Human b-defensin 3 binds to hemagglutinin B (rHagB), a non ... - Nature

Immunology and Cell Biology (2008) 86, 643–649 & 2008 Australasian Society for Immunology Inc. All rights reserved 0818-9641/08 $32.00 www.nature.com/icb

OUTSTANDING OBSERVATION

Human b-defensin 3 binds to hemagglutinin B (rHagB), a non-fimbrial adhesin from Porphyromonas gingivalis, and attenuates a pro-inflammatory cytokine response Lindsey C Pingel1, Karl G Kohlgraf1, Christopher J Hansen1, Christopher G Eastman1, Deborah E Dietrich1, Kindra K Burnell1, Rupasree N Srikantha1, Xiangjun Xiao1, Myriam Be´langer2, Ann Progulske-Fox2, Joseph E Cavanaugh3, Janet M Guthmiller4, Georgia K Johnson5, Sophie Joly1, Zoya B Kurago6, Deborah V Dawson1 and Kim A Brogden1,5 Regulatory mechanisms in mucosal secretions and tissues recognize antigens and attenuate pro-inflammatory cytokine responses. Here, we asked whether human b-defensin 3 (HBD3) serves as an upstream suppressor of cytokine signaling that binds and attenuates pro-inflammatory cytokine responses to recombinant hemagglutinin B (rHagB), a non-fimbrial adhesin from Porphyromonas gingivalis strain 381. We found that HBD3 binds to immobilized rHagB and produces a significantly higher resonance unit signal in surface plasmon resonance spectroscopic analysis, than HBD2 and HBD1 that are used as control defensins. Furthermore, we found that HBD3 significantly attenuates (Po0.05) the interleukin (IL)-6, IL-10, granulocyte macrophage colony stimulating factor (GM-CSF) and tumor-necrosis factor-a (TNF-a) responses induced by rHagB in human myeloid dendritic cell culture supernatants and the extracellular signal-regulated kinases (ERK 1/2) response in human myeloid dendritic cell lysates. Thus, HBD3 binds rHagB and this interaction may be an important initial step to attenuate a pro-inflammatory cytokine response and an ERK 1/2 response. Immunology and Cell Biology (2008) 86, 643–649; doi:10.1038/icb.2008.56; published online 19 August 2008 Keywords: HBD3; pro-inflammatory cytokine suppression; innate immunity; defensin

Antigen recognition and processing at mucosal surfaces determines the extent to which antigens induce an inflammatory response. Numerous mechanisms in mucosal secretions limit inflammatory responses to microbial antigens during their clearance, otherwise they would induce continual cytokine release and unrestricted mucosal inflammation.1–4 In fact, restraint of pro-inflammatory cytokine induction is highly desirable at mucosal surfaces of the oral and nasal cavities that are continually exposed to low levels of microbial antigens from commensal flora and microbial antigens from opportunistic pathogens.5 Mechanisms that suppress proinflammatory cytokine production and inhibit mitogen-activated protein kinase (MAPK) signaling pathways are currently under investigation and serve as targets to improve therapies to treat a wide variety of diseases and inflammatory disorders.6 One such mechanism may involve human b-defensin 3 (HBD3) (also known as DEFB103).

Human b-defensin 3 is a member of a diverse group of cationic antimicrobial peptides called defensins.7–11 Gene DEFB103 is found at chromosomal location 8p23 and has a polymorphic copy number.12,13 It contains 45 amino-acid residues, has a b-sheet structure in solution, forms dimers and has a net +11 charge.14 HBD3 is induced by bacteria; bacterial products; tumor-necrosis factor-a (TNF-a), interleukin (IL)1b and interferon (IFN)-g; and is expressed in epithelia of many organs and in non-epithelial tissues.15–19 It kills gram-negative bacteria, grampositive bacteria, fungi and inhibits some viral infections,15,20–23 yet can also chemoattract monocytes, macrophages, CD45 RA+/CD4+ T lymphocytes, mast cells and immature dendritic cells.15,24,25 In the epithelium and mucosal secretions of the oral and nasal cavities, HBD3 is ideally positioned to interact with an extensive and diverse group of both commensal and pathogenic microbial antigens. This includes those of the oral pathogen, Porphyromonas gingivalis, whose extracellular products are capable of inducing pro-inflammatory

1Dows Institute for Dental Research, College of Dentistry, The University of Iowa, Iowa City, IA, USA; 2Center for Molecular Biology and Department of Oral Biology, College of Dentistry, University of Florida, Gainesville, FL, USA; 3Department of Biostatistics, College of Public Health, The University of Iowa, Iowa City, IA, USA; 4Department of Periodontics, School of Dentistry, The University of North Carolina at Chapel Hill, Chapel Hill, NC, USA; 5Department of Periodontics, College of Dentistry, The University of Iowa, Iowa City, IA, USA and 6Department of Oral & Maxillofacial Pathology, Radiology and Medicine, New York University College of Dentistry, New York, NY, USA Correspondence: Professor KA Brogden, Department of Periodontics, The University of Iowa, Dows Institute for Dental Research, 801 Newton Road, Iowa City, IA 52242, USA. E-mail: [email protected] Received 15 April 2008; revised 3 July 2008; accepted 8 July 2008; published online 19 August 2008

HBD3 attenuates pro-inflammatory cytokine response LC Pingel et al 644

cytokines and producing intense inflammatory responses at mucosal surfaces.26–28 Porphyromonas gingivalis has five hemagglutinins (Hag) that facilitate microbial binding to host cells and erythrocytes.29 hagA encodes a 233.4 kDa protein that contains four repeating segments, each with hemagglutinating activity.30 hagB and hagC encode 42.0 and 39.3 kDa proteins, respectively.30–32 hagD encodes a 187.9 kDa protein32,33 and hagE encodes a 185.7 kDa protein.34 HagA and HagD have 73.8% homology; HagA and HagE have 93% homology; and HagB and HagC have 98% homology.32 HagB is a major virulence factor and among the more closely studied hemagglutinins of P. gingivalis. In the course of our studies, we first observed that HBD3, co-administered intranasally with recombinant HagB (rHagB) in mice, attenuates rHagB-induced pro-inflammatory cytokine responses to rHagB in nasal lavage fluids. We also observed that HBD3 binds to P. gingivalis antigens. This led us to hypothesize that HBD3 directly binds to microbial antigens such as the adhesin rHagB. We further hypothesize that the binding between HBD3 and rHagB alters the binding of rHagB with the dendritic cell surface receptors and thus attenuates the rHagB-induced pro-inflammatory cytokine responses of the dendritic cells by inhibiting MAPK pathways. RESULTS Our hypothesis that HBD3 binds to rHagB was tested using surface plasmon resonance spectroscopy. For this, 100 mg ml
1 rHagB was immobilized to a carboxymethylated dextran CM5 chip surface. Then

1.0, 10.0, 50.0 and 100.0 mg ml
1 HBD3 was passed over the immobilized rHagB surface. Two other b-defensins, HBD1 and HBD2, were used as control peptides. The sensorgrams showing defensin binding to immobilized rHagB were vastly different (Figures 1a and b). Binding with HBD3 was fast (Ka¼1.80104) and stable (Kd¼0.6010–3) with affinity (KD¼17.9 nM). Binding with HBD1 was slow (Ka¼0.21104) and stable (Kd¼0.0110–3) with affinity (KD¼8.76 nM) and binding with HBD2 was fast (Ka¼17.60104), yet less stable (Kd¼9.7410–3) with lesser affinity (KD¼49.96 nM). HBD3 produced a higher resonance unit (RU) signal response than both HBD1 and HBD2 (Figure 1a). Lower RU signal responses were seen using HBD1 (Figure 1b) and HBD2 (Figure 1c). Significant Spearman rank values confirmed correlations among increasing RU signal responses and increasing concentrations for HBD3 (r¼0.92, P¼0.0001) and HBD2 (r¼0.77, P¼0.0053), but not HBD1 (r¼0.21, P¼0.54). The concentration of HBDs played an important role (Figure 1d). At 1.0 mg ml
1, there were differences in the RU signal values among the three peptides tested (P¼0.0464). At 10.0 mg ml
1, differences were apparent (P¼0.0036) and RU signal values of HBD3 were significantly higher than those of HBD1 and HBD2. At 50.0 mg ml
1, differences in the association of the three peptides remained significant (P¼0.0107) and pairwise comparisons showed that the RU signal values of HBD3 were significantly higher than those of HBD1 but not HBD2. Finally, at 100.0 mg ml
1, the RU signal values were very high for HBD3 and suggestive of significance in the overall Kruskal–Wallis test.

Figure 1 Surface plasmon resonance spectroscopy sensorgrams showing concentration dependent binding of b-defensins to immobilized rHagB from P. gingivalis. Figures (a–c) show the overlay of four binding isotherms generated from 100.0 mg ml
1 (top line), 50.0 mg ml
1 (second line), 10.0 mg ml
1 (third line) and 1.0 mg ml
1 (bottom line) of HBD3, HBD1 and HBD2. The binding affinity of each defensin to immobilized rHagB was determined by kinetic analysis calculated from the on (Ka) and off (Kd) rates of the four binding isotherms together in each graph using BIAevaluation software version 4.1 (Biacore, Inc., Piscataway, NJ, USA). The resonance unit (RU) signal responses of (a) HBD3 were much higher than (b) HBD1 and (c) HBD2 at concentrations of 1.0, 10.0, 50.0 and 100.0 mg ml
1. When compared (d), the RU signal responses of HBD3 were significantly higher (n¼3) (*Po0.05, HBD3 vs HBD2 and HBD1). Immunology and Cell Biology

HBD3 attenuates pro-inflammatory cytokine response LC Pingel et al 645

Recombinant HagB induces a cytokine response in a variety of rodent cell types: INF-g, IL-2, IL-4 and IL-10 by rat splenic lymphoid cells;35 IL-4, IL-5 and IFN-g by murine CD4+ T cells;36 and TNF-a, IL-12p40, IFN-g and IL-10 by mouse peritoneal macrophages.27 Human oral keratinocytes exposed to rHagB produce IL-6 and IL-8 (Supplementary Figure 1). The cytokine response of oral mucosa cells including human myeloid dendritic cells to rHagB is largely unknown. Therefore, we examined the cytokine response of human myeloid dendritic cells to rHagB and HBD3. rHagB was diluted tenfold from 0.1 to 0.001 mg ml
1 to determine the minimal concentration of rHagB capable of inducing a significant pro-inflammatory cytokine and chemokine response in human myeloid dendritic cells (Supplementary Figure 2a–c). rHagB at a concentration of 0.1 mg ml
1 induced a significant (Po0.05) IL-6 response and 0.01 and 0.1 mg ml
1 rHagB induced significant (Po0.05) IL-8 and TNF-a responses compared with lesser concentrations of rHagB, 0.01 M phosphate-buffered saline, pH 7.2 without rHagB and similar concentrations of boiled, inactivated rHagB. Significant (Po0.05) results were adjusted for multiple comparisons using Dunnett’s procedure. Based on these findings, 0.05 mg ml
1 rHagB was selected for use with the human myeloid dendritic cell cultures. HBD3 was diluted tenfold (from 1.0 to 0.01 mg ml
1) to determine the highest concentration of HBD3 incapable of inducing proinflammatory cytokine and chemokine responses in human myeloid dendritic cells (Supplementary Figure 3a–c). HBD3, at concentrations of 1.0, 0.1 and 0.01 mg ml
1, did not induce IL-6, IL-8 or TNF-a responses significantly different from each other or the baseline cytokine response induced by 0.01 M phosphate-buffered saline, pH 7.2 without HBD3. Therefore, 1.0 mg ml
1 HBD3 was selected for use with human myeloid dendritic cell cultures. Second, we assessed a broader pro-inflammatory cytokine and chemokine response in human myeloid dendritic cells exposed to HBD3 and rHagB mixtures (Figure 2a). Human myeloid dendritic cells incubated with 0.05 mg ml
1 rHagB produced pro-inflammatory cytokines (IL-6, granulocyte macrophage colony stimulating factor (GM-CSF), TNF-a and IL-12p40), IL-10 and the chemokines IL-8 (CXCL8), IP-10 (CXCL10), MCP-1 (CCL2), MIP-1a (CCL3) and RANTES (CCL5). HBD3 significantly attenuated (Po0.05) only the IL-6 (Figure 2b), IL-10 (Figure 2c), GM-CSF (Figure 2d) and TNF-a (Figure 2e) responses induced by rHagB. The MAPK pathways are important in controlling the type and magnitude of the inflammatory response to P. gingivalis and its extracellular products.28,37 In mouse peritoneal macrophages, rHagB induces extracellular signal-regulated kinases (ERK 1/2), c-Jun N-terminal kinases (JNK 1/2) and p38 protein kinase within 10–30 min.27 In human dendritic cells, the signaling pathways induced by rHagB (and HBD3) are not known. Therefore, rHagB and HBD3 were diluted 10-fold from 10.0 to 1.0 mg ml
1 and added to human myeloid dendritic cells for 10, 20 and 30 min. A concentration of 10 mg ml
1 rHagB induced significant (Po0.05) p38 (Figure 3a), JNK 1/2 (Figure 3b) and ERK 1/2 (Figure 3c) responses in human myeloid dendritic cell lysates within 10–30 min. HBD3 did not activate any of these pathways (Figures 3d–f). After 30 min, HBD3/rHagB mixtures in human myeloid dendritic cells induced significantly lower (Po0.05) signals of phosphorylated ERK 1/2 (Figure 3f) but not signals of phosphorylated JNK 1/2 (Figure 3e) or p38 (Figure 3d). DISCUSSION In this study, we hypothesized that HBD3 directly binds to bacterial antigens, attenuates antigen-induced pro-inflammatory cytokine responses and inactivates select antigen-induced signaling pathways

in human myeloid dendritic cells. Here we found that HBD3 binds to immobilized rHagB and produces a significantly higher RU signal in surface plasmon resonance spectroscopic studies than control defensins, HBD2 and HBD1. The binding of HBD3 to rHagB may be related to its overall net cationic charge of +11. However, it is not known whether a single motif or site within HagB interacts with dendritic cell receptors or if HBD3 binding disrupts a functional interactive conformation of rHagB. We also found that HBD3 significantly attenuates the IL-6, IL-10, GM-CSF and TNF-a cytokine responses induced by rHagB in human myeloid dendritic cell culture supernatants and significantly attenuates the ERK 1/2 response in human myeloid dendritic cell lysates. Thus, HBD3 serves as an upstream suppressor of cytokine signaling and regulates and attenuates pro-inflammatory cytokine responses to microbial antigens in mucosal secretions and tissues. Human b-defensin 3 is a predominant defensin in the oral cavity, produced and stored by cells in the gingival epithelium.19,38,39 Keratinocytes produce high levels of HBD3 protein and mRNA expression in healthy oral tissues suggesting that they likely serve as a source of HBD3 in this mechanism.19,40 HBD3 is produced by functionally active keratinocytes in the basal layer of the tissue from healthy subjects and the upper layer of the tissue from clinically healthy subjects.38 HBD3 concentrations in saliva, are quite high ranging from 0 to 6.21 mg ml
1.41–43 In contrast, HBD2 concentrations in saliva average 0.0095 mg ml
1.41 In a surveillance and regulatory-like function, HBD3 produced in the vicinity of commensal flora or opportunistic pathogens, likely binds to microbial products before presentation to adjacent immature dendritic cells thus attenuating pro-inflammatory cytokine production and maintaining periodontal homeostasis. Of course, if further infection and tissue destruction occurs, the attenuating mechanism would be unable to contain the ensuing inflammatory process. HBD1, HBD2 and HBD3 induced by bacteria, bacterial products, TNF-a, IL-1b and IFN-g would chemoattract and activate monocytes, macrophages, CD45 RA+/CD4+ T lymphocytes, mast cells and immature dendritic cells.15,24,25 The exact mechanism for the HBD3 attenuated rHagB cytokine response by human myeloid dendritic cells is unknown but this study suggests that binding between cationic HBD3 and rHagB may occur and be important early in the process. This is supported by the fact that: (1) HBD3 is known to bind to other pathogens;20 (2) the hemagglutinating activity of HagB can be inhibited by prior incubation of HagB with basic amino acids,44 which are found in abundance in HBD3; and (3) HBD3 strongly binds to immobilized rHagB as detected by surface plasmon resonance spectroscopy in this study. The ability of HBD3 to bind to rHagB could alter the ability of rHagB to bind with its appropriate cellular receptor(s) that is thought to be a protein of approximately 140 kDa (Iskandar I, Be´langer M and Progulske-Fox A. Potential receptor of Porphyromonas gingivalis hemagglutinin B on HCAEC. 82nd General Session IADR, Honolulu, HI. March 10–13, 2004) and thus partially block or modify signaling pathways such as MAPK pathways. The MAPK pathways include p38, JNK 1/2 and ERK 1/2. p38 and JNK are closely associated with inflammation and ERK is associated with growth and proliferation.45 In this study, HBD3 and rHagB mixtures induced significantly lower (Po0.05) levels of ERK 1/2 but did not induce lower levels of JNK 1/2 and p38 in human myeloid dendritic cells. p38 activates DNA transcriptional factor CREB; p38 and JNK activate the transcriptional complex containing activator protein-1 (AP-1); and ERK, JNK and p38 activate the transcriptional factor ELK1.1,45 IL-6, IL-10, GM-CSF and TNF-a are downstream Immunology and Cell Biology

646

Immunology and Cell Biology

HBD3 attenuates pro-inflammatory cytokine response LC Pingel et al

Figure 2 HBD3 attenuates a pro-inflammatory cytokine response in the supernatants of human myeloid dendritic cells to P. gingivalis rHagB. At 24 h, (a) human myeloid dendritic cells exposed to 0.05 mg ml
1 rHagB produced a variety of cytokines (n¼3). Shown are those cytokines induced by rHagB alone (listed on the y-axis as the cytokine (1)) or rHagB with HBD3 (listed on the y-axis as the cytokine (2)). Of these 22 cytokine responses, only the (b) IL-6, (c) IL-10, (d) GM-CSF and (e) TNF-a responses were significantly attenuated (Po0.05, n¼4) (*Po0.05, treated vs rHagB control).

HBD3 attenuates pro-inflammatory cytokine response LC Pingel et al 647

Figure 3 HBD3 attenuates a P. gingivalis rHagB induced ERK response in human myeloid dendritic cells. At 30 min, human myeloid dendritic cells exposed to 1 or 10 mg ml
1 rHagB produced dose and time dependent (a) p38, (b) JNK and (c) ERK responses. Human myeloid dendritic cells exposed to 10 mg ml
1 HBD3+10 mg ml
1 rHagB produced (d) p38, (e) JNK and (f) significantly lower (Po0.05), ERK responses. The lower ERK results were confirmed by using antibody arrays and a specific ELISA (box). Lipopolysaccharide, used as a control, did not induce a significant response. (a–c, *Po0.05, treated vs PBS control; d–f, *Po0.05, treated vs rHagB control).

products of these factors and partial inhibition of ERK 1/2 could explain the selective HBD3-mediated decrease in the production of these pro-inflammatory cytokines by human myeloid dendritic cells exposed to rHagB. Defensins are expressed in epithelia of many organs and in nonepithelial tissues and serve to protect these tissues from infection, prevent inflammation and promote adaptive immune responses. In some cases, defensin deficiencies are thought to be key factors in the pathogenesis of infection and inflammation through a compromise of innate immunity.46 Although early, it is tempting to speculate that polymorphisms in defensins or differences in copy number decreases their ability to protect these sites and increases the susceptibility of those individuals to infection and inflammation. Copy number polymorphisms and expression level variations of the a-defensins47 and b-defensins48 may lead to different defensin profiles and possibly correlate with susceptibility to bacterial infections.49,50 Copy number polymorphisms and expression level variations are likely involved in infections associated with Crohn’s disease,46,51 periodontal disease52 and cystic fibrosis.53 Specifically for HBD3, the three-copy number variant was the most frequent genotype occurring in 65.9% of 44 subjects followed by the two-copy number variant in 30.5% of 44 subjects and the four-copy number variant in 13.6% of 44 subjects.13 Ongoing work will identify the rHagB receptor on human dendritic cells; the molecular mechanism by which HBD3 binds to HagB; and the molecular mechanism attenuating the signaling pathway response induced by the HBD3 and rHagB ‘complex’. Additional work is needed to assess how polymorphisms affect the ability of HBD3 to regulate an inflammatory response. Together these results will lead to a better understanding of the early mechanisms of mucosal inflammation and potentially provide a variety of new therapeutic avenues for treatment and prevention of inflammatory responses.

METHODS Recombinant hemagglutinin B The rHagB gene of P. gingivalis (1.4 kb) was cloned into the vector pQE31 (QIAGEN Inc., Valencia, CA, USA) and the construct was designated pQE31TX1.54 The histidine-tagged rHagB was purified on a nickel-nitrilotriacetic acid affinity column by fast protein liquid chromatography (Bio-Rad Laboratories, Hercules, CA, USA) from E. coli M15(pREP4)pQE-31-TX1.54 The eluted protein was dialyzed against 500 mM sodium chloride and 10 mM Tris, pH 7.4, and was concentrated using polyethylene glycol 8000 (Fisher Scientific, Fair Lawn, NJ, USA).29 The composition and purity of rHagB was verified by SDSpolyacrylamide gel electrophoresis, western blot, mass spectrometry and amino acid analysis (High Resolution Mass Spectrometry Facility, University of Iowa, Iowa City, IA, USA). Lipopolysaccharide content was determined with the QCL-1000 Chromogenic Limulus Amebocyte Lysate Assay (Cambrex, Walkersville, MD, USA) and was found to be B0.095 ng per 0.05 mg rHagB. rHagB was boiled for 15 min and used as an inactivated control.

b-Defensins HBD1, HBD2 and HBD3 were purchased from PeproTech, Inc. (Rocky Hill, NJ, USA). The purity of these peptides was confirmed using an Acclaim PolarAdvantage II, C18, 5 mm, 120 A˚ analytical column (Dionex Corp., San Francisco, CA, USA) on a Summit HPLC system (Dionex) with Chromeleon software (Dionex) and eluted with a gradient of acetonitrile (0–100%) in 0.1% trifluroacetic acid. Lipopolysaccharide content was determined with the QCL1000 Chromogenic Limulus Amebocyte Lysate Assay (Cambrex) and was found to be B0.1–4.5 pg per 1.0 mg of these defensins.

Surface plasmon resonance spectroscopy Stock solutions of rHagB and HBDs were prepared in 10 mM HEPES, pH 7.4 containing 150 mM NaCl, 3 mM EDTA and 0.005% surfactant P20 (HBS-EP) buffer. The mass and composition of rHagB and HBDs were reassessed by MALDI-TOF and amino acid analysis (High Resolution Mass Spectometry Facility, University of Iowa, Iowa City, IA, USA). The final rHagB and HBD solutions were adjusted to contain 100 mg ml
1. Immunology and Cell Biology

HBD3 attenuates pro-inflammatory cytokine response LC Pingel et al 648 Surface plasmon resonance spectroscopy (Biacore 3000, Biacore, Inc., Piscataway, NJ, USA) was performed in CM5 sensor chips. 100 mg ml
1 rHagB was covalently attached to the carboxymethylated dextran/gold surface in two separate flow cells. The interaction between analytes was monitored by measuring the change in the RU signal. Non-specific association of mobile analytes with the carboxymethylated dextran surface or specific association among mobile analytes and immobilized analytes was characterized by an increase in RU signal over time in one of the CM5 flow cells compared with the baseline established by passing HBS-EP in the other CM5 flow cell at 20 ml min
1 for B2–4 min at 25 1C. Dissociation was analyzed for 3 min by passing HBS-EP through the flow cell and monitoring the surface plasmon RU signal. After each analysis, the sensor chip surfaces were regenerated with 2.5 M glycine and equilibrated with HBS-EP before the next analysis. The association of each concentration of HBD with rHagB was assessed in three replications. The kinetic association (ka) and dissociation (kd) rate constants were calculated using BIAevaluation, version 4.1 (Biacore, Inc.). The kinetic dissociation constant KD was calculated as kd/ka.

Human myeloid dendritic cells Human myeloid dendritic cells were generated from peripheral blood monocytes of normal donors as approved by the University of Iowa Institutional Review Board. Monocytes were enriched from the peripheral blood with RosetteSep Monocyte Enrichment Cocktail, isolated in Ficoll-Paque Plus (StemCell Technologies, Vancouver, BC, Canada), suspended in serum-free X-Vivo15 (Cambrex) and allowed to adhere to plastic plates at 37 1C. After 15–30 min incubation, unattached cells were rinsed away. The remaining monocytes, stained for 30 min on ice with fluorochrome-labeled antibodies for flow cytometry in an LSR II flow cytometer (BD Immunocytometry Systems, San Jose, CA, USA), were 99% for CD11c+ and 95–99% for CD14+. These cells were cultured for 6 days in serum-free X-Vivo-15 with gentamicin and supplemented on alternate days with 1000 U ml
1 recombinant human GM-CSF and 1000 U ml
1 IL-4 (R&D Systems, Minneapolis, MN, USA). The resulting cells, also stained with fluorochrome-labeled antibodies for flow cytometry, were 98–100% CD11C+, MHC class II+, CD80dim CD86dim DC-SIGN+, immature human myeloid dendritic cells.

Induction and detection of cytokines Human myeloid dendritic cells were adjusted to contain 1.0105 cells per ml. rHagB and HBD3 were mixed, incubated at 37 1C for 30 min, and added to 150 ml of tissue culture media containing 1.5104 cells per well in 96 well plates (the final concentration was estimated to be B0.5 pg rHagB per dendritic cell). At 24 h post-inoculation, the dendritic cell culture supernatants were removed, centrifuged to pellet loose cells and debris, and stored at
80 1C. Cytokines and chemokines (pg ml
1) in cell culture supernatants were determined using commercial multiplexed fluorescent bead-based immunoassays (Millipore, Billerica, MA, USA) in the Luminex 100 IS Instrument (Luminex, Austin, TX, USA). Kit 48-002 detects IL-1b, IL-6, IL-8, IL-10, IL-12 (p70) and TNF-a and was used to screen initial dendritic cell responses to rHagB and HBD3 (Supplementary Figures 1 and 2). Kit 48-011 detects T helper 1 cytokines (IL-2, IL-12 (p70) and IFN-g), T helper 2 cytokines (IL-3, IL-4, IL-5, IL-10 and IL-13), pro-inflammatory cytokines (IL-1a, IL-1b, IL-6, GM-CSF, TNF-a and IL-12 (p40)), chemokines (CXCL8/IL-8, CXCL10/ IP-10, CCL2/MCP-1, CCL3/MIP-1a, CCL5/RANTES and CCL11/eotaxin) and regulators of T cell and natural killer cell activation and proliferation (IL-7 and IL-15). Concentrations of cytokines in each sample were extrapolated from standards (2.3–5000 pg ml
1) using Beadview software (Millipore).

Induction and detection of MAPK signaling pathways Human myeloid dendritic cells were adjusted to contain 7.5105 cells per ml. rHagB and HBD3 were mixed, incubated at 37 1C for 30 min and added to 200 ml of tissue culture media containing 1.5105 cells per well in 96 well plates (the final concentration was estimated to be B13 pg rHagB per dendritic cell). At 10, 20 and 30 min post-inoculation, the dendritic cell culture supernatants were removed by vacuum filtration. The cells were washed with 100 ml ice-cold Tris-buffered saline and lysed in Cell Signalling Universal Lysis Buffer Immunology and Cell Biology

(Millipore) containing Tris buffered salts, detergents, phosphatase inhibitors and 1 mM sodium orthovanadate. Lysates were stored at
80 1C. A Multi-Pathway Signaling kit (Millipore) for phosphoprotein was used to detect changes in phosphorylated Erk/MAP kinase 1/2 (Thr185/Tyr187), STAT3 (Ser727), JNK (Thr183/Tyr185), p70 S6 kinase (Thr412), IkBa (Ser32), STAT5A/B (Tyr694/699), CREB (Ser133) and p38 (Thr180/Tyr182) in human myeloid dendritic cell lysates using the Luminex 100 IS Instrument (Luminex). The attenuation of ERK2 was verified using the human phospho-MAPK array kit (R&D Systems) and a STAR Phospho-ERK 1/2 (Thr185/Tyr187) ELISA Kit (Millipore). The final concentration was estimated to be B2 pg rHagB per dendritic cell).

Statistics The non-parametric Kruskal–Wallis test was used to compare differences in the distribution of RU signal values among the four concentrations of each defensin and to compare differences among the RU signal values of three defensins at each concentration. Adjustment for multiple pairwise comparisons was accomplished using the adaptation of the Tukey method recommended by Conover55 in conjunction with an overall 0.05 level of significance. The Spearman rank correlation was used to assess the possibility of an increasing or decreasing dose–response relationship between concentrations and RU signal values for each defensin using a nominal 0.05 level of significance. A log10 transformation was applied to the cytokine responses obtained from two-dimensional pattern of ten-fold dilutions of HBD3 mixed with ten-fold dilutions of rHagB to account for the positive skewness in the measurements and to make the assumption of normality more defensible. Analysis of variance (ANOVA) models were fitted to the log-transformed responses using SAS (version 9.1, SAS Institute, Cary, NC, USA). A one way ANOVA was used to analyze the effect of HBD3 on a specific concentration of rHagB (for example, 0.05 mg ml
1) and a Dunnett’s multiple comparison test was used to compare the group differences with that induced by rHagB alone. An overall 0.05 level of statistical significance was used in conjunction with the multiple comparisons adjustment.

ACKNOWLEDGEMENTS We thank JD Herd and EA Schmitt (Cairn Communications, Mahtomedi, MN) for assistance with the figures and manuscript. This work was supported by funds from T32 DE014678 and R01 DE014390 from the National Institute of Dental and Craniofacial Research, National Institutes of Health.

1 Han J, Ulevitch RJ. Limiting inflammatory responses during activation of innate immunity. Nat Immunol 2005; 6: 1198–1205. 2 Henson PM. Dampening inflammation. Nat Immunol 2005; 6: 1179–1181. 3 Mayer L, Shao L. Therapeutic potential of oral tolerance. Nat Rev Immunol 2004; 4: 407–419. 4 Anderson P. Post-transcriptional control of cytokine production. Nat Immunol 2008; 9: 353–359. 5 Fenner JE, Starr R, Cornish AL, Zhang JG, Metcalf D, Schreiber RD et al. Suppressor of cytokine signaling 1 regulates the immune response to infection by a unique inhibition of type I interferon activity. Nat Immunol 2006; 7: 33–39. 6 Bogoyevitch MA, Fairlie DP. A new paradigm for protein kinase inhibition: blocking phosphorylation without directly targeting ATP binding. Drug Discov Today 2007; 12: 622–633. 7 Selsted ME, Ouellette AJ. Mammalian defensins in the antimicrobial immune response. Nat Immunol 2005; 6: 551–557. 8 Brogden KA. Antimicrobial peptides: pore formers or metabolic inhibitors in bacteria? Nat Rev Microbiol 2005; 3: 238–250. 9 Ganz T. Defensins: antimicrobial peptides of innate immunity. Nat Rev Immunol 2003; 3: 710–720. 10 Schutte BC, Mitros JP, Bartlett JA, Walters JD, Jia HP, Welsh MJ et al. Discovery of five conserved b-defensin gene clusters using a computational search strategy. Proc Natl Acad Sci USA 2002; 99: 2129–2133. 11 Lehrer RI. Primate defensins. Nat Rev Microbiol 2004; 2: 727–738. 12 Hollox EJ, Armour JA, Barber JC. Extensive normal copy number variation of a betadefensin antimicrobial-gene cluster. Am J Hum Genet 2003; 73: 591–600. 13 Chen Q, Book M, Fang X, Hoeft A, Stuber F. Screening of copy number polymorphisms in human beta-defensin genes using modified real-time quantitative PCR. J Immunol Methods 2006; 308: 231–240.

HBD3 attenuates pro-inflammatory cytokine response LC Pingel et al 649 14 Schibli DJ, Hunter HN, Aseyev V, Starner TD, Wiencek JM, McCray Jr PB et al. The solution structures of the human beta-defensins lead to a better understanding of the potent bactericidal activity of HBD3 against Staphylococcus aureus. J Biol Chem 2002; 277: 8279–8289. 15 Garcia JR, Jaumann F, Schulz S, Krause A, Rodriguez-Jimenez J, Forssmann U et al. Identification of a novel, multifunctional beta-defensin (human beta-defensin 3) with specific antimicrobial activity. Its interaction with plasma membranes of Xenopus oocytes and the induction of macrophage chemoattraction. Cell Tissue Res 2001; 306: 257–264. 16 Harder J, Bartels J, Christophers E, Schroder JM. Isolation and characterization of human b-Defensin-3, a novel human inducible peptide antibiotic. J Biol Chem 2001; 276: 5707–5713. 17 Harder J, Bartels J, Christophers E, Schroder J-M. A peptide antibiotic from human skin. Nature 1997; 387: 861–862. 18 Joly S, Organ CC, Johnson GK, McCray Jr PB, Guthmiller JM. Correlation between betadefensin expression and induction profiles in gingival keratinocytes. Mol Immunol 2005; 42: 1073–1084. 19 Dunsche A, Acil Y, Dommisch H, Siebert R, Schroder JM, Jepsen S. The novel human beta-defensin-3 is widely expressed in oral tissues. Eur J Oral Sci 2002; 110: 121–124. 20 Hazrati E, Galen B, Lu W, Wang W, Ouyang Y, Keller MJ et al. Human alpha- and betadefensins block multiple steps in herpes simplex virus infection. J Immunol 2006; 177: 8658–8666. 21 Feng Z, Dubyak GR, Lederman MM, Weinberg A. Cutting edge: human beta defensin 3—a novel antagonist of the HIV-1 coreceptor CXCR4. J Immunol 2006; 177: 782–786. 22 Quinones-Mateu ME, Lederman MM, Feng Z, Chakraborty B, Weber J, Rangel HR et al. Human epithelial beta-defensins 2 and 3 inhibit HIV-1 replication. Aids 2003; 17: F39–F48. 23 Joly S, Maze C, McCray Jr PB, Guthmiller JM. Human beta-defensins 2 and 3 demonstrate strain-selective activity against oral microorganisms. J Clin Microbiol 2004; 42: 1024–1029. 24 Yang D, Biragyn A, Kwak LW, Oppenheim JJ. Mammalian defensins in immunity: more than just microbicidal. Trends Immunol 2002; 23: 291–296. 25 Chen X, Niyonsaba F, Ushio H, Hara M, Yokoi H, Matsumoto K et al. Antimicrobial peptides human beta-defensin (hBD)-3 and hBD-4 activate mast cells and increase skin vascular permeability. Eur J Immunol 2007; 37: 434–444. 26 Takahashi Y, Davey M, Yumoto H, Gibson III FC, Genco CA. Fimbria-dependent activation of pro-inflammatory molecules in Porphyromonas gingivalis infected human aortic endothelial cells. Cell Microbiol 2006; 8: 738–757. 27 Zhang P, Martin M, Michalek SM, Katz J. Role of mitogen-activated protein kinases and NF-kappaB in the regulation of pro-inflammatory and anti-inflammatory cytokines by Porphyromonas gingivalis hemagglutinin B. Infect Immun 2005; 73: 3990–3998. 28 Kinane DF, Demuth DR, Gorr SU, Hajishengallis GN, Martin MH. Human variability in innate immunity. Periodontol 2000 2007; 45: 14–34. 29 Song H, Belanger M, Whitlock J, Kozarov E, Progulske-Fox A. Hemagglutinin B is involved in the adherence of Porphyromonas gingivalis to human coronary artery endothelial cells. Infect Immun 2005; 73: 7267–7273. 30 Progulske-Fox A, Tumwasorn S, Holt SC. The expression and function of a Bacteroides gingivalis hemagglutinin gene in Escherichia coli. Oral Microbiol Immunol 1989; 4: 121–131. 31 Progulske-Fox A, Tumwasorn S, Lepine G, Whitlock J, Savett D, Ferretti JJ et al. The cloning, expression and sequence analysis of a second Porphyromonas gingivalis gene that codes for a protein involved in hemagglutination. Oral Microbiol Immunol 1995; 10: 311–318. 32 Lepine G, Ellen RP, Progulske-Fox A. Construction and preliminary characterization of three hemagglutinin mutants of Porphyromonas gingivalis. Infect Immun 1996; 64: 1467–1472. 33 Han N, Lepine G, Whitlock J, Wojciechowski L, Progulske-Fox A. The Porphyromonas gingivalis prtP/kgp homologue exists as two open reading frames in strain 381. Oral Dis 1998; 4: 170–179.

34 Veith PD, Talbo GH, Slakeski N, Dashper SG, Moore C, Paolini RA et al. Major outer membrane proteins and proteolytic processing of RgpA and Kgp of Porphyromonas gingivalis W50. Biochem J 2002; 363: 105–115. 35 Katz J, Black KP, Michalek SM. Host responses to recombinant hemagglutinin B of Porphyromonas gingivalis in an experimental rat model. Infect Immun 1999; 67: 4352–4359. 36 Zhang P, Martin M, Yang QB, Michalek SM, Katz J. Role of B7 costimulatory molecules in immune responses and T-helper cell differentiation in response to recombinant HagB from Porphyromonas gingivalis. Infect Immun 2004; 72: 637–644. 37 Watanabe K, Yilmaz O, Nakhjiri SF, Belton CM, Lamont RJ. Association of mitogenactivated protein kinase pathways with gingival epithelial cell responses to Porphyromonas gingivalis infection. Infect Immun 2001; 69: 6731–6737. 38 Lu Q, Samaranayake LP, Darveau RP, Jin L. Expression of human beta-defensin-3 in gingival epithelia. J Periodontal Res 2005; 40: 474–481. 39 Saitoh M, Abiko Y, Shimabukuro S, Kusano K, Nishimura M, Arakawa T et al. Correlated expression of human beta defensin-1, -2 and -3 mRNAs in gingival tissues of young children. Arch Oral Biol 2004; 49: 799–803. 40 Bissell J, Joly S, Johnson GK, Organ CC, Dawson D, BMcCray PJ et al. Expression of beta-defensins in gingival health and in periodontal disease. J Oral Pathol Med 2004; 33: 278–285. 41 Ghosh SK, Gerken TA, Schneider KM, Feng Z, McCormick TS, Weinberg A. Quantification of human beta-defensin-2 and -3 in body fluids: application for studies of innate immunity. Clin Chem 2007; 53: 757–765. 42 Tao R, Jurevic RJ, Coulton KK, Tsutsui MT, Roberts MC, Kimball JR et al. Salivary antimicrobial peptide expression and dental caries experience in children. Antimicrob Agents Chemother 2005; 49: 3883–3888. 43 Dale BA, Tao R, Kimball JR, Jurevic RJ. Oral antimicrobial peptides and biological control of caries. BMC Oral Health 2006; 6 (Suppl 1): S13. 44 Okuda K, Yamamoto A, Naito Y, Takazoe I, Slots J, Genco RJ. Purification and properties of hemagglutinin from culture supernatant of Bacteroides gingivalis. Infect Immun 1986; 54: 659–665. 45 Johnson GL, Lapadat R. Mitogen-activated protein kinase pathways mediated by ERK, JNK, and p38 protein kinases. Science 2002; 298: 1911–1912. 46 Wehkamp J, Salzman NH, Porter E, Nuding S, Weichenthal M, Petras RE et al. Reduced Paneth cell alpha-defensins in ileal Crohn’s disease. Proc Natl Acad Sci USA 2005; 102: 18129–18134. 47 Aldred PM, Hollox EJ, Armour JA. Copy number polymorphism and expression level variation of the human alpha-defensin genes DEFA1 and DEFA3. Hum Mol Genet 2005; 14: 2045–2052. 48 Wehkamp J, Schmid M, Fellermann K, Stange EF. Defensin deficiency, intestinal microbes, and the clinical phenotypes of Crohn’s disease. J Leukoc Biol 2005; 77: 460–465. 49 Dork T, Stuhrmann M. Polymorphisms of the human beta-defensin-1 gene. 1998; 12: 171–173. 50 Linzmeier RM, Ganz T. Human defensin gene copy number polymorphisms: comprehensive analysis of independent variation in alpha- and beta-defensin regions at 8p22p23. Genomics 2005; 86: 423–430. 51 Fellermann K, Wehkamp J, Herrlinger KR, Stange EF. Crohn’s disease: a defensin deficiency syndrome? Eur J Gastroenterol Hepatol 2003; 15: 627–634. 52 Boniotto M, Hazbon MH, Jordan WJ, Lennon GP, Eskdale J, Alland D et al. Novel hairpin-shaped primer assay to study the association of the -44 single-nucleotide polymorphism of the DEFB1 gene with early-onset periodontal disease. Clin Diagn Lab Immunol 2004; 11: 766–769. 53 Vankeerberghen A, Scudiero O, De Boeck K, Macek Jr M, Pignatti PF, Van Hul N et al. Distribution of human beta-defensin polymorphisms in various control and cystic fibrosis populations. Genomics 2005; 85: 574–581. 54 Kohler JJ, Pathangey L, Hasona A, Progulske-Fox A, Brown TA. Long-term immunological memory induced by recombinant oral Salmonella vaccine vectors. Infect Immun 2000; 68: 4370–4373. 55 Conover W. Practical Nonparametric Statistics, Edn, 3rd edn. Wiley: New York, 1999.

Supplementary Information accompanies the paper on Immunology and Cell Biology website (http://www.nature.com/icb)

Immunology and Cell Biology