Protease-Activated Receptor-2 Neutrophil Proteinase 3 Through

Activation of Human Oral Epithelial Cells by Neutrophil Proteinase 3 Through Protease-Activated Receptor-2

This information is current as Akiko Uehara, Shunji Sugawara, Koji Muramoto and of October 2, 2021. Haruhiko Takada J Immunol 2002; 169:4594-4603; ; doi: 10.4049/jimmunol.169.8.4594 http://www.jimmunol.org/content/169/8/4594 Downloaded from

References This article cites 52 articles, 25 of which you can access for free at: http://www.jimmunol.org/content/169/8/4594.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

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

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average by guest on October 2, 2021

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts Errata An erratum has been published regarding this article. Please see next page or: /content/184/7/4042.full.pdf

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2002 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Activation of Human Oral Epithelial Cells by Neutrophil Proteinase 3 Through Protease-Activated Receptor-21

Akiko Uehara,* Shunji Sugawara,2* Koji Muramoto,† and Haruhiko Takada*

Proteinase 3 (PR3), a 29-kDa serine proteinase secreted from activated neutrophils, also exists in a membrane-bound form, and is suggested to actively contribute to inflammatory processes. The present study focused on the mechanism by which PR3 activates human oral epithelial cells. PR3 activated the epithelial cells in culture to produce IL-8 and monocyte chemoattractant protein-1 and to express ICAM-1 in a dose- and time-dependent manner. Incubation of the epithelial cells for 24 h with PR3 resulted in a significant increase in the adhesion to neutrophils, which was reduced to baseline levels in the presence of anti-ICAM-1 mAb. Activation of the epithelial cells by PR3 was inhibited by serine proteinase inhibitors and serum. The epithelial cells strongly express protease-activated receptor (PAR)-1 and PAR-2 mRNA and weakly express PAR-3 mRNA. The expression of PAR-2 on the cell surface was promoted by PR3, and inhibited by cytochalasin B, but not by cycloheximide. PR3 cleaved the peptide corresponding to the N terminus of PAR-2 with exposure of its tethered ligand. Treatment with trypsin, an agonist for PAR-2, and Downloaded from -a synthetic PAR-2 agonist peptide induced intracellular Ca2؉ mobilization, and rendered cells refractory to subsequent stimu lation with PR3 and vice versa. The production of cytokine induced by PR3 and the PAR-2 agonist peptide was completely abolished by a phospholipase C inhibitor. These findings suggest that neutrophil PR3 activates oral epithelial cells through G protein-coupled PAR-2 and actively participates in the process of inflammation such as periodontitis. The Journal of Immunol- ogy, 2002, 169: 4594–4603. http://www.jimmunol.org/

he four members of the protease-activated receptor is considered that they are involved in several pathophysiological pro- (PAR)3 family are G protein-coupled receptors character- cesses, including growth, development, and inﬂammation. T ized by a proteolytic cleavage of the N terminus that ex- PR3 is a 29-kDa serine proteinase with homology to HLE and poses tethered ligands and autoactivates the receptor function (1– Cat G, all of which are stored in azurophil granules of neutrophils 3). Three of them (PAR-1, PAR-3, and PAR-4) are activated (11–13). PR3 also presents on the cell surface and within secretory mainly by thrombin; the fourth (PAR-2) is activated by other and speciﬁc granules of neutrophils, and exposure of neutrophils to serine proteinases such as trypsin, tryptase, and factor Xa (1–3). It cytokines or chemoattractants induces an increase in cell surface-

is reported that more than one family member can be present in the bound PR3 (14). PR3 is also expressed by monocytes, basophils, by guest on October 2, 2021 same cell: human platelets express PAR-1 and PAR-4 (4, 5); and mast cells (15). PR3 is a major target Ag of anti-neutrophil mouse platelets express PAR-3 and PAR-4 (6); human endothelial cytoplasmic Abs in Wegener’s granulomatosis, a debilitating au- cells express PAR-1, PAR-2, and possibly PAR-3 (7, 8); and hu- toimmune disease characterized by necrotizing vasculitis (15, 16). man oral epithelial cells express PAR-1, PAR-2, and PAR-3 (9). It has recently been shown that PR3 exhibits many biological func- PAR-1 is the most widely expressed receptor among PAR family tions, including the degradation of extracellular matrix proteins members in humans and mice, and it is reported that PAR-1 on (13), regulation of myeloid differentiation (12, 17), enhancement human platelets and endothelial cells is inactivated by neutrophil of TNF-␣ and IL-1␤ release from human monocytic cell lines (18), serine proteinases, human leukocyte elastase (HLE), cathepsin G production of IL-8 and monocyte chemoattractant protein-1 (Cat G), and proteinase 3 (PR3), by cleavage downstream of the teth- (MCP-1) by human endothelial cells (19, 20), and antibacterial ered ligand (10). Because PARs are expressed in a variety of cells, it action (21), all of which indicate that cell-bound and secreted soluble PR3 actively contribute to inflammatory processes, although the underlying mechanism is unclear to date. *Department of Microbiology and Immunology, School of Dentistry, and †Laboratory Oral epithelial cells are the first cells encountered by bacteria in of Biomolecular Function, Graduate School of Life Sciences, Tohoku University, the periodontal tissues. In addition to acting as a physical barrier Sendai, Japan against the invasion of pathogenic organisms, oral (gingival) epi- Received for publication April 29, 2002. Accepted for publication August 15, 2002. thelial cells in inflamed regions appear to express several proin- The costs of publication of this article were defrayed in part by the payment of page flammatory cytokines such as IL-1␤, IL-6, IL-8, TNF-␣, and charges. This article must therefore be hereby marked advertisement in accordance ␤ with 18 U.S.C. Section 1734 solely to indicate this fact. TGF- 1 (22), implying that the cells actively participate in the 1 This work was supported in part by Grants-in-Aid for Scientific Research from the initiation and development of chronic oral inflammation such as Japan Society for the Promotion of Science (12470380 and 13671894). periodontitis. However, only a few studies have demonstrated cy- 2 Address correspondence and reprint requests to Dr. Shunji Sugawara, Department of tokine production by human gingival epithelial cells and related Microbiology and Immunology, Tohoku University School of Dentistry, 4-1 Seiryo- cell lines derived from the oral cavity in response to oral bacteria machi, Aoba-ku, Sendai 980-8575, Japan. E-mail address: [email protected]. tohoku.ac.jp (23–25). In contrast to human colonic epithelial cells, human oral 3 ␣ ␣ epithelial cells are shown to be basically unresponsive to many Abbreviations used in this paper: PAR, protease-activated receptor; 1-AT, 1- antitrypsin; Boc-Ala-ONp, Boc-Ala-p-nitrophenyl ester; Cat G, cathepsin G; CDS, bacterial cell surface components, even in the presence of the sol- cell dissociation solution; EM, extracellular medium; GCF, gingival crevicular fluid; HLE, human leukocyte elastase; MCP-1, monocyte chemoattractant protein-1; PLC, uble form of CD14, a bacterial pattern recognition receptor (26). phospholipase C; PR3, proteinase 3. We have recently shown that oral epithelial cells constitutively

␣ ␣ express a 24-kDa precursor form of IL-18, and that human neu- teinase inhibitors, Pefabloc SC and 1-antitrypsin ( 1-AT), and FCS for 30 trophil PR3 induced the secretion of a bioactive IL-18 from the min at 37°C before use. Cultivation was conducted in triplicate, and levels epithelial cells in combination with LPS after priming with IFN-␥ of IL-8 and MCP-1 in the supernatants were measured using OptEIA ELISA kits (BD PharMingen, San Diego, CA). The concentrations of the (27). We extended the investigation of the underlying mechanism cytokines in the supernatants were determined using the Softmax data anal- by which PR3 activates the epithelial cells, and in the present ysis program (Molecular Devices, Menlo Park, CA). study, we obtained the evidence for the first time that PR3 by itself could activate the epithelial cells in culture to induce production of Flow cytometry IL-8 and MCP-1 and expression of ICAM-1 via the PAR-2 Flow cytometric analyses were performed using a FACScan cytometer pathway. (BD Biosciences, Mountain View, CA), as described (32). Oral epithelial cells were pretreated with or without cytochalasin B (30 nM) for 30 min or 1 ␮g/ml cycloheximide for6hat37°C. Then cells were stimulated with or Materials and Methods without PR3 (10 ␮g/ml) for up to 24 h in the presence or absence of Reagents cytochalasin B (30 nM) or cycloheximide (1 ␮g/ml) at 37°C. After the incubation, cells were collected by nonenzymatic CDS and washed in PBS. Purified human neutrophil PR3 was obtained from HyTest (Turku, Finland) Cells were stained with anti-CD54, anti-PAR-1, anti-PAR-2, or control IgG and from Elastin Products (Owensville, MO). Purified HLE, Cat G, phos- at 4°C for 30 min, followed by FITC-conjugated goat anti-mouse IgG pholipase C (PLC) inhibitor U73122, and control compound U73343 were (BioSource International, Camarillo, CA) at 4°C for a further 30 min. For obtained from Calbiochem-Novabiochem (La Jolla, CA). The purity of Ͼ PAR-3 staining, cells were incubated with rabbit anti-PAR-3 polyclonal Ab three enzymes (PR3, HLE, and Cat G) was 95% by SDS-PAGE, accord- or control IgG for 30 min, followed by FITC-conjugated swine anti-rabbit ing to the manufacturer, and further confirmed by Western blotting using IgG (DAKO, Kyoto, Japan) at 4°C for a further 30 min. To calculate the anti-human PR3 (Elastin Products), HLE, and Cat G (Santa Cruz Biotech- percentage of positive cells, the baseline cursor was set at a channel that nology, Santa Cruz, CA) polyclonal Abs. The result showed that no cross- yielded less than 2% of events positive with the isotype Ab control. Flu- Downloaded from reactivity was observed in each preparation (data not shown). Human ␣ ␣ orescence to the right was counted as specific binding. The arithmetic mean rIL-1 and rTNF- were supplied by Dainippon Pharmaceutical (Osaka, was used in the computation of the mean fluorescence intensity. Japan). Boc-Ala-p-nitrophenyl ester (Boc-Ala-ONp) was obtained from Bachem (Bubendorf, Germany). A low toxic serine proteinase inhibitor, Adhesion assay Pefabloc SC, was obtained from Roche Diagnostics (Mannheim, Ger- many). Cycloheximide was obtained from Biomol (Plymouth Meeting, HSC-2 cells (104 cells/well) were cultured in 96-well plates for 24 h until PA). Anti-CD54 (ICAM-1) (mouse IgG1) was purchased from Immuno- confluent. Thereafter, cells were washed twice with PBS and cultured in tech (Marseille, France). Neutralizing anti-CD54 84H10 (mouse IgG1) was RPMI 1640 medium alone or medium containing either PR3 (10 ␮g/ml) or http://www.jimmunol.org/ obtained from Beckman Coulter (Miami, FL) and dialyzed against PBS. TNF-␣ (5 ng/ml). Cells were cultured for 24 h and washed with warmed Anti-human PAR-1 mAb ATAP2 (mouse IgG) raised against aa 42–55 of PBS. Neutrophils (3.2 ϫ 105 cells/ml each) were then added to HSC-2 cells human PAR-1, anti-human PAR-2 mAb SAM11 (mouse IgG2a) raised in RPMI 1640 medium and allowed to adhere during 30 min at 37°C under against aa 37–50 of human PAR-2, and rabbit anti-human PAR-3 poly- static conditions. The wells were then washed three times with warmed clonal Ab raised against aa 1–103 of human PAR-3 were obtained from PBS to remove nonadherent cells. Adhesion of neutrophils was quantified Santa Cruz Biotechnology. Nonenzymatic cell dissociation solution (CDS) using a modified myeloperoxidase assay, as described (33). Briefly, HSC-2 ϩ was obtained from Sigma-Aldrich (St. Louis, MO). Cell-permeant fura cells plus adherent neutrophils were washed twice with PBS without Ca2 ϩ 2-AM was obtained from Molecular Probes (Eugene, OR). PAR-1 and and Mg2 (pH 6.0), and subsequently HSC-2 cells plus adhering cells were PAR-2 agonist peptides (SFLLRN and SLIGKV, respectively) were syn- permeabilized in 50 ␮l PBS containing 0.5% hexadecyltrimethyl ammo- thesized by Takara (Otsu, Japan). All other reagents were obtained from nium bromide for 30 min at room temperature. Next, 250 ␮l warmed O- by guest on October 2, 2021 Sigma-Aldrich, unless otherwise indicated. dianisidine dihydrochloride (0.2 mg/ml in PBS, pH 6.0) containing H2O2 (0.4 mM) was added. After 15 min of incubation at 37°C, OD was read at Cells and cell cultures 450 nm. Serial dilutions of neutrophils were used as a standard to calculate the number of adherent neutrophils. For the inhibition assay, HSC-2 cells The human oral epithelial cell lines HSC-2 (28) and KB (29), established were incubated with anti-CD54 mAb 84H10 or isotype control mAb (10 from squamous cell carcinoma, were obtained from the Cancer Cell Re- ␮g/ml each) for 120 min at 37°C before the addition of neutrophils. Next, pository, Institute of Development, Aging and Cancer, Tohoku University the adhesion assay was performed as described. The mAb remained present (HSC-2; Sendai, Japan) and from the Health Science Research Resources during the adhesion assay. Bank (KB; Tokyo, Japan). HSC-2 was grown in RPMI 1640 with 10% heat-inactivated FCS (Life Technologies, Grand Island, NY), and KB was Measurement of enzymatic activity and inhibition assay grown in ␣-MEM with 10% FCS with a medium change every 3 days. Human gingival epithelial cells were prepared from explants of normal The amidolytic activity of PR3 was assayed with 0.625 mM Boc-Ala-ONp human gingival tissues with informed consent, as described previously for 10–30 min at 25°C in 0.1 M HEPES buffer containing 0.1 M NaCl, 10 (27). The experimental procedure was approved by the Ethical Review mM CaCl2, 0.005% Triton X-100, and 5% DMSO, pH 7.5, as described Board, Tohoku University School of Dentistry. previously (18, 34). The formation of the p-nitrophenol product was mon- Neutrophils from heparinized (10 U/ml) peripheral venous blood were itored at 405 nm using the Softmax data analysis program (Molecular De- isolated by density-gradient centrifugation on Mono-Poly resolving me- vices). One unit of enzymatic activity was defined as the liberation of 1 dium (ICN Biomedical, Costa Mesa, CA) at 300 ϫ g for 30 min at room ␮mol p-nitrophenol from the substrate/min at 25°C. To inhibit the enzy- temperature (30). The fraction containing neutrophils was harvested and matic activity of PR3, the enzyme was preincubated with different con- washed three times with PBS at 4°C and suspended in RPMI 1640 me- centrations of inhibitors for 30 min before use. dium. The viability of these cells was greater than 98%, as judged by trypan blue dye exclusion. The purity of neutrophils was above 95% RT-PCR assay morphologically. Total cellular RNA was extracted from the cells using Isogen (Nippon PBMCs from heparinized (10 U/ml) peripheral venous blood were iso- Gene, Tokyo, Japan), according to the manufacturer’s instructions. Ran- lated by Lympholyte-H (Cedarlane Laboratories, Hornby, Ontario, Can- dom hexamer-primed reverse transcription was performed on 2.5 ␮l total ada) gradient centrifugation at 800 ϫ g for 20 min at room temperature RNA in a 50-␮l reaction volume, and all PCR procedures were performed (31). The isolated PBMCs were washed three times with PBS at 4°C. The in a 20-␮l vol, as described previously (27). The primers used for PCR viability of these cells was greater than 98%, as judged by trypan blue dye were as follows: IL-8, 5Ј-GATTGAGAGTGGACCACACT-3Ј,5Ј-TCTC exclusion. CCGTGCAATATCTAGG-3Ј; MCP-1, 5Ј-AACTGAAGCTCGCACTCT Measurement of cytokines CG-3Ј,5Ј-TCAGCACAGATCTCCTTGGC-3Ј; ICAM-1, 5Ј-CAGTCAC CTATGGCAACGAC-3Ј,5Ј-ATTCAGCGTCACCTTGGCTC-3Ј; PAR-1, Confluent oral epithelial cells were collected by CDS and washed three 5Ј-TGTGAACTGATCATGTTTATG-3Ј,5Ј-TTCGTAAGATAAGAGAT times in PBS. The cells (104 cells/200 ␮l) were seeded in culture medium ATGT-3Ј (35); PAR-2, 5Ј-GCAGCCTCTCTCTCCTGCAGTGG-3Ј,5Ј- in 96-well plates (Falcon; BD Labware, Lincoln Park, NJ). After incuba- CTTGCATCTGCTTTACAGTGCG-3Ј (36); PAR-3, 5Ј-ATAACGTTTA Ј Ј Ј tion for 1 day at 37°Cina5%CO2 incubator, the cells were stimulated AGAGACGGGACT-3 ,5-TAGCAGTAGATGATAAGCACA-3 (7); with test materials in 200 ␮l medium without serum for a given period. To PAR-4, 5Ј-GACGAGAGCGGGAGCACC-3Ј,5Ј-CCCGTAGCACAGCA inhibit the enzymatic activity of PR3, it was preincubated with serine pro- GCATGG-3Ј (4); and GAPDH, 5Ј-CTACAATGAGCTGCGTGTGG-3Ј, 4596 ACTIVATION OF ORAL EPITHELIAL CELLS BY PR3 VIA PAR-2

5Ј-AAGGAAGGCTGGAAGAGTGC-3Ј (27). The primers for IL-8, MCP-1, ICAM-1, PAR-1, PAR-2, PAR-3, PAR-4, and GAPDH were con- structed to generate fragments of 422, 257, 243, 708, 1,066, 858, 725, and 527 bp, respectively. Cycling conditions were as follows: IL-8, 25 cycles at 94°C for 1 min, 63°C for 1 min, and 72°C for 3 min; MCP-1, ICAM-1, and GAPDH, 35 cycles at 94°C for 1.5 min, 60°C for 1 min, and 72°C for 3 min; PAR-1, 34 cycles at 95°C for 1 min, 55°C for 1 min, and 72°C for 1 min; and PAR-2, PAR-3, and PAR-4, 36 cycles at 95°C for 1 min, 55°C for 1 min, and 72°C for 3 min. Amplified samples were visualized on 2.0% agarose gels stained with ethidium bromide and photographed under UV light. Analysis of peptide cleavage A peptide corresponding to a region spanning the cleavage site of the PAR-2, residues 32–45 (32SSKGRSLIGKVDGT45) (37), was synthesized by Takara. The peptide (200 ␮M) was incubated with proteinases for 30 min at 37°C in PBS. Each digest was separated by reversed-phase HPLC on a Wakosil 5C4-200 column (5 mm, 4.6 ϫ 250 mm) (WAKO, Osaka, Japan) using a linear gradient from 0 to 30% acetonitrile in 0.1% triflu- oroacetic acid, and the amino acid sequences of peptide fragments were analyzed by a gas phase protein sequencer (PSQ-1; Shimadzu, Kyoto, Ja- pan) and a matrix-assisted laser desorption ionization time of flight mass

spectrometry (Kompact MALDI I; Shimadzu), according to a previously Downloaded from described procedure (38). Calcium mobilization Conﬂuent oral epithelial cells were collected by nonenzymatic CDS, washed twice with PBS, and suspended at 2 ϫ 106 cells/ml in an extra-

cellular medium (EM; 121 mM NaCl, 5.4 mM KCl, 0.8 mM MgCl2, 1.8

mM CaCl2, 6 mM NaHCO3, 5.5 mM glucose, 25 mM HEPES, 0.1% w/v http://www.jimmunol.org/ BSA, pH 7.3). Cells were loaded with 1 ␮M fura 2-AM with shaking for 30 min at room temperature. After being washed with EM, cells were resuspended in EM and incubated for 30 min at room temperature. After another wash with EM, loaded cells suspended at 2 ϫ 106 cells/ml in EM without BSA were transferred to stirred quartz cuvettes in a CAF-100 spectrofluorometer (Jasco, Tokyo, Japan) at 37°C. Fura 2 fluorescence was measured at 340 and 380 nm excitation and 510 nm emission, and the ratio of the fluorescence at the two excitation wavelengths, which is proportional 2ϩ to [Ca ]i, was calculated.

Coculture of oral epithelial cells with stimulated neutrophils by guest on October 2, 2021 Confluent HSC-2 cells in 96-well plates were cocultured with the indicated number of purified neutrophils in 200 ␮l serum-free medium in the presence or absence of various concentrations of FMLP for 24 h. The cocul- tures were also performed in the presence or absence of Pefabloc SC, ␣ 1-AT, and FCS for 24 h. After the incubation, levels of IL-8 and MCP-1 in the supernatants were measured by ELISA. FIGURE 1. Effect of PR3 on the production of IL-8 and MCP-1 by human oral epithelial cells. Confluent HSC-2 and KB were incubated for Data analysis 24 h in the presence of various concentrations of PR3 (A), or incubated in All experiments in this study were performed at least three times to confirm the presence (ϩ) or absence (Ϫ) of PR3 (10 ␮g/ml) for the time indicated the reproducibility of the results. In most experiments, values are repre- (B). C, Confluent HSC-2 cells were incubated in the presence of various sented as means Ϯ SD of triplicate assays. The statistical significance of concentrations of PR3, HLE, and Cat G for 24 h. D, Confluent HSC-2 cells differences between the two means was evaluated by one-way ANOVA, were incubated in the presence of various concentrations of PR3 from using the Bonferroni or Dunn method, and p values less than 0.05 were HyTest (PR3 (H)) or PR3 from Elastin Products (PR3 (E)) for 24 h. Con- considered significant. centrations of IL-8 and MCP-1 in the culture supernatants were determined by ELISA. E, Confluent HSC-2 and KB were incubated for 8 h with me- Results dium alone (lane 2), 10 ␮g/ml PR3 (lane 3), or 1000 IU/ml IFN-␥ for IL-8 Effect of PR3 on the production of IL-8 and MCP-1 by oral or 10 ng/ml IL-1␣ for MCP-1 (lane 4), and the expression of IL-8, MCP-1, epithelial cells and GAPDH mRNA was analyzed by RT-PCR. Water control was loaded We first examined the effect of PR3 on the production of IL-8 and in lane 1. F, Confluent primary oral epithelial cells were incubated for 24 h MCP-1 by oral epithelial cell lines, HSC-2 and KB. Incubation of in the presence of various concentrations of PR3. Concentrations of IL-8 p Ͻ ,ء .and MCP-1 in the culture supernatants were determined by ELISA both cell lines in the presence of various concentrations of PR3 for pϽ 0.01 compared with the respective control (medium ,ءء and ,0.05 24 h resulted in a dose-dependent increase in IL-8 and MCP-1 alone). The results presented were representative of three different exper- ␮ (Fig. 1A). PR3 at 5–10 g/ml was most effective in stimulating the iments demonstrating similar results. production. Incubation of HSC-2 cells with 10 ␮g/ml (345 nM) PR3 resulted in a time-dependent increase in the production of IL-8 and MCP-1 (Fig. 1B). A significant increase in production was observed from 8-h incubation, and a marked increase at 24 and different sources showed the same results (Fig. 1D). IL-8 mRNA 48 h. In contrast to PR3, two other neutrophil serine proteinases, was already expressed in untreated cells, but MCP-1 mRNA was HLE and Cat G, showed only marginal activity in oral epithelial not expressed in untreated cells, and the expression of IL-8 and cells as assessed by MCP-1 production (Fig. 1C), which was con- MCP-1 mRNA was increased and induced by PR3, respectively sistent with our recent observation (27). The PR3 from the two (Fig. 1E). IFN-␥ (1000 IU/ml) and IL-1␣ (10 ng/ml) were used as The Journal of Immunology 4597 a positive control for IL-8 and MCP-1 mRNA expression, respec- trophils was observed as compared with the cells incubated in the tively. Oral epithelial cells in primary culture also produce IL-8 medium alone (Fig. 2E). TNF-␣ (5 ng/ml), as a positive control, and MCP-1 in response to PR3, although the cytokine levels were also significantly promoted the adhesion to neutrophils and to lower than those in the epithelial cell lines (Fig. 1F). HSC-2 cells. Furthermore, pretreatment of HSC-2 cells with anti- ICAM-1 mAb resulted in a complete reduction of adhesion of PR3-induced ICAM-1 mediates adhesion of oral epithelial cells neutrophils, indicating that the ICAM-1 expressed on oral epithe- to neutrophils lial cells by PR3 was a major factor in the adhesion to neutrophils. We next examined the effect of PR3 on the expression of an ad- KB showed the same results (data not shown). The findings shown hesion molecule, ICAM-1, on HSC-2 cells by flow cytometry. As- in Figs. 1 and 2 indicate that PR3 activates oral epithelial cells, sessment of the surface expression of ICAM-1 on the untreated consequently inducing production of IL-8 and MCP-1, and expres- cells revealed no reactivity (less than 2%), but the cells treated sion of ICAM-1. with increasing concentrations of PR3 for 24 h showed a dose- dependent increase of ICAM-1 expression, with a plateau in the Requirement of enzymatic activity for PR3 to induce IL-8 level of ICAM-1 expression reached at 1 ␮g/ml PR3 (Fig. 2A). and MCP-1 production and ICAM-1 expression in oral TNF-␣ was used as a positive control and also enhanced ICAM-1 epithelial cells expression. Incubation of HSC-2 cells with 10 ␮g/ml PR3 resulted We then examined whether the PR3-induced activation of oral in a time-dependent increase in the expression of ICAM-1 (Fig. epithelial cells was due to the enzymatic activity of PR3. It has 2B). The expression was up-regulated from 8-h incubation, then been reported that the enzymatic activity of PR3 was inhibited by

increased linearly until 48 h, at which point almost all the cells sulfonyl fluoride-type serine proteinase inhibitors such as PMSF Downloaded from expressed ICAM-1. A representative FACS profile of ICAM-1 ex- and diisopropyl fluorophosphate, and by a naturally occurring ␣ pression on HSC-2 cells after given doses of PR3 treatment for serine proteinase inhibitor, 1-AT (19, 21). Therefore, we first ex- 24 h is shown in Fig. 2C. ICAM-1 mRNA was not expressed in amined the inhibitory effects of a low toxic sulfonyl fluoride-type ␣ ␣ untreated cells and induced by PR3 (Fig. 2D). TNF- (5 ng/ml) serine proteinase inhibitor, Pefabloc SC, and 1-AT on the enzy- was used as a positive control. We then examined whether the matic activity of PR3 using Boc-Ala-ONp as a substrate. PR3 ICAM-1 induced by PR3 in HSC-2 cells is involved in the adhe- showed substantial enzymatic activity (9.7 U/mg protein), and http://www.jimmunol.org/ sion to neutrophils. After the incubation of HSC-2 cells with 10 these inhibitors almost completely inhibited it (Fig. 3A). Con- ␮ ␣ g/ml PR3 for 24 h, a 9.1-fold increase in the adhesion of neu- current with this, Pefabloc SC and 1-AT significantly inhibited by guest on October 2, 2021

FIGURE 2. PR3-induced ICAM-1 expression on oral epithelial cells mediates adhesion of neutrophils to oral epithelial cells. Confluent HSC-2 cells were incubated for 24 h in the presence of various concentrations of PR3 or TNF-␣ (A), or incubated in the presence or absence of PR3 (10 ␮g/ml) for the time indicated (B). The surface expression of ICAM-1 was assessed by flow cytometry. Data were expressed as percentage of ICAM-1-positive cells. C,A representative FACS profile of ICAM-1 expression on HSC-2 cells after treatment with the given doses of PR3 for 24 h is shown. D, Confluent HSC-2 cells were incubated for 8 h with medium alone (lane 2), 10 ␮g/ml PR3 (lane 3), or 5 ng/ml TNF-␣ (lane 4), and the expression of ICAM-1 and GAPDH mRNA was analyzed by RT-PCR. Water control was loaded in lane 1. E, Confluent HSC-2 cells were stimulated with medium alone or medium containing PR3 (10 ␮g/ml) or TNF-␣ (5 ng/ml) as a positive control for 24 h. HSC-2 cells washed with PBS were incubated with or without anti-CD54 mAb 84H10 or isotype control IgG (10 ␮g/ml each) for 30 min, and neutrophils (3.2 ϫ 105 cells/ml each) were added. After 30 min, nonadherent cells were removed, p Ͻ0.01 compared with the respective control (medium alone). The ,ءء .and the number of adherent cells was quantified using a myeloperoxidase assay results presented were representative of three different experiments demonstrating similar results. 4598 ACTIVATION OF ORAL EPITHELIAL CELLS BY PR3 VIA PAR-2

FIGURE 3. Effect of serine proteinase inhibitors and FCS on the activation of oral epithelial cells by PR3. A, ␣ PR3 was pretreated with Pefabloc SC, 1-AT, or FCS at the dose indicated for 30 min before use. Enzymatic activity was measured using Boc-Ala-ONp as a substrate, and the results are expressed as percentage of activity. The activity of PR3 was 9.7 U/mg protein. B–E, HSC-2 cells were stimulated with or without PR3 (10 ␮g/ml) (B–D) or rIL-1␣ (10 ng/ml) (E) for 24 h at 37°C. PR3 (10 ␮g/ml) and IL-1␣ (10 ng/ml) were pre- ␮ ␣ treated with or without Pefabloc SC (40 M), 1-AT (1 ␮g/ml), or FCS (1 and 10%) for 30 min at 37°C before use. The supernatants were collected, and the concen- Downloaded from trations of IL-8 (B) and MCP-1 (C and E) were measured by ELISA. The cells were collected by CDS, and the surface expression of ICAM-1 was evaluated by ﬂow cytometry (D). The percentage of cytokine production and ICAM-1 expression was calculated on the basis of the value obtained without inhibitors (B, 2.9 Ϯ 0.2 ng/ml; C, 14.0 Ϯ 1.1 ng/ml; D, mean ﬂuorescence http://www.jimmunol.org/ intensity ϭ 465; E, 14.2 Ϯ 0.9 ng/ml). Error bars in- p Ͻ 0.01 compared with PR3 alone. The ,ءء .dicate SD results presented were representative of three different experiments demonstrating similar results. by guest on October 2, 2021

PR3-induced IL-8 and MCP-1 production, and ICAM-1 expression by cytochalasin B, an inhibitor of actin polymerization (39), but not by HSC-2 cells (Fig. 3, B–D). In addition, FCS almost completely by cycloheximide, a protein synthesis inhibitor. The up-regulation by inhibited the enzymatic activity of PR3 (Fig. 3A), PR3-induced PR3 was detected at 30-min incubation, and reached a plateau at 1 h IL-8, and MCP-1 production (Fig. 3, B and C), and ICAM-1 ex- (Fig. 4C). Exposure of the cells to PR3 for up to 24 h did not further pression (Fig. 3D), probably due to naturally occurring proteinase up-regulate the expression of PAR-2, and did not induce expression of inhibitors in the serum. By contrast, the serine proteinase inhibitors PAR-1 and -3 (data not shown). Furthermore, trypsin, an agonist for and serum did not inhibit IL-1␣-induced production of MCP-1 PAR-2 (1–3), and a PAR-2 agonist peptide SLIGKV also up-regu- (Fig. 3E). These results indicate that the enzymatic activity of PR3 lated PAR-2 expression (Fig. 4D), but did not up-regulate PAR-1 and is critical to the activation of oral epithelial cells. -3 (data not shown) on the epithelial cell surface. These results indicate that PAR-2 is an inducible receptor from internal storage by PR3 Analysis of the PAR family in oral epithelial cells and induction and known PAR-2 agonists. of PAR-2 expression on the cell surface by PR3 Because it has been shown that PARs on human platelets and Cleavage of human PAR-2 peptide with exposure of its tethered endothelial cells are activated or inactivated by various proteinases ligand by PR3 (1–3), the expression of the PAR family in oral epithelial cells was The above observation suggests that PR3 cleaves PAR-2 at the analyzed. PBMCs were used as a positive control (5). Oral epi- specific site and exposes its tethered ligand. To examine this pos- thelial cells in culture strongly expressed PAR-1 and PAR-2 sibility, a peptide corresponding to region surrounding the cleav- mRNA, weakly expressed PAR-3 mRNA, and did not express age site of the human PAR-2 (Fig. 5A) was incubated with PR3, PAR-4 mRNA, as assessed by RT-PCR analysis (Fig. 4A), which and proteolytic fragments were analyzed. Trypsin, an agonist for was consistent with the findings of immunohistochemistry using PAR-2 (1–3), was used as a positive control. The PAR-2 peptide sections of human gingival tissues (9). By flow cytometric analyses, was rapidly cleaved at the site, R36-S37, by 5 nM PR3 or trypsin weak expression of PAR-2 was detected on the untreated cell surface, (Fig. 5B). The major peptide fragment was identified to be but the expression of PAR-1 and -3 on the cell surface was below the SLIGKVDGT, the PAR-2 tethered ligand, by sequencing. The detectable limit (Fig. 4B). The expression of PAR-2 was markedly measured molecular mass (890.8) was in good agreement with the augmented by PR3 treatment for 1 h, an effect that was inhibited calculated value (890.2). The fragment, SSKGR, was not detected The Journal of Immunology 4599 Downloaded from http://www.jimmunol.org/

FIGURE 4. Expression of the PAR family in oral epithelial cells. A, Total RNA was extracted from KB (lane 2), HSC-2 (lane 3), and PBMCs (lane 4), and cDNA was prepared and analyzed for the expression of PAR-1, PAR-2, PAR-3, PAR-4, and GAPDH by RT-PCR. Water control was loaded in lane 1. B, Confluent HSC-2 cells were pretreated with or without 30 nM cytochalasin B (cyto B) for 30 min or 1 ␮g/ml cycloheximide (CHX) for6hat 37°C. Then cells were stimulated with or without PR3 (10 ␮g/ml) for1hinthepresence or absence of 30 nM cytochalasin B or 1 ␮g/ml cycloheximide. After the incubation, cells were collected by CDS, and the expression of PAR-1, PAR-2, and PAR-3 of the cells was analyzed by flow cytometry. C and D, Confluent HSC-2 cells were stimulated with or without 10 ␮g/ml PR3 for the time indicated (C), or 1 ␮M trypsin or 100 ␮M PAR-2 agonist peptide (SLIGKV, PAR-2AP) for1h(D). After the incubation, cells were collected by CDS, and the expression of PAR-2 of the cells was analyzed by flow cytometry. The number in B–D indicates the degree of increases in PAR expression after the treatments evaluated on the basis of the value obtained with by guest on October 2, 2021 untreated cells. The results presented were representative of four different experiments demonstrating similar results.

probably because it was further cleaved to SSK and GR. Upon These results clearly indicated that the PR3-induced activation of digestion with 500 nM PR3 or trypsin, the large fragment was the cells is mediated by PAR-2. cleaved at the site, K41-V42, to give another major peak, which PR3 induced only marginal response in KB cells at 0.01 ␮M, corresponded to SLIGK (measured value, 517.9; calculated value, and induced a good response at 1 ␮M, which pattern was the same 517.8). PR3 from two different sources showed the identical result. as that with trypsin, as assessed by Ca2ϩ mobilization (Fig. 7), In contrast, thrombin, which does not activate PAR-2 (1–3), did indicating that the potency and efficacy are basically the same as not cleave the PAR-2 peptide. those of trypsin. PAR-2 agonist peptide SLIGKV required 100 times greater concentration to induce the same response. Involvement of PAR-2 in the PR3-induced activation of oral In common with many G protein-coupled receptors, the princi- epithelial cells pal mechanism of PAR-mediated activation is through G␣q pro- To confirm that PR3-induced activation of oral epithelial cells is teins, resulting in activation of PLC (1–3). To examine that PLC is mediated by PAR-2, Ca2ϩ mobilization in the cells was measured also involved in PR3-induced activation of oral epithelial cells, on exposure of KB cells to PR3, trypsin, and synthetic PAR ago- HSC-2 cells were stimulated with PR3 in the presence of the PLC nist peptides. Trypsin induced a Ca2ϩ response, and abolished the inhibitor U73122 or the control compound U73343 for 24 h. The response to a second application of trypsin, although the cells re- PAR-1 and -2 agonist peptides were used as control, and both sponded to the PAR-1 agonist peptide, SFLLRN (Fig. 6A). PR3 as peptides induced production of IL-8 and MCP-1 (Fig. 8), indicat- well as the PAR-2 agonist peptide SLIGKV induced the mobili- ing that the epithelial cells are activated through not only PAR-2, zation of Ca2ϩ in the cells the same as trypsin (Fig. 6, B and F). but also PAR-1. The inhibition of PLC completely abolished the SLIGKV stimulation rendered cells refractory to subsequent stim- production of cytokine induce by both agonist peptides and PR3. ulation with PR3, but not SFLLRN (Fig. 6C), while stimulation with SFLLRN had no such effect (Fig. 6D). Furthermore, trypsin Activation of oral epithelial cells by stimulated neutrophils inhibited a subsequent response to PR3 and to SLIGKV (Fig. 6E), To examine whether stimulated neutrophils activate oral epithelial and PR3 also desensitized responses to PAR-2 agonists, trypsin cells, confluent monolayers of HSC-2 were cocultured with indi- (Fig. 6G), and SLIGKV (Fig. 6H). HSC-2 and PR3 from two dif- cated number of neutrophils in the presence or absence of FMLP ferent sources showed the same results, although the Ca2ϩ mobi- for 24 h. Neutrophils at 0.5 ϫ 105 cells with 0.01 ␮M FMLP lization in HSC-2 was lower than that in KB (data not shown). started to induce production of IL-8 from the cells, and increasing 4600 ACTIVATION OF ORAL EPITHELIAL CELLS BY PR3 VIA PAR-2

cleaved the PAR-2 peptide with exposure of its tethered ligand; 4) trypsin and a synthetic PAR-2 agonist peptide induced the mobilization of intracellular Ca2ϩ, and rendered cells refractory to subsequent stimulation with PR3; and 5) a PAR-2 agonist peptide as well as PR3 induced cytokine production by the cells, which was inhibited by the PLC inhibitor. Lactate dehydrogenase activity was not detected in the super- natant of the epithelial cells after PR3 treatment, as previously reported (27), which indicates that PR3 has no adverse effect on the epithelial cell viability. Evidence for the activation of oral epithelial cells by PR3 includes: 1) the expression of IL-8, and MCP-1 and ICAM-1 mRNA was increased and induced by PR3, respectively (Figs. 1 and 2), and 2) PR3 induced the mobilization of intracellular Ca2ϩ in the epithelial cells (Fig. 6). In contrast to PR3, two other neutrophil serine proteinases, HLE and Cat G, showed only marginal activity on oral epithelial cells, as assessed from the production of MCP-1 (Fig. 1C). HLE and Cat G also induced the mobilization of intracellular Ca2ϩ in oral epithelial cells to a small extent as compared with PR3 and trypsin, and PR3, Downloaded from trypsin, and a PAR-2 agonist peptide rendered cells refractory to subsequent exposure to HLE and Cat G (data not shown), although it is reported that HLE and Cat G also inactivate PAR-1 by cleaving downstream of the tethered ligand (10). The findings indicate that PR3 is most involved in activating oral epithelial cells through PAR-2 among the three neutrophil serine proteinases. It is reported that PR3 has an elastase-like specificity for Ala, http://www.jimmunol.org/ Ser, and Val at the P1 site (13), but there was no evidence that it cleaves Arg-X and Lys-X bonds. The present study showed that FIGURE 5. Analysis of cleavage products of PAR-2 peptide. A, The peptide sequence of human PAR-2 used in this study. The number refers to PR3 rapidly cleaved between Arg and Ser and relatively ineffi- the sequence region of the intact receptor. The slash indicates a putative ciently between Lys and Val, which was the same specificity as 36 37 trypsin cleavage site that leads to exposure of the tethered ligand domain trypsin and tryptase (Fig. 5B) (37), indicating that the R -S site (underlined residues) in the intact receptor (37). B, The peptide (200 ␮M) and surrounding the site of the PAR-2 are structurally accessible was incubated with the indicated concentrations of PR3, trypsin, and by trypsin, tryptase, and PR3, but not by thrombin. thrombin for 30 min at 37°C, and separated by reversed-phase HPLC. The The present study showed that PR3 promoted the surface ex- by guest on October 2, 2021 flow rate was 1 ml/min. AUFS, absorbance units at full scale. Each peak pression of PAR-2 on oral epithelial cells, which was inhibited by (P1 and P2) was then analyzed by a protein sequencer and a mass spec- cytochalasin B, but not by cycloheximide (Fig. 4B). The longer trometry. P1, no cleaved peptide; P2, SLIGKVDGT. The results presented incubation with PR3 for up to 24 h neither up-regulated PAR-2 were representative of three different experiments demonstrating similar expression nor induced PAR-1 and -3 on the cell surface (data not results. shown). This indicates that PAR-2 was inducible receptor from intracellular storage, and not required de novo synthesis among the PAR family. In support of this, it was reported that the expression concentrations of FMLP resulted in a dose-dependent increase of of PAR-2 on human endothelial cells was also up-regulated by the IL-8 production (Fig. 9A). A significant increase in the pro- LPS as well as IL-1␣ and TNF-␣ without an effect on PAR-1 duction of MCP-1 was observed at 5 ϫ 105 cells of neutrophils expression (40), and that PAR-2 expression in asthmatic bronchial with 0.01 ␮M FMLP, and the MCP-1 production reached a plateau epithelium was significantly increased in comparison with normal at 0.1 ␮M FMLP (Fig. 9B). Unstimulated neutrophils lacked ac- epithelium (41). Recently, the PAR-2-induced activation of kera- tivity regardless of number of neutrophils. In contrast to the sol- tinocytes was shown to be mediated by the mitogen-activated pro- uble PR3, as shown in Fig. 3, the production of IL-8 and MCP-1 tein kinase and NF-␬B pathways (42), which are important for the induced by stimulated neutrophils was only partially inhibited by gene expression of cytokines. Thus, it is conceivable that the ac- serine proteinase inhibitors and serum (Fig. 9, C and D). tivation of oral epithelial cells by PR3 is mediated by these pathways involved in producing inflammatory cytokines and express- Discussion ing adhesion molecules on the cell surface. The present study showed that human neutrophil PR3 activates PR3 is secreted as a soluble form by activated neutrophils, and oral epithelial cells to induce production of IL-8 and MCP-1 and also exists as a membrane-bound form on neutrophils. Previous expression of ICAM-1 through PAR-2. Although it is reported that study demonstrated that each azurophil granule of neutrophils con- PR3 inactivates PAR-1 by cleaving downstream of the tethered tains PR3 at 13.4 mM, and the activation of neutrophils resulted in ligand (10), there was no evidence that it does the same to PAR-2. about a 10-fold increase in membrane-bound PR3 (14), indicating The present study proves for the first time that the PR3-induced that the local concentration of PR3 around activated neutrophils is activation of oral epithelial cells is mediated by PAR-2, based on high enough to activate PAR2. In addition, although serum con- the following evidence: 1) oral epithelial cells strongly express tains abundant naturally occurring proteinase inhibitors, mem- PAR-1 and PAR-2 mRNA and weakly express PAR-3 mRNA, and brane-bound PR3 is substantially resistant to inhibition by natu- ␣ PAR-2 expression on the cell surface was induced by PR3; 2) rally occurring inhibitors such as 1-AT and elafin, even when activation of the cells by PR3 was inhibited by serine proteinase these inhibitors are used at a 100- to 300-fold molar excess over inhibitors and serum; 3) PR3 and trypsin, an agonist for PAR-2, the enzyme (14). In support of this possibility, the results in Fig. 9 The Journal of Immunology 4601

FIGURE 6. Effect of trypsin, PR3, and PAR agonist peptides on calcium mobilization in oral epithelial cells. A–H, Fura 2-loaded KB cells were exposed to 1 ␮M trypsin, 10 ␮g/ml PR3, 100 ␮M PAR-1 agonist peptide (SFLLRN, PAR-1AP), and 100 ␮M PAR-2 agonist peptide (SLIGKV, PAR- 2AP) as indicated sequence, and the change in intracellular calcium was monitored. The results presented were representative of three different experiments demonstrating similar results. Downloaded from http://www.jimmunol.org/

showed that activation of oral epithelial cells by FMLP-stimulated neutrophils was only partially inhibited by serine proteinase inhibitors and serum, suggesting that the activation is likely to occur in vivo. The present study showed that oral epithelial cells are activated to produce IL-8 and MCP-1 and highly express ICAM-1 on the by guest on October 2, 2021 cell surface in response to PR3 through the PAR-2 pathway. IL-8 is a major chemokine responsible for the activation of neutrophils and migration of neutrophils and T cells to inﬂammatory sites (43). MCP-1 plays a critical role in the activation and migration of monocytes, T cells, and NK cells, and is an important factor in the development of Th1 and Th2 responses (44, 45). ICAM-1 is one of

FIGURE 8. Effect of PLC inhibition on the PR3-induced production of IL-8 and MCP-1 by oral epithelial cells. Conﬂuent HSC-2 cells were pretreated with or without U73122 or U73343 at the dose indicated for 30 min. FIGURE 7. Potency and efﬁcacy of PR3 compared with those of other Then cells were stimulated with the PAR-1 agonist peptide (PAR-1AP; 100 PAR-2 agonists. Fura 2-loaded KB cells were exposed to PR3, trypsin, and ␮M), the PAR-2 agonist peptide (PAR-2AP; 100 ␮M), or PR3 (10 ␮g/ml) PAR-2 agonist peptide (SLIGKV, PAR-2AP) at the concentrations indi- for 24 h. Concentrations of IL-8 and MCP-1 in the culture supernatants p Ͻ 0.01 compared ,ءء .cated, and the change in intracellular calcium was monitored. PR3, 1 were determined by ELISA. Error bars indicate SD ␮M ϭ 29 ␮g/ml; trypsin, 1 ␮M ϭ 24 ␮g/ml; PAR-2AP, 100 ␮M ϭ 61.6 with the respective control (PAR-1AP, PAR-2AP, or PR3 alone). The re- ␮g/ml. The results presented were representative of three different exper- sults presented were representative of three different experiments demon- iments demonstrating similar results. strating similar results. 4602 ACTIVATION OF ORAL EPITHELIAL CELLS BY PR3 VIA PAR-2

FIGURE 9. Activation of oral epithelial cells with stimulated neutrophils. Conﬂuent HSC-2 cells were cocultured with the indicated number of neutrophils in the presence or absence of FMLP at the dose indicated for 24 h, and concentrations of IL-8 (A) and MCP-1 (B) in the culture supernatants were determined by ELISA. Conﬂuent HSC-2 cells were cultured with 106 of neutrophils and 10 ␮M FMLP in the presence or absence of ␮ ␣ ␮ Pefabloc SC (40 M), 1-AT (1 g/ml), or FCS (1 and 10%) for 24 h, and concentrations of IL-8 (C) and MCP-1 (D) in the culture supernatants were determined by ELISA. The percentage of cytokine production was calculated based on the value obtained without inhibitors (C, 4.9 Ϯ 0.03 ng/ml; D, 7.3 Ϯ 0.09 ng/ml). Error Ͻ ءء bars indicate SD. , p 0.01 compared with no inhibi- Downloaded from tor. The results presented were representative of three different experiments demonstrating similar results. http://www.jimmunol.org/

␤ the major adhesion molecules interacting with the 2 integrin fam- cells (12, 17–20), although the underlying mechanism was unclear. ily, including LFA-1 and Mac-1 present on neutrophils, mono- Therefore, the present study may provide one of the mechanisms cytes, and T cells (46). Fig. 2 shows that ICAM-1 expressed on of the cell activation by PR3 through PAR-2, and suggests that oral epithelial cells in response to PR3 contributes extensively to PR3 could activate the PAR-2-expressing cells, and regulate a

the interaction with neutrophils (via Mac-1 and LFA-1), and the number of inflammatory processes, and that the control of PAR- by guest on October 2, 2021 ICAM-1-mediated interaction may further augment the activation 2-activating proteinases including PR3 at inflammatory sites might of oral epithelial cells through PAR-2. IL-18 is another cytokine be beneficial in the regulation of inflammation. closely involved in controlling Th1 and Th2 responses (47), and we recently showed that oral epithelial cells produce an active Acknowledgments IL-18 when stimulated by PR3 in combination with LPS after We thank N. Takahashi and Y. Iwami (Tohoku University School of Den- IFN-␥ priming (27). The inflamed site is characterized by an in- tistry) for helpful advice concerning calcium mobilization assay and for filtration of neutrophils. Infiltration of neutrophils into gingival generously allowing us the use of the spectrofluorometer, and E. Nemoto tissues is an early event in gingival inflammation, and neutrophils (Tohoku University School of Dentistry) for helpful discussion. are the predominant leukocytes in the gingival crevicular fluid References (GCF) (ϳ90%) (48). It is also evident that GCF has neutrophil 1. De´ry, O., C. U. Corvera, M. Steinhoff, and N. W. Bunnett. 1998. Proteinase- serine proteinase activities against MeOSuc-Ala-Ala-Pro-Val-p- activated receptors: novel mechanisms of signaling by serine proteases. nitroanilide for HLE and Suc-Ala-Ala-Pro-Phe-p-nitroanilide for Am. J. Physiol. 274:C1429. 2. Coughlin, S. R. 2000. Thrombin signalling and protease-activated receptors. Na- Cat G (49). Because MeOSuc-Ala-Ala-Pro-Val-p-nitroanilide is a ture 407:258. PR3 substrate as well (13), PR3 is most likely present in GCF. 3. O’Brien, P. J., M. Molino, M. Kahn, and L. F. Brass. 2001. Protease activated Therefore, the present study as well as our previous study (27) receptors: theme and variations. Oncogene 20:1570. 4. Xu, W.-F., H. Andersen, T. E. Whitmore, S. R. Presnell, D. P. Yee, A. Ching, suggest that the PAR-2-mediated activation of oral epithelial cells T. Gilbert, E. W. Davie, and D. C. Foster. 1998. Cloning and characterization of by PR3, which is released by active neutrophils, further augments human protease-activated receptor-4. Proc. Natl. Acad. Sci. USA 95:6642. accumulation of neutrophils and immune cells by controlling Th1 5. Kahn, M. L., M. Nakanishi-Matsui, M. J. Shapiro, H. Ishihara, and S. R. Coughlin. 1999. Protease-activated receptors 1 and 4 mediate activation of or Th2 responses at periodontitis sites, and consequently plays an human platelets by thrombin. J. Clin. Invest. 103:879. important role in host defense against periodontal pathogens. 6. Kahn, M. L., Y. W. Zheng, W. Huang, V. Bigornia, D. Zeng, S. Moff, R. V. Farese, Jr., C. Tam, and S. R. Coughlin. 1998. A dual thrombin receptor PAR-2 is expressed in the gastrointestinal tract, pancreas, kid- system for platelet activation. Nature 394:690. ney, liver, airway, prostate, ovary, eye, and skin, and is found in 7. Schmidt, V. A., W. C. Nierman, D. R. Maglott, L. D. Cupit, K. A. Moskowitz, epithelial and endothelial cells, smooth muscle cells, keratinocytes, J. A. Wainer, and W. F. Bahou. 1998. The human proteinase-activated receptor-3 (PAR-3) gene: identification within a PAR gene cluster and characterization in T cell lines, and certain tumor cells (1–3). It was recently reported vascular endothelial cells and platelets. J. Biol. Chem. 273:15061. that stimulation of PAR-2 with an agonist peptide activates human 8. O’Brien, P. J., N. Prevost, M. Molino, M. K. Hollinger, M. J. Woolkalis, keratinocytes (42), eosinophils (50), and respiratory epithelial cells D. S. Woulfe, and L. F. Brass. 2000. Thrombin responses in human endothelial cells: contributions from receptors other than PAR1 include the transactivation of (51) to induce inflammatory mediators, and up-regulates keratin- PAR2 by thrombin-cleaved PAR1. J. Biol. Chem. 275:13502. ocyte phagocytosis (52). PR3 is a major target Ag of anti- 9. Lourbakos, A., J. Potempa, J. Travis, M. R. D’Andrea, P. Andrade-Gordon, R. Santulli, E. J. Mackie, and R. N. Pike. 2001. Arginine-specific protease from neutrophil cytoplasmic Abs (15, 16) and degrades extracellular Porphyromonas gingivalis activates protease-activated receptors on human oral matrix proteins (13). PR3 is also shown to activate many types of epithelial cells and induces interleukin-6 secretion. Infect. Immun. 69:5121. The Journal of Immunology 4603

10. Renesto, P., M. Si-Tahar, M. Moniatte, V. Balloy, A. Van Dorsselaer, D. Pidard, 30. Ferrante, A., and Y. H. Thong. 1980. Optimal conditions for simultaneous puri- and M. Chignard. 1997. Specific inhibition of thrombin-induced cell activation by fication of mononuclear and polymorphonuclear leukocytes from human blood by the neutrophil proteinases elastase, cathepsin G, and proteinase 3: evidence for the Hypaque-Ficoll method. J. Immunol. Methods 36:109. distinct cleavage sites within the aminoterminal domain of the thrombin receptor. 31. Sugawara, S., E. Nemoto, H. Tada, K. Miyake, T. Imamura, and H. Takada. 2000. Blood 89:1944. Proteolysis of human monocyte CD14 by cysteine proteinases (gingipains) from 11. Owen, C. A., and E. J. Campbell. 1999. The cell biology of leukocyte-mediated Porphyromonas gingivalis leading to lipopolysaccharide hyporesponsiveness. proteolysis. J. Leukocyte Biol. 65:137. J. Immunol. 165:411. 12. Bories, D., M.-C. Raynal, D. H. Solomon, Z. Darzynkiewicz, and Y. E. Cayre. 32. Sugawara, S., A. Sugiyama, E. Nemoto, H. Rikiishi, and H. Takada. 1998. Het- 1989. Down-regulation of a serine protease, myeloblastin, causes growth arrest erogeneous expression and release of CD14 by human gingival fibroblasts: char- and differentiation of promyelocytic leukemia cells. Cell 59:959. acterization and CD14-mediated interleukin-8 secretion in response to lipopoly- 13. Rao, N. V., N. G. Wehner, B. C. Marshall, W. R. Gray, B. H. Gray, and saccharide. Infect. Immun. 66:3043. J. R. Hoidal. 1991. Characterization of proteinase-3 (PR-3), a neutrophil serine 33. Bath, P. M. W., R. F. G. Booth, and D. G. Hassall. 1989. Monocyte-lymphocyte proteinase: structural and functional properties. J. Biol. Chem. 266:9540. discrimination in a new microtitre-based adhesion assay. J. Immunol. Methods 14. Campbell, E. J., M. A. Campbell, and C. A. Owen. 2000. Bioactive proteinase 3 118:59. on the cell surface of human neutrophils: quantification, catalytic activity, and 34. Hoidal, J. R., N. V. Rao, and B. Gray. 1994. Myeloblastin: leukocyte proteinase susceptibility to inhibition. J. Immunol. 165:3366. 3. Methods Enzymol. 244:61. 15. Braun, M. G., E. Csernok, W. L. Gross, and H. K. Muller-Hermelink. 1991. 35. Vu, T. H., D. T. Hung, V. I. Wheaton, and S. R. Coughlin. 1991. Molecular Proteinase 3, the target antigen of anticytoplasmic antibodies circulating in We- cloning of a functional thrombin receptor reveals a novel proteolytic mechanism gener’s granulomatosis: immunolocalization in normal and pathologic tissues. of receptor activation. Cell 64:1057. Am. J. Pathol. 139:831. 36. Storck, J., B. Kusters, M. Vahland, C. Morys-Wortmann, and E. R. Zimmermann. 16. Lu¨demann, J., B. Utecht, and W. L. Gross. 1990. Anti-neutrophil cytoplasm 1996. Trypsin induced von Willebrand factor release from human endothelial antibodies in Wegener’s granulomatosis recognize an elastinolytic enzyme. cells is mediated by PAR-2 activation. Thromb. Res. 84:463. J. Exp. Med. 171:357. 37. Molino, M., E. S. Barnathan, R. Numerof, J. Clark, M. Dreyer, A. Cumashi, 17. Sko¨ld, S., B. Rosberg, U. Gullberg, and T. Olofsson. 1999. A secreted proform J. A. Hoxie, N. Schechter, M. Woolkalis, and L. F. Brass. 1997. Interactions of of neutrophil proteinase 3 regulates the proliferation of granulopoietic progenitor mast cell tryptase with thrombin receptors and PAR-2. J. Biol. Chem. 272:4043. cells. Blood 93:849. 38. Tateno, H., A. Saneyoshi, T. Ogawa, K. Muramoto, H. Kamiya, and Downloaded from 18. Coeshott, C., C. Ohnemus, A. Pilyavskaya, S. Ross, M. Wieczorek, H. Kroona, M. Saneyoshi. 1998. Isolation and characterization of rhamnose-binding lectins A. H. Leimer, and J. Cheronis. 1999. Converting enzyme-independent release of from eggs of steelhead trout (Oncorhynchus mykiss) homologous to low density ␣ ␤ tumor necrosis factor and IL-1 from a stimulated human monocytic cell line lipoprotein receptor superfamily. J. Biol. Chem. 273:19190. in the presence of activated neutrophils or purified proteinase 3. Proc. Natl. Acad. 39. Korn, E. D. 1981. Useful tool for characterizing some of the polymerization Sci. USA 96:6261. properties of actin. Physiol. Rev. 62:672. 19. Berger, S. P., M. A. J. Seelen, P. S. Hiemstra, J. S. J. Gerritsma, E. Heemskerk, 40. Nystedt, S., V. Ramakrishnan, and J. Sundelin. 1996. The proteinase-activated F. J. van der Woude, and M. R. Daha. 1996. Proteinase 3, the major autoantigen receptor 2 is induced by inflammatory mediators in human endothelial cells: of Wegener’s granulomatosis, enhances IL-8 production by endothelial cells in comparison with thrombin receptor. J. Biol. Chem. 271:14910. vitro. J. Am. Soc. Nephrol. 7:694. http://www.jimmunol.org/ 41. Knight, D. A., S. Lim, A. K. Scaffidi, N. Roche, K. F. Chung, G. A. Stewart, and 20. Taekema-Roelvink, M. E. J., C. van Kooten, S. van der Kooij, E. Heemskerk, and P. J. Thompson. 2001. Protease-activated receptors in human airways: up-regu- M. R. Daha. 2001. Proteinase 3 enhances endothelial monocyte chemoattractant lation of PAR-2 in respiratory epithelium from patients with asthma. protein-1 production and induces increased adhesion of neutrophils to endothelial J. Allergy Clin. Immunol. 108:797. cells by up-regulating intercellular cell adhesion molecule-1. J. Am. Soc. Nephrol. 42. Kanke, T., S. R. Macfarlane, M. J. Seatter, E. Davenport, A. Paul, R. C. McKenzie, 12:932. and R. Plevin. 2001. Proteinase-activated receptor-2-mediated activation of stress- 21. Campanelli, D., P. A. Detmars, C. F. Nathan, and J. E. Gabay. 1990. Azurocidin activated protein kinases and inhibitory ␬B kinases in NCTC 2544 keratinocytes. and a homologous serine protease from neutrophils: differential antimicrobial and J. Biol. Chem. 276:31657. proteolytic properties. J. Clin. Invest. 85:904. 22. Lundqvist, C., V. Baranov, S. Teglund, S. Hammarstro¨m, and M.-L. Ham- 43. Baggiolini, M., B. Dewald, and B. Moser. 1994. Interleukin-8 and related che- marstro¨m. 1994. Cytokine profile and ultrastructure of intraepithelial ␥␦ T cells motactic cytokines; CXC and CC chemokines. Adv. Immunol. 55:97. in chronically inflamed human gingiva suggest a cytotoxic effector function. 44. Gu, L., S. Tseng, R. M. Horner, C. Tam, M. Loda, and B. J. Rollins. 2000. J. Immunol. 153:2302. Control of TH2 polarization by the chemokine monocyte chemoattractant pro- by guest on October 2, 2021 23. Huang, G. T.-J., S. K. Haake, J.-W. Kim, and N.-H. Park. 1998. Differential tein-1. Nature 404:407. expression of interleukin-8 and intercellular adhesion molecule-1 by human gin- 45. Huang, D., J. Wang, P. Kivisakk, B. J. Rollins, and R. M. Ransohoff. 2001. gival epithelial cells in response to Actinobacillus actinomycetemcomitans or Absence of monocyte chemoattractant protein 1 in mice leads to decreased local Porphyromonas gingivalis infection. Oral Microbiol. Immunol. 13:301. macrophage recruitment and antigen-specific T helper cell type 1 immune re- 24. Yumoto, H., H. Nakae, K. Fujinaka, S. Ebisu, and T. Matsuo. 1999. Interleukin-6 sponse in experimental autoimmune encephalomyelitis. J. Exp. Med. 193:713. (IL-6) and IL-8 are induced in human oral epithelial cells in response to exposure 46. Van Kooyk, Y., and C. G. Figdor. 2000. Avidity regulation of integrins: the to periodontopathic Eikenella corrodens. Infect. Immun. 67:384. driving force in leukocyte adhesion. Curr. Opin. Cell Biol. 12:542. 25. Sugiyama, A., A. Uehara, K. Iki, K. Matsushita, R. Nakamura, T. Ogawa, 47. Nakanishi, K., T. Yoshimoto, H. Tsutsui, and H. Okamura. 2001. Interleukin-18 S. Sugawara, and H. Takada. 2002. Activation of human gingival epithelial cells regulates both Th1 and Th2 responses. Annu. Rev. Immunol. 19:423. by cell-surface components of black-pigmented bacteria: augmentation of pro- 48. Suchett-Kaye, G., J.-J. Morrier, and O. Barsotti. 1998. Interactions between non- duction of interleukin-8, granulocyte colony-stimulating factor and granulocyte- immune host cells and the immune system during periodontal disease: role of the macrophage colony-stimulating factor and expression of intercellular adhesion gingival keratinocyte. Crit. Rev. Oral Biol. Med. 9:292. molecule 1. J. Med. Microbiol. 51:27. 49. Mailhot, J. M., J. Potempa, S. H. Stein, J. Travis, J. D. Sterrett, P. J. Hanes, and 26. Uehara, A., S. Sugawara, R. Tamai, and H. Takada. 2001. Contrasting responses C. M. Russell. 1998. A relationship between proteinase activity and clinical pa- of human gingival and colonic epithelial cells to lipopolysaccharide, lipoteichoic rameters in the treatment of periodontal disease. J. Clin. Periodontol. 25:578. acids and peptidoglycans in the presence of soluble CD14. Med. Microbiol. Im- 50. Miike, S., A. S. McWilliam, and H. Kita. 2001. Trypsin induces activation and munol. 189:185. inflammatory mediator release from human eosinophils through protease-acti- 27. Sugawara, S., A. Uehara, T. Nochi, T. Yamaguchi, H. Ueda, K. Hanzawa, vated receptor-2. J. Immunol. 167:6615. K. Kumagai, H. Okamura, and H. Takada. 2001. Neutrophil proteinase 3-medi- 51. Asokananthan, N., P. T. Graham, J. Fink, D. A. Knight, A. J. Bakker, ated induction of bioactive IL-18 secretion by human oral epithelial cells. J. Im- A. S. McWilliam, P. J. Thompson, and G. A. Stewart. 2002. Activation of pro- munol. 167:6568. tease-activated receptor (PAR)-1, PAR-2, and PAR-4 stimulates IL-6, IL-8, and 28. Momose, F., T. Araida, A. Negishi, H. Ichijo, S. Shioda, and S. Sasaki. 1989. prostaglandin E2 release from human respiratory epithelial cells. J. Immunol. Variant sublines with different metastatic potentials selected in nude mice from 168:3577. human oral squamous cell carcinomas. J. Oral Pathol. Med. 18:391. 52. Sharlow, E. R., C. S. Paine, L. Babiarz, M. Eisinger, S. Shapiro, and M. Seiberg. 29. Eagle, H. 1955. Propagation in a fluid medium of a human epidermoid carcinoma, 2000. The protease-activated receptor-2 up-regulates keratinocyte phagocytosis. strain KB. Proc. Soc. Exp. Biol. Med. 89:362. J. Cell Sci. 113:3093. The Journal of Immunology

Letter of Retraction

We wish to retract the article titled “Activation of Human Oral Epithelial Cells by Neutrophil Proteinase 3 Through Protease-Activated Receptor-2,” by Akiko Uehara, Shunji Sugawara, Koji Muramoto, and Haruhiko Takada, The Journal of Immunology, 2002, 169: 4594–4603. This retraction follows an investigation by Tohoku University into scientific misconduct. The investigation pointed out the following: 1. Fig. 1E: HSC-2, an oral epithelial cell line, is different from KB cells, but the patterns of their band staining for IL-8, MCP-1, and GAPDH cDNA are the same. 2. Fig. 2D: Bands for MCP-1 and GAPDH cDNA in Fig. 1E and ICAM-1 and GAPDH cDNA in this figure are the same. 3. Fig. 4A: Total RNA was extracted from KB (lane 2), HSC-2 (lane 3), and PBMCs (lane 4), and cDNA was prepared and analyzed for the expression of PAR1-4 and GAPDH by RT-PCR. However, three bands of PAR3 and a band of PAR4 are the same. Furthermore, the patterns of GAPDH in this article and those in Fig. 1B and 2A of The Journal of Immunology, 2003, 170: 5690–5696 and in Fig. 2A of Clinical and Diagnostic Laboratory Immunology, 2003, 10: 286–292 are the same. 4. Fig. 4B and 4D: Three panels of the expression of PAR1 with PR3 and PR3 + cytochalasin B (Cyto B) and PAR3 with PR3 + Cyto B are the same. Two panels of the expression of PAR2 with PR3 + cycloheximide (CHX) in Fig. 4B and with trypsin in Fig. 4D are the same. The first author, who conducted these experiments, could not counter the argument by adducing raw data at the investigation, and the investigation recognized them as scientific misconduct. Therefore, we wish to retract the article. We deeply regret these errors and apologize to the scientific community for the need to retract the article.

Shunji Sugawara Tohoku University Graduate School of Dentistry Sendai, Japan Koji Muramoto Tohoku University Graduate School of Life Sciences Sendai, Japan

Copyright Ó 2010 by The American Association of Immunologists, Inc. 0022-1767/10/$16.00 www.jimmunol.org/cgi/doi/10.4049/jimmunol.1090011