Dental pulp inflammation often results from dissemination of periodontitis caused mostly by Porphyromonas gingivalis infection. Calcitonin gene-related peptide and substance P are proinflammatory neuropeptides that increase in inflamed pulp tissue. To study an involvement of the periodontitis pathogen and neuropeptides in pulp inflammation, we investigated human dental pulp cell neuropeptide release by arginine-specific cysteine protease (RgpB), a cysteine proteinase of P. gingivalis, and participating signaling pathways. RgpB induced neuropeptide release from cultured human pulp cells (HPCs) in a proteolytic activity-dependent manner at a range of 12.5–200 nM. HPCs expressed both mRNA and the products of calcitonin gene-related peptide, substance P, and proteinase-activated receptor-2 (PAR-2) that were also found in dental pulp fibroblast-like cells. The PAR-2 agonists, SLIGKV and trypsin, also induced neuropeptide release from HPCs, and HPC PAR-2 gene knockout by transfection of PAR-2 antisense oligonucleotides inhibited significantly the RgpB-elicited neuropeptide release. These results indicated that RgpB-induced neuropeptide release was dependent on PAR-2 activation. The kinase inhibitor profile on the RgpB-neuropeptide release from HPC revealed a new PAR-2 signaling pathway that was mediated by p38 MAPK and activated transcription factor-2 activation, in addition to the PAR-2-p44/42 p38MAPK and -AP-1 pathway. This new RgpB activity suggests a possible link between periodontitis and pulp inflammation, which may be modulated by neuropeptides released in the lesion.
Porphyromonas gingivalis, a Gram-negative anaerobic bacterium, is one of the major pathogens of periodontitis, which causes periodontal tissue destruction and consequent teeth loss. Cell surface-associated and secretory trypsin-like cysteine proteinases (gingipains) produced by P. gingivalis are important virulence factors and are implicated in the development of periodontitis (1, 2). Periodontitis often leads to acute pulpitis and pulpal necrosis by dissemination of the pathogens and/or their virulence components (3, 4, 5, 6, 7, 8, 9, 10). Arginine-specific cysteine protease (RgpB)2 is a gingipain (11) and has various proinflammatory activities including vascular permeability enhancement and edema induction (1, 2), which can occur in acute pulpitis (12, 13). RgpB may be associated with pulpal inflammation.
Calcitonin gene-related peptide (CGRP) and substance P (SP) are neuropeptides that are widely codistributed in central and peripheral neurons (14, 15, 16). Both neuropeptides induce vasodilation (16), local blood flow increase (15), vascular permeability enhancement (17), and pain (18, 19), thus modulating inflammation (20, 21). CGRP and SP are present in pulpal and gingival tissues (17, 22, 23) and are increased markedly in human teeth exposing pulp due to advanced caries and/or those with painful pulpitis (24, 25), which suggests a modulation of pulpal inflammation by these peptides. CGRP and SP are released from neurons by activation of proteinase-activated receptors (PAR) (14), a family of four subtypes of seven transmembrane domain receptors coupled with G protein (26). RgpB induces platelet cytosolic Ca2+ increase and aggregation via PAR-1 and -4 (27) and gingival epithelial cell IL-6 secretion via PAR-1 and -2 (28). RgpB may induce the release of CGRP and SP from pulp cells through PAR activation and participate in the pulpitis that accompanies periodontitis.
To study an involvement of RgpB and the neuropeptides in dental pulp inflammation, we investigated neuropeptide release from human pulp cells by RgpB and its mechanism including activation of PAR and the following signaling pathway.
Materials and Methods
The synthetic PAR-2 agonist peptide H-SLIGKV-NH2, corresponding to the tethered ligand, was purchased from Asahi Technoglass. Leupeptin was purchased from Peptide Institute. SB202190 and SP600125 were purchased from Calbiochem. U0126 was purchased from Promega. Trypsin 250 was purchased from Difco Laboratories. Other reagents were supplied by Sigma-Aldrich.
Anti-human PAR-1 and PAR-2 mAbs and anti-PAR-2, PAR-3, and PAR-4 polyclonal Abs were purchased from Santa Cruz Biotechnology. Rabbit anti-SP and anti-CGRP Abs were purchased from ICN Biomedicals and Peninsula Laboratories, respectively. mAb 1B10 was purchased from Sigma-Aldrich. MAPK assay kits (containing polyclonal Abs against p38, JNK/stress-activated protein kinase, and p44/42, phospho-p38, phospho-JNK/stress-activated protein kinase, and phospho-p44/p42) and anti-phospho-ATF-2 Ab were purchased from Cell Signaling Technology.
Purification and activation of RgpB
RgpB was isolated as previously described by Potempa et al. (29). Purified RgpB was activated with 10 mM l-cysteine in 0.2 M HEPES buffer, pH 8.0, containing 5 mM CaCl2 at 37°C for 5 min and then kept at room temperature. The proteinase was diluted with 50 mM Tris-HCl, pH 7.4, containing 0.1 M NaCl and 5 mM CaCl2 immediately before use. The amount of active enzyme in each 0purified proteinase was determined by active site titration with d-phenylalanylprolylarginylcholoromethyl ketone (d-Phe-Pro-Arg-CH2Cl). The concentration of active RgpB was calculated from the amount of the inhibitor needed for complete inactivation of the proteinase (30).
Human pulp cells (HPCs) were prepared as described previously (31). Briefly, normal human dental pulp tissue was obtained from extracted first premolars and was cultured in α-MEM (ICN Biomedicals) supplemented with 10% FBS. The cells at subculture passage 5–9 (∼80% confluent) were used in this study. All experiments using inhibitor or/and stimulators were performed in α-MEM containing 0.2% FBS. Human gingival fibroblasts (HGF) were obtained from healthy gingival tissues cultured in DMEM (Life Technologies) supplemented with 10% FBS. Outgrown cells from the tissue cultures were subcultured and used at later passages 5–9. Human alveolar epithelial cells (A549) cultured as described previously (32) were seeded into collagen (type I)-coated tissue culture dishes and were grown to confluence in DMEM supplemented with 10% FBS at 37°C in 5% CO2.
Preparation/concentration of samples
HPCs (5 × 104 cells/well) were suspended onto 24-well cell culture plates (Nunclon) in the appropriate growth medium supplemented with 10% FBS. After an 18-h incubation in serum-free α-MEM, the 80% confluent cells were stimulated with RgpB in the presence or absence of leupeptin. The supernatants (2 ml) were concentrated according to the ELISA kit protocol for sample preparation. Briefly, cartridges of Sep-Pak C18 (Waters) were equilibrated with acetonitrile, followed by 1% trifluoroacetic acid (TFA). Samples diluted with an equal volume of 1% TFA were slowly loaded onto the column, and the column was washed with 10 ml of 1% TFA. Adsorbed peptides were then eluted with 3 ml of 60% (v/v) acetonitrile. The eluant was concentrated by microcentrifugal vacuum concentrator (MV-100; TOMY TECH) and stored at −20°C.
CGRP and SP assay
Samples reconstituted with 0.5 ml of assay buffer were assayed for CGRP and SP using ELISA kits for CGRP (BACHEM/Peninsula Laboratories) and SP (Assay Designs) according to the manufacturer’s instructions. Briefly, CGRP assay is based on the competitive binding of CGRP and the biotinylated CGRP in standard solutions or samples to an anti-CGRP Ab. The standard CGRP solutions (0.04–1,000 ng/ml) and samples were measured in duplicate. An appropriately diluted Ab against CGRP and nonbiotinylated CGRP (either standard or 0.05 ml of unknown concentrated samples) were mixed in a microtiter plate well. After incubation at room temperature for 1 h, 0.025 ml of biotinylated CGRP was poured to the well and further incubated for 2 h. Unbound biotinylated CGRP was removed from the well by washing five times with the assay buffer, and the well was dried. The 0.1 ml of streptavidin-conjugated HRP was added to the well and incubated for 1 h to make an immobilized primary Ab-biotinylated peptide complex. After excess streptavidin-conjugated-HRP was washed away, the bound complex was reacted with 0.1 ml of 3,3′,5,5′-tetramethylbenzidine dihydrochloride for 30 min. The enzyme-substrate reaction was stopped, and the yellow color generated was measured at 450 nm. The minimum detection limit of the assay is 0.04–0.06 ng/ml with intra- and interassay coefficient of variation <5% and 14%, respectively. The SP assay is based on the competitive binding to a limited quantitative SP-specific rabbit Ab between SP and alkaline phosphate-labeled SP. The standard solutions of SP (9.76–10,000 pg/ml) and samples were measured in duplicate. An appropriately diluted standard SP and 0.05 ml of a sample was mixed in a microtiter plate well. The alkaline phosphatase-conjugated SP for 0.05 ml was then added to each well. After a 2-h incubation with 0.05 ml of rabbit anti-SP at room temperature, the excess reagents were washed away from the well, followed by an addition of p-nitrophenyl phosphate. After a short-time incubation, the enzyme reaction was stopped, and the yellow color generated was measured at 405 nm with a Multilabel counter using the Wallac 1420 ARVO program. The minimum detection limit of the assay is 8.04 pg/ml with intra- and interassay coefficient variation being <6.7% and 4.2%, respectively.
In both assays, the OD values in a well were inversely proportional to the concentrations of CGRP or SP. Peptide concentrations of samples were estimated from the standard best fit lines made from the plots of the OD values of neuropeptides with known concentrations.
To confirm reliability of the ELISA, the supernatant of cells stimulated by RgpB at 200 nM was concentrated and reconstituted with the assay buffer as described above. The supernatant was serially 2-fold diluted, and neuropeptide concentrations of diluted samples were measured by ELISA. Best fit lines were obtained from the OD values of serial 2-fold diluted samples of the supernatant and compared with corresponding standard best fit lines.
mRNA detection of the neuropeptides and PAR-1, -2, -3, and -4
Total RNA was extracted using TRIzol reagent according to the manufacturer’s instructions (Invitrogen). First-strand cDNA was synthesized by reverse transcriptase using a commercial RT-PCR kit (Takara Biomedicals), and the reaction was performed following the manufacturer’s instructions. Briefly, 1 μg of RNA was added to a 20-μl reaction volume containing 1 μl of random primers, 2 μl of 10× first-strand buffer, 0.16 μl of RNase inhibitor (1 U/μl), 8 μl of 1 mM dNTP, and 0.14 μl of reverse transcriptase (0.25 U/μl). The reaction mix was incubated at 37°C for 60 min and then at 90°C for 5 min. The resulting cDNA mixture was amplified with Taq polymerase, and the following specific primers were synthesized by Hokkaido System Science: CGRP, 5′-GAACTATTGCTAAATGCAGAACAAGCT-3′ and 5′-ATTTGCTACCAGATAAGCCAATGAGA-3′ (PCR product, 119 bp; GenBank accession no. X02330); SP, 5′-GACAGCGACCAGATCAAGGAGGAA-3′ and 5′-CAGCATCCCGTTTGCC-3′ (PCR product, 121 bp; GenBank accession no. U37529); PAR-1, 5′-TACACCGGAGTGTTTGTAGT-3′ and 5′-TTGAGGACGAGAGGCACTAC-3′ (PCR product, 395 bp; GenBank accession no. M62424); PAR-2, 5′-GGTAAGGTTGATGGCACATC-3′ and 5′-TGGTCTGCTTCACGACATAC-3′ (PCR product, 509 bp; GenBank accession no. U34038); PAR-3, 5′-ATCTCATAGCTTTGTGCCTG-3′ and 5′-CACGCCTGTAATCCAGCACT-3′ (PCR product, 488 bp; GenBank accession no. U92971); PAR-4, 5′-AGTCTGTGCCAATGACAGTG-3′ and 5′-TCATGGCAGAGCACGCGATC-3′ (PCR product, 534 bp; GenBank accession no. AF055917); GAPDH, 5′-CATCACCATCTTCCAGGAGC-3′ and 5′-CATGAGTCCTTCCACGATACC-3′ (PCR product, 286 bp). Amplification conditions were 35 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 90 s for PAR-1 and PAR-2; 35 cycles of 94°C for 30 s, 59°C for 30 s, and 72°C for 90 s for PAR-3 and PAR-4; and 30 cycles of 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s with primer extension time at 72°C for 30 s for CGRP; 30 cycles of 94°C for 30 s, 62°C for 30 s, and 72°C for 30 s with primer extension time at 72°C for 30 s for SP; and 28 cycles of 94°C for 30 s, 56°C for 30 s, and 72°C for 90 s for GAPDH. To eliminate the possibility of false positive results of neuropeptide expression by contaminating DNA, RT-PCR was performed in the absence of reverse transcriptase, primers or RNA was shown as controls.
Detection of PARs on HPCs by flow cytometric analysis
HPCs and A549 cells were grown in monolayers in cell culture dishes in α-MEM or DMEM supplemented with 10% FBS, respectively. Cell monolayers were dispersed and resuspended at a final concentration of 3 × 106 cells/ml. Cells were incubated with 20 μg/ml anti-PARs Abs or with nonimmune serum for 1 h, followed by the FITC-conjugated second Ab (ICN Pharmaceuticals) for 30 min. Fluorescence was analyzed with a FACSCan (Beckman Coulter).
Dental pulp tissue preparation
Dental pulp tissues (n = 20) were obtained from patients with informed consent according to the guideline approved by the Ethical Committee at Kagoshima University Dental School. Teeth were excluded from the study if any physiological root resorption was found. After extraction, the groove was cut off from the buccal aspect of the tooth, and the pulp tissue taken carefully from its chamber was fixed in 4% paraformaldehyde and embedded in paraffin.
Paraffin-embedded sections (5 μm) of pulp tissues were deparaffinized in xylene and rehydrated through decreasing concentrations of ethanol. To block endogenous peroxidase activity, we treated sections with 3% (v/v) H2O2 for 20 min and then processed them for immunostaining using primary Abs and LSAB+ Kit (DAKO). Sections were incubated in each primary Ab for 1 h at room temperature (anti-PAR-1, -2, -1B10 mouse mAbs, 100× dilution; anti-PAR-2 goat Ab, 50× dilution; anti-PAR-3 and -4 goat Abs, 100× dilution; anti-CGRP rabbit Ab, 250× dilution; anti-SP rabbit Ab, 200× dilution). After a washing, sections were incubated in LINK (biotinylated anti-rabbit, anti-mouse, and anti-goat Igs). Sections were washed and further incubated in streptavidin conjugated to HRP for 30 min. As negative controls, each isotype nonimmune serum (2 μg/ml) of the same species was used instead of the primary Ab. The specificity of the Abs for PAR-2 was confirmed by preincubation with its specific blocking peptides (Santa Cruz Biotechnology) (1/50 dilution) at 37°C for 1 h. All incubation procedures were performed in a humid chamber. The peroxidase bound to the tissue sections was detected using 3,3′-diaminobenzidine and H2O2. The tissue sections were counterstained with hematoxylin, mounted in Aquatex (Merck) and examined with an Olympus BH 2 light microscope.
HPCs were grown overnight in a Lab-Tech chamber slide (Nalge Nunc International). Cells were fixed with 4% paraformaldehyde and permeabilized in PBS with 0.1% Triton X-100 for 15 min at room temperature. Nonspecific binding was blocked by incubating the cells with 1% normal goat serum in PBS for 30 min. Slides were incubated in each primary Ab (anti-CGRP, -SP, or -PAR-2 Ab; 50× dilution and anti-1B10 Ab; 100× dilution) in PBS containing 0.1% Triton X-100 for 12 h at 4°C. After a washing, slides were incubated with tetramethylrhodamine (Alexa Fluor 546; Molecular Probes)-conjugated goat Ab against rabbit IgG (1/100 dilution with PBS containing 0.1% Triton X-100) for 30 min in the dark. Finally, immunoreactivity was visualized using a confocal microscope (Leica TCS4D; Leica).
Transient transfection experiments
Human sense or antisense synthetic phosphorothioate-modified oligonucleotides homologous to PAR-2 and c-jun were purified by HPLC. All PAR-2 oligonucleotides used in this study were evaluated for homology to preexisting genes using the BLAST homology search program and the current version of GenBank database. The PAR-2 nucleotide sequences used were: sense, 5′-ATGCGGAGCCCCAGCGCGGC-3′; antisense, 5′-GCCGCGCTGGGGCTCCGCAT-3′, nucleotides 148–167; GenBank accession no. U34038 (Sigma Genosis). The phosphorothioate-modified oligonucleotides are indicated by an underline. The control and antisense c-jun fully phosphorothioate-modified oligonucleotides directed against the first 18 bases of human c-jun mRNA were purchased from BIOMOL Research Laboratories. The sequences used were: control c-jun oligonucleotide, 5′-ACTGCAAAGATGGAAACG-3′; antisense c-jun oligonucleotide, 5′-CGTTTCCATCTTTGCAGT-3′.
Dominant negative activating transcription factor-2 (replacement of Thr69 or Thr71 with Ala, dn-ATF-2) was kindly provided by Dr. R. J. Davis (Howard Hughes Medical Institute) as described previously (33).
LipofectAMINE 2000 (Invitrogen Life Technologies) was used as a carrier of the oligonucleotides or dn-ATF-2, and we modified the manufacturer’s protocol for transfection (34). Briefly, HPCs were washed with prewarmed (37°C) serum-free OptiMEM medium (Invitrogen Life Technologies). The oligonucleotides (0.2 μg) or dn-ATF-2 (0.1 μg) were premixed with 2 μl of LipofectAMINE 2000 in 50 μl of OptiMEM medium for 30 min at 37°C and overlaid onto the washed cells. HPCs were incubated with either oligonucleotide-LipofectAMINE 2000 complex (for 48 h) or dn-ATF-2-LipofectAMINE 2000 complex (for 24 h) at 37°C in 5% CO2. After each transfection experiment, cells were incubated with RgpB for 1 h. The supernatants were collected for ELISA. Total RNA was extracted from the PAR-2 oligonucleotide-transfected cells for RT-PCR. Nuclear protein was extracted from the c-jun oligonucleotide-transfected cells for EMSA.
Preparation of cell lysates and Western blotting analysis
MAPKs activation was analyzed according to the method of Sarker et al. (35), with modification. In brief, 120 μl of cell suspension (5 × 105 cells/dish) were seeded onto 60-mm cell culture dishes, and cell lysates were obtained by adding 120 μl of SDS sample buffer (containing 50 mM DTT, 1 mM PMSF, and 0.5 mM Na2VO3). An equal volume of cell lysate was applied to each lane of a 12% SDS-polyacrylamide gel. After protein transfer, the nitrocellulose membrane (Schleicher & Schuell) was washed with TBST, followed by blocking with 5% nonfat milk plus 1% BSA for 1 h. After another washing, the membrane was incubated with respective Abs at 4°C overnight. After a third washing with TBST, the membrane was incubated with HRP-conjugated anti-rabbit Ab (diluted 1/3000 in TBST) for 1 h at room temperature. Finally, the membrane was washed with TBST and developed with an ECL kit (Amersham Pharmacia Biotech).
Preparation of nuclear extracts and EMSA
After treatment of cells with RgpB in the presence or absence of an inhibitor, the nuclear extract was obtained from the precipitate as described previously (36). Briefly, cells were washed with ice-cold Ca2+-Mg2+-free PBS and suspended in a buffer containing 1 M HEPES (pH 7.9), 1 M MgCl2, 1 M KCl, 100 mM PMSF, and 1 M DTT. The suspension was centrifuged at 2000 × g, and the nuclei were extracted in iced low salt buffer (1 M HEPES, 100% glycerol, 1 M MgCl2, 0.5 M EDTA, 100 mM PMSF, 1 M DTT) for 15 min. After centrifugation at 5000 × g, the supernatants were quickly mixed with high salt buffer (same as low salt buffer except 1 M KCl was added) and stored at −80°C. The DNA-binding form of AP-1 was assayed using synthetic AP-1 oligonucleotide probe 5′-GATCAGCATGAGTCACTTC-3′. The underlined letters are the core consensus sequences for the AP-1-binding sites on SP promoter (37, 38, 39). Nuclear extracts (10 μg) were incubated with 2 × 10 4 cpm/μl [α-32P]deoxy-CTP-AP-1 oligonucleotides labeled with Klenow fragment for 40 min at 30°C. A competitive reaction was performed by adding a 100-fold molar excess of an unlabeled AP-1 probe to the reaction mixture 20 min before the addition of the labeled AP-1 probe. A mutated AP-1 element probe was identical with the AP-1 oligonucleotide probe, with exception of a substitution of TG for CA in the AP-1-binding motif. The supershift assays were conducted using a specific polyclonal Ab against c-Jun/AP-1 (Santa Cruz Biotechnology). The reaction was identical with those described above, except for the addition of 2 μl of Ab to the binding assay system. The complexes were separated by 5% nondenaturing PAGE at 175 V for 70 min. The gels were dried with a gel dryer (Iwaki Glass) and visualized by exposing them to an imaging plate (BAS 1000 Mac).
All data from ELISA study were calculated from triplicate wells and are expressed as the means ± SD. The Bonferoni correction was used for multiple t test comparison, and p was determined using StatView version 5.0 for the Macintosh. p < 0.05 was considered statistically significant. At least three separate experiments were performed in each assay except ELISA.
CGRP and SP in the supernatant of HPC
To investigate the effect of RgpB on HPCs, we cultured cells in the presence of RgpB for 1 h and measured the amounts of CGRP and SP in the supernatant by ELISA. Both neuropeptides increased in proportion to RgpB concentrations to at least 200 nM and are at similar levels (Fig. 1,A). Considering the molecule of CGRP is >3 times bigger than that of SP (37 aa vs 11 aa), SP is released at ∼3 times the level of CGRP on a molar basis. Leupeptin, an RgpB inhibitor (2), reduced the RgpB-induced CGRP increase significantly (p < 0.05) and blocked the SP increase completely (Fig. 1,B). These results indicate that RgpB elicits a neuropeptide release from HPCs in a proteolytic activity-dependent manner. The reliability of the ELISA was supported by the result that the best fit lines obtained from serial dilution of a representative sample paralleled to standard best fit lines of neuropeptides (Fig. 1, C and D). The neuropeptide assays were also performed by RIA. The results obtained by RIA were almost identical with those by ELISA (data not shown), and we adopted ELISA for following experiments.
Expression of PAR-2, CGRP, and SP in HPCs and human dental pulp tissues
To study whether RgpB-induced CGRP and SP release from HPCs is mediated by PAR activation, we investigated HPC PAR mRNA expression. HPCs expressed only PAR-2 mRNA, whereas, as reported previously (32, 40, 41), HGF expressed PAR-1, -2, and -3 mRNA and A549 cells expressed mRNA of all four PARs (Fig. 2,A). Consistent with the mRNA expression result, HPCs expressed only PAR-2 protein on the cell membrane, and A549 cells expressed all of the PAR proteins (Fig. 2 B).
To determine whether HPCs produced the neuropeptides, we investigated the expression of CGRP and SP mRNA by RT-PCR. HPCs were found to express both CGRP and SP mRNA (Fig. 3,A, lanes b and d), whereas no product was found when RT-PCR was performed in the absence of reverse transcriptase, primers or RNA, indicating production of these molecules by HPCs. The presence of these neuropeptides was confirmed by immunocytochemical staining for HPCs (Fig. 3 B). From the granular pattern of the fluorescence for neuropeptides and linear fluorescence for PAR-2, it is likely that neuropeptides localize in cytoplasm granules and PAR-2 are mainly on the membrane.
To explore expression of SP, CGRP, and PAR-2 in dental pulp cells in vivo, we performed immunohistochemical staining for human dental pulp tissues. Dental pulp cells are spindle-shaped stromal cell-like, and most of them were positive for SP (Fig. 3,C, upper left), CGRP (Fig. 3,C, upper right), and PAR-2 (Fig. 3,C, middle left), whereas no cells were stained with nonspecific IgG (Fig. 3,C, middle right). The spindle-shaped cells were also stained with the IgM mAb (1B10) that reacts with cell membrane surface molecules of human fibroblasts (Fig. 3,C, lower left), whereas no cells were stained with nonspecific IgM (Fig. 3 C, lower right). These results indicate that HPCs are derived from dental pulp stroma cells that have a fibroblast-like nature and express PAR-2, SP, and CGRP.
CGRP and SP release from HPCs by PAR-2 agonists
We examined whether activation of PAR-2 with two different types of PAR-2 agonists can mimic RgpB-induced neuropeptide release. HPCs were incubated with SLIGKV, a synthetic peptide corresponding to the tethered ligand of PAR-2 (42). The peptide induced neuropeptide release in a dose-dependent manner (Fig. 4,A). Similar results were obtained with trypsin (Fig. 4,B), which activates PAR-2 by cleaving the receptor at —SKGR36↓S37LIGKV— (where ↓designates the trypsin cleavage site), exposing the N-terminal-tethered ligand SLIGKV (42, 43). RgpB appeared to be less potent than trypsin but more effective than SLIGKV in neuropeptides release from HPCs (Figs. 1,A, F44A, and B). Induction of the neuropeptide release from HPCs by PAR-2 agonists suggests that RgpB-induced CGRP and SP release occurs through PAR-2 activation.
CGRP and SP in the supernatant of PAR-2-knockout HPC cultured with RgpB
To confirm the PAR-2 mediation in RgpB-induced CGRP and SP release, HPCs were transiently transfected with two different parts of PAR-2 sense and antisense oligonucleotides. In the PAR-2 antisense oligonucleotide I-transfected HPC, no PAR-2 mRNA expression was observed when compared with the nontransfected or sense oligonucleotide-transfected mRNA (Fig. 5,A), and the RgpB-induced CGRP and SP release was reduced to ∼81% and ∼54% (p < 0.05), respectively (Fig. 5,B). In the PAR-2 sense oligonucleotide I-transfected HPC, the RgpB-induced CGRP and SP release (Fig. 5 B) was not affected. Almost identical results were obtained in experiments using PAR-2 sense or antisense oligonucleotide II transfection. Neither of these oligonucleotides exerts a cytotoxic effect on HPC when determined by MTT assay (data not shown). These results indicate that PAR-2 mediates most of the RgpB-induced CGRP release and more than one-half of the SP release in HPC.
p44/42 and p38 MAPK activation in RgpB-induced CGRP and SP release in HPC
MAPKs are involved in PAR-2-initiated signal transduction in some cells (44, 45). To study the signal transduction in RgpB-stimulated HPC, we investigated the activation of p44/42, p38 MAPK, and JNK by Western blotting analysis using Abs that specifically recognize each kinase or its phosphorylated form. The ratios of phosphorylated vs unphosphorylated p44/42 and p38 MAPK were increased by RgpB, but the ratio of JNK was not affected by the proteinase (Fig. 6,A). Phosphorylation of both kinases in HPC were detected within 3.5 min after exposure to RgpB and sustained until 60 min after exposure (Fig. 6,A). SLIGKV and trypsin also induced phosphorylation of p44/42 and p38 MAPK but not JNK, in a time course similar to that seen in the case of RgpB (Fig. 6, B and C).
To determine the contribution of p44/42 or p38 MAPK-mediated signaling to RgpB-induced CGRP and SP release, we examined the effect of kinase inhibitors. U0126, a potent inhibitor specific for MEK1/2 (the upstream kinase of p44/42), caused ∼64% and ∼57% reduction of CGRP and SP release, respectively (p < 0.05) (Fig. 6,D). SB 202190, a potent inhibitor specific for p38 MAPK, inhibited ∼71% and ∼57% of RgpB-induced CGRP and SP release, respectively (p < 0.05; Fig. 6,D). SLIGKV-induced neuropeptides release from HPC was also inhibited by U0126 or SB 202190 in similar ratios as seen in the case of RgpB (Fig. 6 E). These results suggest that phosphorylation of p44/42 and p38 MAPK contributes to more than one-half of the signaling pathway initiated by RgpB-elicited PAR-2 activation, leading to the neuropeptide release.
ATF-2 and AP-1 activation in RgpB-induced CGRP and SP release in HPC
ATF-2 and AP-1 are enhanced through phosphorylation of MAPKs (46, 47, 48, 49); we therefore investigated the signal transduction step following activated-p44/42 and-p38 MAPK in RgpB-induced neuropeptide release. RgpB-induced phosphorylated ATF-2 was blocked by SB202190 but not by U0126 or SP600125 (Fig. 7,A). The involvement of ATF-2 phosphorylation was confirmed by the result that a pronounced reduction of RgpB-induced neuropeptide release was seen in cells transfected with the kinase-inactive form of ATF-2 (dn-ATF-2) (Fig. 7 B).
The AP-1 DNA-binding activity strongly increased in nuclear extracts of RgpB-stimulated cells (Fig. 7,C). This binding activity was induced by the same concentration of nuclear protein extracted from recombinant human vascular endothelial growth factor-treated cells as shown previously (31). The AP-1 DNA-binding complex was completely displaced by the excess addition of an unlebeled AP-1 oligonucleotide probe but not with the mutated AP-1 element in the competition assay. No radioactivity staining was detected in the absence of nuclear protein (Fig. 7,C). These results indicated the specificity of the AP-1 DNA-binding activity in the nuclear extracts of RgpB-stimulated HPC. A supershift assay demonstrated that c-jun, a component of the AP-1 binding activity (34), participated in the bound AP-1 complex after the addition of anti-c-jun/AP-1 Ab to the nuclear extracts, whereas the additional of specific Ab for the p50 subunit of NF-κB transcription factor showed no effect (Fig. 7,C). To ascertain the involvement of AP-1 neuropeptide release, nuclear protein was extracted from either antisense c-jun or control c-jun oligonucleotide-transfected cells. In the antisense c-jun oligonucleotide-transfected cells, AP-1 DNA-binding was almost totally blocked when compared with the nontransfected or the control c-jun oligonucleotide-transfected cells (Fig. 7,C). The AP-1 potent inhibitor, curcumin (50) suppressed RgpB-stimulated AP-1 binding activity (Fig. 7,D). The result that activation of AP-1 by RgpB was blocked by MAPK inhibitors (SB202190, U0126) but not by a JNK inhibitor (SP600125; Fig. 7,D) supports the hypothesis that p44/42 and p38 MAPKs mediate the RgpB-induced AP-1 activation. The significant reduction of RgpB- or SLIGKV-elicited neuropeptide release by curcumin (Fig. 7, E and F), and by the antisense knockout of c-jun (data not shown) suggests that PAR-2-mediated AP-1 activation is associated with the RgpB-induced cell event. Together, these data demonstrate that RgpB stimulation of ATF-2/AP-1 activation is an important intermediate in its regulation of CGRP and SP release in HPC.
This study revealed a new activity of RgpB, induction of CGRP and SP release from HPCs (Fig. 1,A), which suggests that dental pulp inflammation is associated with the periodontal disease pathogen P. gingivalis and is modified by neuropeptides released in the dental pulp tissue. Neuropeptides are produced and released mainly by the nervous system (14, 15) and also by cells other than neurocytes, e.g., macrophages (51) and eosinophils (52). CGRP and SP are abundantly copresent in nerve fibers and perivascular plexuses in the human dental pulp tissues (22, 25). From the results that CGRP- and SP-positive cells in the dental pulp stroma and HPCs were also positive for 1B10 Ab (Fig. 3,C), it is likely that HPCs are derived not from neurocytes but from the fibroblast-like cells, which are consistent with the fact that primary pulp cells exhibited fibroblast-like nature (53). Although no study has clearly shown neuropeptide synthesis by fibroblast-like cells, the result that HPCs expressed mRNA of both CGRP and SP (Fig. 3,A) indicates production of these neuropeptides by HPCs. However, the data that the neuropeptides increased in HPC culture medium within 1 h after RgpB stimulation (Fig. 1,A) and no neuropeptides were detected in the medium before stimulation, whereas the neuropeptides are present in HPC cytoplasm granules (Fig. 3 B), suggest that the RgpB effect on HPCs is to induce release rather than production of these peptides.
The colocalization of PAR-2, CGRP, and SP and the mediation of the neuropeptides in PAR-2-induced edema have been reported in neurocytes (14). The dependency of PAR-2 activation in RgpB-elicited neuropeptide release from HPC was shown by the results that 1) HPC express only PAR-2 (Fig. 2), 2) PAR-2 agonists induce HPC neuropeptide release (Fig. 4), and 3) a pronounced reduction of HPC neuropeptide release by RgpB occurred in PAR-2 gene knockout HPC (Fig. 5 B). The RgpB-induced PAR-2 activation was also seen in neutrophils, oral epithelial cells, and osteoblast-like cells (28, 54, 55). PAR-2 agonists induce CGRP and SP release in sensory neurons (14). It seems relevant that RgpB activates HPC PAR-2 and elicits release of these neuropeptides. Lourbakos et al. (56) have reported that RgpB stimulated PAR-2, leading to neutrophil activation indicated by an increase in Ca2+ with the half-maximal response (EC50) of 49.8 ± 3.9 nM. The concentration of RgpB in the gingival crevicular fluid of patients with adult periodontitis was ∼0.1 μM (57), the highest concentration used for HPC stimulation in the present study. Accordingly, it is likely that HPC PAR-2 activation by RgpB occurs in vivo.
PAR-2 activation induces phosphorylation of p44/42 and p38 MAPK in smooth muscle cells, fibroblasts (44), keratinocyte cell line (55), and cardiomyocytes (58), and JNK was also activated in the latter two types of cells (55, 58). The result that JNK phosphorylation was not affected in HPC neuropeptide release by RgpB as well as PAR-2 agonist (Fig. 6, A–C) suggests a similarity of the PAR-2 signaling pathway in HPC to that of smooth muscle cells or fibroblasts. As found in RgpB-elicited HPC neuropeptides release (Fig. 7, C–F), eosinophil PAR-2 activation by tryptase leads to activation of the MAPKs/AP-1 pathway, causing cytokine release (59). The MAPKs/AP-1 pathway may be linked to PAR-2-mediated releasing events. ATF-2 phosphorylation induction in PAR-2-initiated signal transduction has not been reported thus far, whereas this transcription factor can mediate neuropeptides expression in the sensory neurons (60). We show for the first time that RgpB-elicited HPC neuropeptide release, in which the PAR-2-p38 MAPK-ATF-2 pathway is involved in addition to the PAR-2- p44/42 p38 MAPK and -AP-1 pathway.
CGRP and SP at a range from picomolar to micromolar have been reported to stimulate nonneuronal cells (61, 62, 63). The values of CGRP (208 pg/ml, 54 pM) and SP (182 pg/ml, 140 pM) were obtained from 50,000 RgpB-treated HPC in 0.5 ml assay buffer, and 0.1 ml of the aliquot was used for each ELISA (Fig. 1). Therefore, released CGRP and SP contents in 10,000 HPC/0.1 ml are 20.8 and 18.2 pg, respectively. Under an assumption that a HPC is a ball-like structure with a diameter of 10 μm, a HPC volume [3/4 × 3.14 × (0.0005 cm)3] is estimated to be 0.3 × 10−9 ml. Supposing that a neuropeptide evenly distributes in a HPC, CGRP and SP concentrations in a cell are 7 μg/ml (1.8 μM) and 6 μg/ml (4.6 μM), respectively. These neuropeptide concentrations should be much higher in the cell granules (Fig. 3 B). Accordingly, we can predict that the concentrations of neuropeptides released from HPCs by RgpB are high enough to exert biological activities of the neuropeptides in the dental pulp tissue. The activities of CGRP and SP to induce vasodilation, increased local blood flow, enhanced vascular permeability, and pain (14, 15, 16, 17, 18, 19) would augment these activities of bradykinin released by RgpB through activation of the kallikrein/kinin pathway (2). Taken together, the new RgpB activity to induce HPC neuropeptides release can be implicated in dental pulp inflammation disseminated from periodontitis caused by P. gingivalis infection, which exacerbates pulpitis.
We thank Ko-ichi Kawahara (Department of Laboratory and Vascular Medicine, Kagoshima University Graduate School of Medical and Dental Science, Kagoshima, Japan) for his useful advice, support, and discussion throughout this study.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Abbreviations used in this paper: RgpB, arginine-specific cysteine protease; HPC, human pulp cell; PAR, proteinase-activated receptor; SP, substance P; CGRP, calcitonin gene-related peptide; HGF, human gingival fibroblast; ATF-2, activated transcription factor-2; dn-ATF-2, dominant negative ATF-2; TFA, trifluoroacetic acid.