding and functional assays (Supplemental Fig. 1).

Negative staining and transmission electron microscopy

The binding between p33 and AMPs was visualized by negative staining and electron microscopy as previously described (21). AMPs were conjugated with 5 nm colloidal gold particles according to routine protocols (22). Conjugates were incubated with p33 for 30 min at RT and negatively stained with 0.75% uranyl formate. Specimens were examined in a Philips/FEI CM100 BioTwin transmission electron microscope at a ×100,000 magnification.

Viable count assay

The E. coli ATCC 25922 strain was grown overnight in 3% (w/v) Tryptic Soy Broth medium at 37°C on rotation. A volume of 350 μl was transferred to 10 ml prewarmed Tryptic Soy Broth medium and grown to midlog phase (A620 = 0.4). The bacteria were washed (2300 × g, 10 min, 4°C), resuspended in 10 ml cold 10 mM Tris containing 5 mM glucose, pH 7.5, and further diluted in cold 10 mM Tris containing 5 mM glucose to 2 × 106 CFUs/ml. Fifty microliters bacteria were incubated with 50 μl various AMPs (1.5–6 μM), with or without 10 μM MBP-p33 in 10 mM Tris-HCl containing 10 mM maltose (1 h, 37°C, rotation). Antibacterial activity was determined by counting the number of CFUs on serial dilutions of the incubation mixtures plated on Todd Hewitt agar plates.

SDS-PAGE and Western blot analysis

SDS-PAGE was performed as described by Laemmli (23) using a polyacrylamide concentration of 10%. The separated proteins were transferred to a PVDF membrane (Amersham Biosciences). The membrane was rinsed in PBST before 30-min incubation with blocking buffer (PBST containing 5% milk powder), incubated with primary Ab (clone 60.11, MMS-606R; Nordic BioSite AB) diluted 1:1000 in blocking buffer for 60 min at RT. The membrane was washed three times in PBST, followed by an incubation with a peroxidase-labeled secondary Ab (goat anti-mouse IgG, 1:2,500; Bio-Rad Laboratories) in blocking buffer for 60 min at RT. Membranes were washed three times in PBST and developed using a SuperSignal West Pico Chemiluminescence kit (Thermo Scientific) according to manufacturer’s instructions.

Fluorescence microscopy

Endothelial cells (EA.hy926) were grown confluent on coverslips in 12-well plates. Cells were fixed in 4% paraformaldehyde for 15 min on ice and 45 min at RT. After fixation, cells were washed three times in 10 mM PBS before a blocking step with PBST including 2% BSA for 30 min at 37°C. Primary Ab mouse anti-p33 IgG1 (clone 60.11; Abcam) or an isotype control mouse IgG1 (Invitrogen) was added overnight at 4°C. The following day, cells were incubated with secondary goat anti-mouse IgG1 Alexa Fluor 488 (Invitrogen) for 60 min at 37°C. Coverslips were transferred to glass slides with one drop DAPI Prolong Gold Antifade reagent (Invitrogen) and let dry in the dark at RT overnight. One washing step was performed after every incubation step.

Immunohistochemistry

Slides were dried at 60°C for 90 min and deparaffinized with Tissue Clear (Histolab). Slides were transferred to a cuvette containing DIVA Decloaker (Biocare Medical) before Ag retrieval. Thereafter, slides were cooled and washed in TBST. Slides were blocked with Background Punisher Blocking reagent (Histolab) and washed again in TBST followed by addition of total IgG from p33 rabbit anti-serum or a rabbit IgG isotype control Ab for 2 h at RT. Slides were washed with TBST and Rabbit-on-Rodent AP polymer (Histolab) was added, followed by another washing step. Staining was visualized with Vulcan Fast Red Chromogen Kit 2 (BMA Biomedicals). Harris hematoxylin (Sigma) was used for counterstaining followed by a wash step in tap water. Slides were dipped in 0.05% HCl diluted in 70% ethanol, washed in tap water, and placed in MilliQ water containing ammonium chloride (4 drops in 500 ml). Slides were washed in tap water, dried, and mounted on coverslips using Pertex (Histolab).

Hemolysis assay

One milliliter citrate-blood was centrifuged (800 × g, 10 min, RT), and plasma was removed and replaced by an equal volume of PBS. The washing step was repeated two times, and the washed blood cells were kept in RT. AMPs were diluted to 40 μM in PBS in the absence or presence of 40 μM MBP-p33. Samples were preincubated (37°C, 60 min, on rotation) followed by an addition of 3 μl washed blood cells (5% v/v) and another incubation (60 min at 37°C on rotation). Tox 7-Lysis buffer (1:10) and PBS served as positive and negative control, respectively. Samples were then centrifuged (800 × g, 10 min, RT), and the supernatant was transferred to a 96-well plate. The absorbance of hemoglobin was measured at 540 nm and was expressed as the percentage of Tox-7 lysis-induced hemolysis.

Lactate dehydrogenase assay

EA.hy926 cells were grown confluent in 96-well plates in DMEM containing 10% FBS, 100 mM hypoxanthine, 0.4 mM aminopterin, 16 mM thymidine, 100 mg/ml streptomycin, and 100 U/ml penicillin. Medium was removed and cells were washed with 200 μl phenol-free DMEM. AMPs were diluted in phenol-free DMEM and preincubated with MBP-p33 for 1 h at 37°C. Samples were transferred to the 96-well plates and incubated for 16 h in 37°C, 5% CO2. The following day, 90 μl of the supernatant was transferred to a new 96-well plate, and lactate dehydrogenase (LDH) substrate mix was added. LDH release was measured at 490 nm using the Tox-7 kit (Sigma-Aldrich) according to manufacturers’ instructions. LDH release was calculated as percentage of total lysis with Tox-7 Lysis buffer.

Colocalization of endogenous p33 and AMPs

HUVECs were grown to confluency on 96-well plates in EBM-2 medium with EGM-2 BulletKit, CC-3162 (Lonza). Cells were washed in PBS and incubated with LL37 (2.5 or 20 μM) or β-defensin 3 (2.5 or 20 μM) for 16 h in 37°C, 5% CO2. CHAPS detergent (1%) was added for 30 min to dissolve the cells into membrane blebs followed by addition of EM-fix. Abs against p33, LL37, and β-defensin 3 were conjugated with gold particles (p33: 5 nm, LL37: 10 nm, and β-defensin 3: 10 nm) and incubated with the cell membrane blebs for 30 min at 4°C, and negatively stained with uranyl formate before electron microscopy.

Statistical analysis

Statistical analysis was performed using GraphPad Prism, Version 5.00. The p value was determined by using the unpaired t test (comparison of two groups). All experiments were performed at least three times, if not otherwise mentioned. The bars in the figures indicate SEM.

Results

p33 binds to AMPs

Structural analysis revealed that p33 is a doughnut-shaped protein consisting of three monomers with an asymmetric charge distribution and patches with an accumulation of negatively charged amino acids (24). It has been suggested that these domains play an important role in the interaction with p33 ligands (25). Notably, many peptides with antimicrobial activity have like HKH20 (pI 11.2), a positive net charge and a pI, which is in the similar range or even higher (26). We therefore wished to test whether it is a common feature of p33 to bind to these peptides. To this end, an indirect ELISA was established where a panel of AMPs with a positive net charge (Table I) was coated onto microtiter plates and probed with MBP-p33. Fig. 1 shows that p33 binds to HKH20, NAT26, GGL27, the β-defensins 2, 3 and 4, and to some extent to LL37, but not to β-defensin 1 and the α-defensins 3, 5, and 6, respectively. GGL27(S) is a derivative of GGL27 in which selected positive amino acids were replaced by a serine. This modification led to a change of the net charge from +8 to +1 and a complete loss of GGL27(S)’s affinity for p33 (Fig. 1). Together, these data demonstrate that p33 binds to some but not all AMPs tested.

View this table:
Table I. AMPs screened for p33 binding
FIGURE 1.
FIGURE 1.

p33 binds to AMPs. Microtiter plates were coated overnight with 5 μM AMPs and probed with 2.5 μg/ml MBP-p33. Binding was detected with a mAb against p33 and visualized by a peroxidase-conjugated secondary Ab. One representative experiment out of three is shown.

MBP-p33 has an apparent m.w. of 75.5 kDa (pI 4.6) as determined by SDS-PAGE analysis. After removal of the MBP tag by factor Xa cleavage, recombinant p33 migrates as a 33-kDa monomer when examined by SDS-PAGE under denaturing and nondenaturing conditions (data not shown). However, additional gel filtration experiments suggest that p33 forms homotrimers (data not shown), which was confirmed when the protein was negatively stained and visualized by transmission electron microscopy (Fig. 2). For further electron microscopy analysis, AMPs were labeled with colloid 5 nm gold particles and incubated with p33. Fig. 2 shows that p33 can bind up to three peptides (HKH20, GGL27, LL37, and the β-defensins 2 and 3, respectively). No binding was detected when α-defensin 3 was tested.

FIGURE 2.
FIGURE 2.

Negative staining electron microscopy of p33 in complex with AMPs. (upper panel) An overview of p33 is shown. (lower panels) p33 molecules in the absence of a ligand or bound to gold-labeled AMPs are depicted. Arrowheads point to gold-labeled AMPs (last image for each AMP). Scale bars, 100 nm (upper panel), 15 nm (lower panels).

In the next series of experiments, we sought to map the AMPs binding site(s) in p33. For this purpose, a panel of 15 overlapping peptides was synthesized, each consisting of 20 aa and spanning the entire p33 sequence (Table II). These peptides were tested in a competitive ELISA for their ability to displace the binding of p33 to the AMPs. Fig. 3A and Supplemental Fig. 2 show that TEA20 (aa 115–135) with a positive net charge of +2, inhibited the binding of p33 to all AMPs. DGE20 (aa 143–163) with a net charge of −7, in contrast, did not have any inhibitory activity, and CHY20 (aa 185–205 and net charge of −7) was able to block the binding of some, but not all, AMPs. Other peptides, such as TGE20 (aa 213–233), TDS20 (aa 227–247), and LEH20 (aa 262–282), respectively, had an effect on one or two AMPs. These findings may indicate that the AMP binding site in p33 is formed by a discontinuous epitope involving only parts of the binding peptides that are negatively charged (so-called binding spots), and indeed, Fig. 3B revealed that the three major binding epitopes are located in close vicinity (Fig. 3B).

View this table:
Table II. Overlapping p33 peptides
FIGURE 3.
FIGURE 3.

Mapping of the AMP binding sites in p33. (A) Microtiter plates were coated with 5 μM AMPs, followed by incubation with MBP-p33 and a panel of p33-derived peptide (10 μM; see Table II). Binding was detected by using a specific Ab against p33 and a secondary peroxidase-conjugated Ab. Values are described as percentage of absorbance compared with MBP-p33 alone. ***<25% absorbance, **25–50% absorbance, *50–75% absorbance. (B) The three-dimensional structure of p33 is shown, and the AMP-binding peptides are indicated: TEA20 in red, CHY20 in blue, and TGE20 in green (Jmol: an open-source Java viewer for chemical structures in 3D; http://www.jmol.org).

p33 inhibits the antimicrobial activity of AMPs

The AMPs used in this study have been shown to explore a broad spectrum of antimicrobial activity (27–30). Thus, a set of experiments was designed to determine whether p33 is able to counteract the ability of the AMPs to kill the E. coli strain ATCC 25922. Fig. 4 shows that all peptides tested, except the α-defensins and GGL27(S) (data not shown), showed antimicrobial activity against the ATCC 25922 strain, which was in the same range as described by others. Fig. 4 also depicts that bacterial killing was almost completely blocked when equimolar concentrations of p33 were applied to the reaction mixture. p33 did not affect bacterial survival in the absence of AMPs, and no binding of p33 to the ATCC 25922 strain was detected that could theoretically provide an alternative explanation for a protective effect that does not require a binding of p33 to AMPs (data not shown). Taken together, these data implicate that binding of p33 to AMPs neutralizes their antimicrobial activity.

FIGURE 4.
FIGURE 4.

p33 inhibits the antimicrobial activity of AMPs. The E. coli ATCC 25922 strain was grown to midlog phase, diluted to 2 × 106 CFU/ml, and incubated with different AMPs: HKH20 (1.5 μM), NAT26 (1.5 μM), GGL27 (1.5 μM), LL37 (6 μM), β-defensin 2 (3 μM), β-defensin 3 (3 μM), β-defensin 4 (1.5 μM) in the presence or absence of 10 μM MBP-p33. Samples were diluted and plated out on Todd Hewitt agar plates, and colonies were counted the following day. ***p < 0.001. Data are mean ± SE values from three separate experiments.

Localization of p33

The ability of p33 to dampen the antimicrobial effect of AMPs seems at first glance counterproductive. We therefore wished to test whether this feature of p33 is of pathophysiological importance. Notably, the concentration of AMPs in plasma can increase under inflammatory conditions (28, 29, 31, 32); thus, we measured whether this will also lead to increased plasma levels of p33. To this end, we used ELISA and Western blot analysis, but we were unable to detect p33 with both assays in plasma samples from healthy donors or sepsis patients (detection limit 6 ng/ml, data not shown). We noted, however, that human endothelial cells (EA.hy926 cells, an umbilical vein cell line), but not erythrocytes, express p33 express as determined by Western blot analysis (Fig. 5A). Further immunohistological examination revealed that p33 is located at the cell membrane of EA.hy926 cells (Fig. 5B). These findings are in line with a report by Guo and colleagues (32), who used bone marrow endothelial cells in their study. To test whether p33 is also expressed in vivo, we immunostained lung tissue from BALB/c mice and also here we saw specific expression of p33 in the endothelial outer lining of the blood vessels (Fig. 5C). These findings suggest that the interaction between p33 and AMPs takes place at cellular surfaces rather than in the circulation.

FIGURE 5.
FIGURE 5.

EA.hy926 cells, but not human erythrocytes, express p33. (A) EA.hy926 cells and human erythrocytes were lysed (see Materials and Methods) and ran on 10% SDS-PAGE. Recombinant p33 served as control. (B) EA.hy926 cells were fixed with paraformaldehyde and stained with an isotype control mAb (left panel) or an mAb to p33 (right panel) followed by a secondary goat anti-mouse mAb Alexa Fluor 488. Nuclei are visualized in blue (DAPI) and p33 in green. (C) Lungs from BALB/c mice were fixed, sectioned, and stained with a rabbit isotype control polyclonal Ab (left panel) or purified total IgG from rabbit anti-p33 sera (right panel). A specific p33 staining is visualized in the endothelial lining of the blood vessels (arrows). Scale bars, 10 μm.

p33 protects host cells from lysis by AMPs

It has been reported that AMPs at high concentrations can exert cytotoxic activities against host cells such as keratinocytes, lung epithelial cells, and endothelial cells, and intratracheal administration of β-defensin 3 into C57BL/6 mice has been shown to cause pulmonary inflammation and tissue injury (7, 33). To test whether p33 can prevent AMP-induced damage of eukaryotic cells, we performed hemolysis assay with human erythrocytes, because these cells have not been reported to express endogenous p33, which was also seen in our assays (Fig. 5A). Purified human erythrocytes were incubated with AMPs, cells were then pelleted, and the release of hemoglobin from lysed erythrocytes was measured in the supernatants. Fig. 6A shows that LL37 and β-defensin 3 had hemolytic activity, whereas HKH20, NAT26, GGL27, and the β-defensins 2 and 4 showed no effect. Importantly, the activity of the two hemolytic AMPs was decreased to baseline levels when preincubated with p33, and it was found that the effect of AMPs and p33 was dose dependent (data not shown). Similar results were obtained when the effect of AMPs on endothelial cells was studied. EA.hy926 cells were incubated with different AMPs, and membrane disruption was monitored by measurement of LDH leakage into the cell medium. As shown in Fig. 6B, LL37 and β-defensin 3 caused in these experiments significant cell damage; also in these experiments, preincubation with p33 abolished the cytotoxic effect of these peptides.

FIGURE 6.
FIGURE 6.

p33 inhibits the cytolytic activity of LL37 and β-defensin 3 against erythrocytes and endothelial cells. (A) AMPs (40 μM) were incubated (60 min at 37°C) with washed and diluted human erythrocytes (5% v/v) in the absence or presence of 40 μM MBP-p33. Samples were centrifuged and the release of hemoglobin was measured in the supernatant at 540 nm. Hemolysis was calculated as percentage of cells treated with Tox-7 Lysis buffer. ***p < 0.001. Data are mean ± SE values from three separate experiments. (B) EA.hy926 cells were grown confluent in 96-well plates. AMPs (20 μM) were added to the EA.hy926 cells in the absence or presence of 20 μM MBP-p33 for 16 h at 37°C, 5% CO2. The release of LDH was measured in the supernatants, and cell lysis was calculated as percentage of cells treated with Tox-7 Lysis buffer. **p < 0.01. Data are mean ± SE values from three separate experiments.

Endogenous p33 colocalizes with LL37 and β-defensin 3 on the surface of endothelial cells

To test whether endogenous p33 on endothelial cells interacts with LL37 and β-defensin 3, we prepared membrane fractions as described in Materials and Methods. These fractions were then incubated with low concentrations of LL37 and β-defensin 3 (2.5 μM) followed by immunodetection with gold-labeled Abs against the two peptides (10 nm) and p33 (5 nm). Ab binding was visualized by negative staining electron microscopy. Fig. 7A and 7B illustrate that at low concentration, LL37 and β-defensin 3 bind in close proximity to p33, and that this interaction did not cause membrane damage. When using higher concentrations (20 μM) of LL37 and β-defensin 3, we noticed that the peptides could also attach to the membrane without binding to p33 (Fig. 7C, 7D). This interaction led to membrane perforation leading to released blebs, suggesting that once the levels of LL37 and β-defensin 3 exceed the concentration of endogenous p33, endothelial cells are not any longer protected from a cytolytic attack and started to become perforated.

FIGURE 7.
FIGURE 7.

Endogenous p33 colocalizes with LL37 and β-defensin 3. Electron microscopic analysis of cell membranes treated with low dose (2.5 μM; A) or high dose (20 μM; B) of LL37. Membrane damage after incubation with high-dose LL37 is indicated by the formation of blebs, which are secreted from the plasma membrane. (C and D) Membrane damage caused by treatment with low (2.5 μM) or high dose (20 μM) of β-defensin 3 is shown. Endogenous p33 was immunostained by using a gold-labeled (5 nm gold) polyclonal Ab, whereas LL37 (A, B) and β-defensin 3 (C, D) were labeled with 10 nm gold particles. Scale bar, 50 nm.

Taken together, our data suggest that p33 has an important regulatory function in the early immune response to infection because it prevents the otherwise self-destructive activity of AMPs against host cells.

Discussion

Today, it is generally believed that many, if not most, peptides with antimicrobial activity are multifunctional, and they are therefore also referred to as host defense peptides (34). LL37, as well as α- and β-defensins, are probably the best characterized AMPs and apart from their ability to kill a huge variety of bacterial species, they have been shown to be proinflammatory and anti-inflammatory mediators, promote wound healing, and play important roles in chemotaxis, fertility, and cancer (for reviews, see Refs. 35–37). Under pathological conditions, AMP levels can become upregulated and their overexpression has been correlated with clinical complications such as psoriasis, circulatory derangement in severe infectious diseases, chronic obstructive pulmonary disease, and tumorigenesis (11, 38–40). However, thus far, not much is known about whether the human host has established mechanisms that can counteract these deleterious side effects.

In this study, we show that p33, a multicompartmental negatively charged protein, binds to a series of AMPs and is able to neutralize their lytic activity, thereby protecting eukaryotic cells from AMP-mediated lysis. To study the interaction between p33 and AMPs, we characterized the protein–protein interaction at the molecular level and correlated these findings with the physiological function of p33. To this end, we performed ELISA, competitive binding studies, and electron microscopy analyses to identify p33-binding AMPs and map AMP-binding sites in p33. Our results show that p33 binds to many, but not all, AMPs tested. By using synthetic peptides, we found that the main interaction site in p33 is located at an epitope that is close to the N terminus of the protein that forms a β-sheet structure, but other binding epitopes also were identified. Interestingly, all interaction sites are located in close proximity when looking at the three-dimensional structure of p33 (Fig. 3B), suggesting that the AMP binding site is formed by a discontinuous epitope. Our results further revealed that p33 is efficient in neutralizing the lytic effect of AMPs toward bacteria and eukaryotic cells, which raises the question about p33’s physiological function. The neutralization of the antimicrobial effect of AMPs would be counterproductive and seems therefore unlikely. This is also further supported by our findings that we were unable to detect soluble p33 in plasma samples from healthy donors or sepsis patients. We therefore believe that p33 has an important function as membrane-bound protein where it protects eukaryotic cells from an AMP attack. For example, LL37 and β-defensin 3 were able to lyse endothelial cells and erythrocytes in our experiments, and the addition of p33 was able to prevent this damage. Both AMPs (LL37 in form of its cathelicidin precursor) are stored in blood cells including neutrophils, monocytes, and thrombocytes, and can be released upon stimulation (40, 41).

Our results show that p33 is expressed on endothelial cells and in the endothelial lining of mice lungs, which is a common site for bacterial infection and increased antimicrobial activity. Under infectious conditions, neutrophils attach to endothelial cells to breach the vascular barrier and reach the site of infection. This is a critical step, because adhesion and transmigration trigger activated neutrophils to degranulate and release their content in close proximity to the endothelial cells. As a consequence, endothelial cells will be locally exposed to high levels of AMPs including, for instance, LL37, which may by far exceed the concentrations of AMPs found in the plasma of septic patients). Interestingly, it has been reported that endothelial cells or other cells can upregulate p33 on their surface under inflammatory conditions. It is therefore tempting to speculate that this presents a self-defense mechanism to protect cell damage caused by adhering neutrophils or platelets. Taken together, our results suggest that p33 has an important function in the early immune response to infection because it helps AMPs to attack the invading pathogen without causing harm to host cells.

Disclosures

The authors have no financial conflicts of interest.

Acknowledgments

We thank Monica Heidenholm, Helena Siller, and Maria Baumgarten for excellent technical assistance.

Footnotes

  • This work was supported by the foundations of Alfred Österlund, Crafoord, Greta and Johan Kock, and Knut and Alice Wallenberg; the Ragnar Söderberg Foundation; the Medical Faculty, Lund University; and the Swedish Research Council.

  • The online version of this article contains supplemental material.

  • Abbreviations used in this article:

    AMP
    antimicrobial peptide
    HK
    high molecular weight kininogen
    LDH
    lactate dehydrogenase
    MBP
    maltose-binding protein
    PBST
    PBS containing 0.05% Tween 20 and 0.5% BSA
    pI
    isoelectric point
    RT
    room temperature.

  • Received March 4, 2013.
  • Accepted September 25, 2013.

References

  1. ↵
    1. Kim D. D.,
    2. W. C. Song
    . 2006. Membrane complement regulatory proteins. Clin. Immunol. 118: 127–136.
    OpenUrlCrossRefPubMed
  2. ↵
    1. Martin B. K.
    2007. Transcriptional control of complement receptor gene expression. Immunol. Res. 39: 146–159.
    OpenUrlCrossRefPubMed
  3. ↵
    1. Longhi M. P.,
    2. C. L. Harris,
    3. B. P. Morgan,
    4. A. Gallimore
    . 2006. Holding T cells in check—a new role for complement regulators? Trends Immunol. 27: 102–108.
    OpenUrlCrossRefPubMed
  4. ↵
    1. Pasupuleti M.,
    2. A. Schmidtchen,
    3. M. Malmsten
    . 2012. Antimicrobial peptides: key components of the innate immune system. Crit. Rev. Biotechnol. 32: 143–171.
    OpenUrlCrossRefPubMed
  5. ↵
    1. Matsuzaki K.
    1999. Why and how are peptide-lipid interactions utilized for self-defense? Magainins and tachyplesins as archetypes. Biochim. Biophys. Acta 1462: 1–10.
    OpenUrlCrossRefPubMed
  6. ↵
    1. Boman H. G.
    2003. Antibacterial peptides: basic facts and emerging concepts. J. Intern. Med. 254: 197–215.
    OpenUrlCrossRefPubMed
  7. ↵
    1. Niyonsaba F.,
    2. H. Ushio,
    3. I. Nagaoka,
    4. K. Okumura,
    5. H. Ogawa
    . 2005. The human beta-defensins (-1, -2, -3, -4) and cathelicidin LL-37 induce IL-18 secretion through p38 and ERK MAPK activation in primary human keratinocytes. J. Immunol. 175: 1776–1784.
    OpenUrlAbstract/FREE Full Text
    1. Lehrer R. I.,
    2. A. K. Lichtenstein,
    3. T. Ganz
    . 1993. Defensins: antimicrobial and cytotoxic peptides of mammalian cells. Annu. Rev. Immunol. 11: 105–128.
    OpenUrlCrossRefPubMed
    1. Klüver E.,
    2. S. Schulz-Maronde,
    3. S. Scheid,
    4. B. Meyer,
    5. W. G. Forssmann,
    6. K. Adermann
    . 2005. Structure-activity relation of human beta-defensin 3: influence of disulfide bonds and cysteine substitution on antimicrobial activity and cytotoxicity. Biochemistry 44: 9804–9816.
    OpenUrlCrossRefPubMed
  10. ↵
    1. Kasetty G.,
    2. P. Papareddy,
    3. M. Kalle,
    4. V. Rydengård,
    5. M. Mörgelin,
    6. B. Albiger,
    7. M. Malmsten,
    8. A. Schmidtchen
    . 2011. Structure-activity studies and therapeutic potential of host defense peptides of human thrombin. Antimicrob. Agents Chemother. 55: 2880–2890.
    OpenUrlAbstract/FREE Full Text
  11. ↵
    1. Berkestedt I.,
    2. H. Herwald,
    3. L. Ljunggren,
    4. A. Nelson,
    5. M. Bodelsson
    . 2010. Elevated plasma levels of antimicrobial polypeptides in patients with severe sepsis. J. Innate Immun. 2: 478–482.
    OpenUrlCrossRefPubMed
  12. ↵
    1. Herwald H.,
    2. J. Dedio,
    3. R. Kellner,
    4. M. Loos,
    5. W. Müller-Esterl
    . 1996. Isolation and characterization of the kininogen-binding protein p33 from endothelial cells. Identity with the gC1q receptor. J. Biol. Chem. 271: 13040–13047.
    OpenUrlAbstract/FREE Full Text
  13. ↵
    1. Peerschke E. I.,
    2. T. K. Murphy,
    3. B. Ghebrehiwet
    . 2003. Activation-dependent surface expression of gC1qR/p33 on human blood platelets. Thromb. Haemost. 89: 331–339.
    OpenUrlPubMed
  14. ↵
    1. Ghebrehiwet B.,
    2. B. L. Lim,
    3. R. Kumar,
    4. X. Feng,
    5. E. I. Peerschke
    . 2001. gC1q-R/p33, a member of a new class of multifunctional and multicompartmental cellular proteins, is involved in inflammation and infection. Immunol. Rev. 180: 65–77.
    OpenUrlCrossRefPubMed
  15. ↵
    1. Ghebrehiwet B.,
    2. E. I. B. Peerschke
    . 1998. Structure and function of gC1q-R: a multiligand binding cellular protein. Immunobiology 199: 225–238.
    OpenUrlCrossRefPubMed
  16. ↵
    1. Nordahl E. A.,
    2. V. Rydengård,
    3. M. Mörgelin,
    4. A. Schmidtchen
    . 2005. Domain 5 of high molecular weight kininogen is antibacterial. J. Biol. Chem. 280: 34832–34839.
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Oehmcke S.,
    2. O. Shannon,
    3. M. von Köckritz-Blickwede,
    4. M. Mörgelin,
    5. A. Linder,
    6. A. I. Olin,
    7. L. Björck,
    8. H. Herwald
    . 2009. Treatment of invasive streptococcal infection with a peptide derived from human high-molecular weight kininogen. Blood 114: 444–451.
    OpenUrlAbstract/FREE Full Text
  18. ↵
    1. Nguyen T.,
    2. B. Ghebrehiwet,
    3. E. I. B. Peerschke
    . 2000. Staphylococcus aureus protein A recognizes platelet gC1qR/p33: a novel mechanism for staphylococcal interactions with platelets. Infect. Immun. 68: 2061–2068.
    OpenUrlAbstract/FREE Full Text
    1. Ghebrehiwet B.,
    2. L. Tantral,
    3. M. A. Titmus,
    4. B. J. Panessa-Warren,
    5. G. T. Tortora,
    6. S. S. Wong,
    7. J. B. Warren
    . 2007. The exosporium of B. cereus contains a binding site for gC1qR/p33: implication in spore attachment and/or entry. Adv. Exp. Med. Biol. 598: 181–197.
    OpenUrlCrossRefPubMed
  20. ↵
    1. Braun L.,
    2. B. Ghebrehiwet,
    3. P. Cossart
    . 2000. gC1q-R/p32, a C1q-binding protein, is a receptor for the InlB invasion protein of Listeria monocytogenes. EMBO J. 19: 1458–1466.
    OpenUrlAbstract/FREE Full Text
  21. ↵
    1. Bober M.,
    2. C. Enochsson,
    3. M. Collin,
    4. M. Mörgelin
    . 2010. Collagen VI is a subepithelial adhesive target for human respiratory tract pathogens. J. Innate Immun. 2: 160–166.
    OpenUrlCrossRefPubMed
  22. ↵
    1. Baschong W.,
    2. N. G. Wrigley
    . 1990. Small colloidal gold conjugated to Fab fragments or to immunoglobulin G as high-resolution labels for electron microscopy: a technical overview. J. Electron Microsc. Tech. 14: 313–323.
    OpenUrlCrossRefPubMed
  23. ↵
    1. Laemmli U. K.
    1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680–685.
    OpenUrlCrossRefPubMed
  24. ↵
    1. Jiang J.,
    2. Y. Zhang,
    3. A. R. Krainer,
    4. R. M. Xu
    . 1999. Crystal structure of human p32, a doughnut-shaped acidic mitochondrial matrix protein. Proc. Natl. Acad. Sci. USA 96: 3572–3577.
    OpenUrlAbstract/FREE Full Text
  25. ↵
    1. Ghebrehiwet B.,
    2. J. Jesty,
    3. R. Vinayagasundaram,
    4. U. Vinayagasundaram,
    5. Y. Ji,
    6. A. Valentino,
    7. N. Tumma,
    8. K. H. Hosszu,
    9. E. I. Peerschke
    . 2013. Targeting gC1qR domains for therapy against infection and inflammation. Adv. Exp. Med. Biol. 735: 97–110.
    OpenUrlCrossRefPubMed
  26. ↵
    1. Kasetty G.,
    2. P. Papareddy,
    3. M. Kalle,
    4. V. Rydengård,
    5. B. Walse,
    6. B. Svensson,
    7. M. Mörgelin,
    8. M. Malmsten,
    9. A. Schmidtchen
    . 2011. The C-terminal sequence of several human serine proteases encodes host defense functions. J. Innate Immun. 3: 471–482.
    OpenUrlCrossRefPubMed
  27. ↵
    1. Lehrer R. I.
    2004. Primate defensins. Nat. Rev. Microbiol. 2: 727–738.
    OpenUrlCrossRefPubMed
  28. ↵
    1. Papareddy P.,
    2. M. Kalle,
    3. G. Kasetty,
    4. M. Mörgelin,
    5. V. Rydengård,
    6. B. Albiger,
    7. K. Lundqvist,
    8. M. Malmsten,
    9. A. Schmidtchen
    . 2010. C-terminal peptides of tissue factor pathway inhibitor are novel host defense molecules. J. Biol. Chem. 285: 28387–28398.
    OpenUrlAbstract/FREE Full Text
  29. ↵
    1. Papareddy P.,
    2. V. Rydengård,
    3. M. Pasupuleti,
    4. B. Walse,
    5. M. Mörgelin,
    6. A. Chalupka,
    7. M. Malmsten,
    8. A. Schmidtchen
    . 2010. Proteolysis of human thrombin generates novel host defense peptides. PLoS Pathog. 6: e1000857.
    OpenUrlCrossRefPubMed
  30. ↵
    1. Frick I. M.,
    2. P. Akesson,
    3. H. Herwald,
    4. M. Mörgelin,
    5. M. Malmsten,
    6. D. K. Nägler,
    7. L. Björck
    . 2006. The contact system—a novel branch of innate immunity generating antibacterial peptides. EMBO J. 25: 5569–5578.
    OpenUrlAbstract/FREE Full Text
  31. ↵
    1. Hiratsuka T.,
    2. H. Mukae,
    3. H. Iiboshi,
    4. J. Ashitani,
    5. K. Nabeshima,
    6. T. Minematsu,
    7. N. Chino,
    8. T. Ihi,
    9. S. Kohno,
    10. M. Nakazato
    . 2003. Increased concentrations of human beta-defensins in plasma and bronchoalveolar lavage fluid of patients with diffuse panbronchiolitis. Thorax 58: 425–430.
    OpenUrlAbstract/FREE Full Text
  32. ↵
    1. Guo W. X.,
    2. B. Ghebrehiwet,
    3. B. Weksler,
    4. K. Schweitzer,
    5. E. I. Peerschke
    . 1999. Up-regulation of endothelial cell binding proteins/receptors for complement component C1q by inflammatory cytokines. J. Lab. Clin. Med. 133: 541–550.
    OpenUrlCrossRefPubMed
  33. ↵
    1. Saito A.,
    2. S. Ariki,
    3. H. Sohma,
    4. C. Nishitani,
    5. K. Inoue,
    6. N. Ebata,
    7. M. Takahashi,
    8. Y. Hasegawa,
    9. K. Kuronuma,
    10. H. Takahashi,
    11. Y. Kuroki
    . 2012. Pulmonary surfactant protein A protects lung epithelium from cytotoxicity of human β-defensin 3. J. Biol. Chem. 287: 15034–15043.
    OpenUrlAbstract/FREE Full Text
  34. ↵
    1. Schmidtchen A.
    2012. The multiple faces of host defence peptides and proteins. J. Innate Immun. 4: 325–326.
    OpenUrlCrossRefPubMed
  35. ↵
    1. Hazlett L.,
    2. M. Wu
    . 2011. Defensins in innate immunity. Cell Tissue Res. 343: 175–188.
    OpenUrlCrossRefPubMed
    1. Semple F.,
    2. J. R. Dorin
    . 2012. β-Defensins: multifunctional modulators of infection, inflammation and more? J. Innate Immun. 4: 337–348.
    OpenUrlCrossRefPubMed
  37. ↵
    1. Bucki R.,
    2. K. Leszczyńska,
    3. A. Namiot,
    4. W. Sokołowski
    . 2010. Cathelicidin LL-37: a multitask antimicrobial peptide. Arch. Immunol. Ther. Exp. (Warsz.) 58: 15–25.
    OpenUrlCrossRefPubMed
  38. ↵
    1. Morizane S.,
    2. R. L. Gallo
    . 2012. Antimicrobial peptides in the pathogenesis of psoriasis. J. Dermatol. 39: 225–230.
    OpenUrlCrossRefPubMed
    1. Liao Z.,
    2. J. Dong,
    3. X. Hu,
    4. T. Wang,
    5. C. Wan,
    6. X. Li,
    7. L. Li,
    8. L. Guo,
    9. D. Xu,
    10. F. Wen
    . 2012. Enhanced expression of human β-defensin 2 in peripheral lungs of patients with chronic obstructive pulmonary disease. Peptides 38: 350–356.
    OpenUrlCrossRefPubMed
  40. ↵
    1. Kawsar H. I.,
    2. A. Weinberg,
    3. S. A. Hirsch,
    4. A. Venizelos,
    5. S. Howell,
    6. B. Jiang,
    7. G. Jin
    . 2009. Overexpression of human beta-defensin-3 in oral dysplasia: potential role in macrophage trafficking. Oral Oncol. 45: 696–702.
    OpenUrlCrossRefPubMed
  41. ↵
    1. van der Does A. M.,
    2. P. Bergman,
    3. B. Agerberth,
    4. L. Lindbom
    . 2012. Induction of the human cathelicidin LL-37 as a novel treatment against bacterial infections. J. Leukoc. Biol. 92: 735–742.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section