Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

A four-member family, known as proteinase-activated receptors (PAR-1-4), has been identified (Fox et al., 1997; Molino et al., 1997; Dery et al., 1998). Serine proteinases cleave an extracellular domain of the PAR, exposing a tethered ligand that undergoes a conformational change and activates the receptor. Thrombin activates PAR-1, -3, and -4, whereas PAR-2 and -4 are activated by trypsin and mast cell tryptase (Fox et al., 1997; Molino et al., 1997; Dery et al., 1998).

PARs are widely distributed in the body and have numerous effects, such as the activation of endothelial cells, vasodilatation, and initiation of inflammatory responses (Dery et al., 1998). Tryptase, a product of mast cells, and thrombin can increase intracellular calcium mobilization in human dermal fibroblasts and keratinocytes, respectively, via a PAR-dependent mechanism (Schechter et al., 1998). PAR-2 is highly expressed on the apical and basolateral membranes of enterocytes and its activation induces prostaglandin E₂ synthesis (Kong et al., 1997), an important regulator of intestinal function. However, the physiological significance of PAR-2 activation in the gut remains unclear.

Thrombin, trypsin, and tryptase can activate more than one PAR. A number of small peptides, known as PAR-activating peptides (PAR-APs) designed to mimic the sequence of the active site of the tethered ligand, have been used to specifically activate one type of PAR. Due to the complication of multiple PAR activation by serine proteinases, these PAR-APs are more useful than the proteinases themselves in the evaluation of physiological responses to individual PAR activation in vivo.

It is likely that mast cells play a role in PAR activation because they are a major source of PAR-activating enzymes and contain several other mast cell-specific serine proteinases (Kido et al., 1985; Schechter et al., 1990; Caughey et al., 1997; Lutzelschwab et al., 1997). Human mast cell tryptase can activate PAR-2 in the muscularis externa of rat colon (Corvera et al., 1997), an effect that seems to suppress colonic peristalsis. Thrombin activates cultured mouse bone marrow-derived mast cells, involving mobilization of intracellular calcium and mediator release (Razin and Marx, 1984; Pervin et al., 1985; Baranes et al., 1986), and some studies have reported that it induces histamine release from rat peritoneal mast cells (Strukova et al., 1996). We therefore hypothesized that mast cells express PARs, which could modulate mast cells function via paracrine or autocrine pathways triggered by the release of proteinases from mast cells or other cells. Our data show that mast cells express mRNA encoding PAR-1 and PAR-2. Moreover, we demonstrate that stimulation of mast cells with PAR-1AP or PAR-2AP results in -hexosaminidase release via a mechanism that seems to be distinct from PAR-1 and PAR-2.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents. Compound 48/80, calcium ionophore A23187, neuraminidase (type V from Clostridium perfringens), benzalkonium chloride, amastatin, porcine pancreatic trypsin, thrombin (human plasma), and 4-methylumbelliferyl-N-acetyl--D-glucosaminide (-hexosaminidase substrate) were purchased from Sigma-Aldrich (St. Louis, MO). Tryptase purified from human lung was obtained from Cortex Biochem (San Leandro, CA). TRIzol, M-MLV reverse transcriptase (RT), oligo-(dT)_12-18 primers, and Taq polymerase were purchased from Invitrogen (Burlington, ON, Canada). Fura-2 acetoxymethyl ester was supplied by Molecular Probes (Eugene, OR). PAR-1- and PAR-2-activating and control peptides, as outlined in Table 1, were synthesized by the Peptide Synthesis Facility (Faculty of Medicine, University of Calgary, Calgary, AB, Canada). These peptides were determined to be >95% pure by mass spectrometry and high-performance liquid chromatography. The PCR primers were synthesized by the DNA Services Laboratory (University of Alberta, Edmonton, AB, Canada). A polyclonal antiserum (B5) was raised in rabbits to a peptide corresponding to rat PAR-2 as described previously (Kong et al., 1997).

View this table: [in this window] [in a new window]   TABLE 1 PAR1 and PAR2 activating peptides The selectivity of PAR peptides used in this study, as determined in other biological systems (Kawabata et al., 1997, 1999; Vergnolle et al., 1998; Vergnolle, 1999). Reverse sequence control peptides are not always complete negative controls because in other biological systems (Kawabata et al., 1997,

Culture of Human Embryonic Kidney Cells (HEK293). HEK293 cells, shown to naturally express PAR-1 and PAR-2 (Kawabata et al., 1997), were cultured in Dulbecco's minimal medium supplemented with 10% fetal bovine serum, penicillin/streptomycin (40 U/ml), and glutamine (2 mM) (37°C, 5% CO₂) and used once they reached approximately 90% confluence. Before use, cells were rinsed free of growth medium with PBS containing 5 mM EDTA (37°C). Cells were then detached from the plates by allowing them to sit for 2 min in PBS containing 5 mM EDTA, the flask was tapped, and the disaggregated cells washed.

Animals. Male Sprague-Dawley rats were housed in the University of Alberta animal housing facility under 12-h light/dark cycles (7:00 AM-7:00 PM) and given food and water ad libitum. For some experiments, rats were sensitized with 3000 L3 larvae of Nippostrongylus brasiliensis (Befus et al., 1979). The University of Alberta Animal Ethics Committee in accordance with guidelines of the Canadian Council for Animal Care approved the study.

Harvesting and Purification of Peritoneal Mast Cells. Rats were sacrificed by exposure to high concentration of CO₂, followed by cervical dislocation and exsanguination. Peritoneal mast cells were isolated by injecting 20 ml of ice-cold HEPES-Tyrode's buffer (HTB) into the peritoneal cavity, and the abdomen was massaged for 1 min, opened, and the liquid aspirated into ice-cold polypropylene tubes (Bissonnette and Befus, 1990). Cells were washed by centrifugation (5 min, 150g, 4°C) and resuspended in 5 ml of HTB. Mast cells were enriched by centrifugation through a discontinuous density gradient of Percoll (>95% purity) (Bissonnette and Befus, 1990). Cell viability was >95% as determined by trypan blue exclusion.

RNA Isolation and Solution Phase RT-PCR. PAR-1 and PAR-2 mRNA expression by peritoneal mast cells was assessed by RT-PCR and RT-in situ PCR. For RT-PCR, total RNA was isolated from peritoneal mast cells after heparinase treatment as described previously (Gilchrist et al., 1997). One microgram of RNA was transcribed using M-MLV and oligo-dT. PCR amplification of PAR-1 and PAR-2 was performed using the following primer sequences: for 10 µM PAR-1 primers, 5' primer: 5-GATCAGCTACTACTTCTCCGGC-3 and 3' primer: 5-TGGCCGGTGCTGTTGCAACTGT-3 (731-bp PAR-1 product); for 5 µM PAR-2 primers, 5' primer: 5-CTGAAGATCTCCTACCACCTCC-3 and 3' primer: 5-ATGCACGACGAGCAGCACGTTG-3 (542-bp PAR-2 product). To ensure equal quantities of RNA were used for PCR amplification, we amplified -actin mRNA as controls using the following primers: -actin 5' primer, 5-GTGGGGCGCCCCAGGCACCA-3 and 3' primer: 5-GTCCTTAATGTCACGCACGATTTC-3 (526-bp -actin product). Annealing temperature was 56°C for PAR-1 and PAR-2 and was 64°C for -actin. The PCR cycle numbers were optimized as the fewest PCR cycles that allowed the detection of strong PAR-1 (35), PAR-2 (35), and -actin bands (30), which were not yet on the plateau of PCR amplification.

Sequencing of RT-PCR Products. Fresh amplified products for PAR-1 and PAR-2 from peritoneal mast cells were ligated into the pCR2.1 plasmid vector using the T/A cloning kit according to the manufacturer's protocol (Invitrogen, Carlsbad, CA). Sequencing was conducted using an ABI 373A automated sequencer (Applied Biosystems, Foster City, CA) by a dideoxy-chain-termination method.

RT-in Situ PCR. RT-in situ PCR was modified and performed as described previously (Nohara et al., 1999). The cells were fixed for 16 h in 10% buffered neutral formaldehyde (BDH, Toronto, ON, Canada) at 22°C, washed twice with diethyl pyrocarbonate-treated water, and then placed on silane-coated glass slides (PerkinElmer Instruments, Norwalk, CT), each with spots for test, and positive and negative controls. Cycling conditions for -actin were 5 min at 94°C and 30 cycles of 94°C for 1 min, 64°C, for 1 min, and 72°C for 1.5 min. Cycling conditions for PAR-1 and PAR-2 were 5 min at 94°C and 35 cycles of 94°C for 1 min, 56°C for 1 min, and 72°C for 1.5 min. Test spots with DNase digestion and the RT step showed target mRNA expression in the cytoplasm. Positive controls without the RT step and DNase treatment to monitor the length of protease digestion showed nuclear DNA priming, and negative controls, where the cells were treated with DNase and the RT step was eliminated, showed that no priming of genomic DNA was detectable. The primers used for RT-in situ PCR PAR-1, PAR-2, and -actin detection were the same as for solution phase RT-PCR as described above.

Flow Cytometry. Cells were incubated for 30 min at 4°C with 1/500 dilution of B5 anti-PAR-2 antibody in flow buffer (PBS with 0.1% bovine serum albumin and 0.05% sodium azide), washed twice, and then incubated for 30 min at 4°C with goat antimouse polyethylene-conjugated antibody. The cells were subsequently washed twice and analyzed with a FACScan flow cytometer (BD Biosciences, San Jose, CA). For blocking experiments, the antibody was preincubated with 50 µg/ml of the immunizing peptide for 30 min at 4°C before it was added to the cells.

Western Blotting. Concentrated A549 supernatants or lysed cells were separated on a 12% SDS-polyacrylamide gel and transferred electrophoretically (25 V, 35 min) to a polyvinylidene difluoride membrane (Bio-Rad, Hercules, CA) using the SemiDry Trans Blot system. The membranes were incubated overnight at 4°C in 5% milk (Bio-Rad) in Tris-buffered saline containing 0.02% Tween and then blocked in the same buffer containing 5% goat serum for 1 h. The membranes were probed with 1/1000 dilution of B5 anti-PAR-2. The membranes were then incubated with donkey anti-rabbit horseradish peroxidase-conjugated antibody (1:5000). Protein bands were detected by enhanced chemiluminescence.

Induction and Quantification of -Hexosaminidase Secretion. Purified peritoneal mast cells were suspended at 2.5 × 10⁵ cells/ml in HTB and stimulated at 37°C with PAR-1AP and PAR-2AP or PAR-1 and PAR-2 control peptides to induce -hexosaminidase release. Compound 48/80 (0.75 µg/ml), calcium ionophore A23187 (5 µM), and the secretory/excretory N. brasiliensis worm antigen (10 worm equivalents/ml) (Befus et al., 1979) were used as positive controls. -Hexosaminidase was measured in the supernatants and cell pellets, as described previously (Schwartz and Austen, 1980). Briefly, equal volumes of sample and -hexosaminidase substrate (1 mM 4-methylumbelliferyl-N-acetyl--D-glucosaminide dissolved in dimethyl sulfoxide and 0.2 M sodium citrate) were mixed and incubated for 2 h at 37°C. One hundred microliters of 0.2 M Tris base stopped the incubation. Samples were read using a CytoFluor 2350 fluorescent spectrophotometer at 450 nm (excitation 356 nm). Results are expressed as -hexosaminidase released as a percentage of total -hexosaminidase. In some experiments, peritoneal mast cells were treated with amastatin (0.1 or 1 µM; 30 min, 37°C), 3 µg/ml benzalkonium chloride (15 min, 37°C), or 2 U/ml neuraminidase (1 h, 37°C) in HTB before activation. These incubation times and concentrations were based on optimal values reported in literature and were further optimized for each experiment.

Measurement of Intracellular Calcium Using Fura-2 Fluorescence. The fluorescent probe fura-2 was used to measure fluxes in free intracellular calcium concentration ([Ca²⁺]_i) (Tsien, 1988). The fura-2 acetoxymethyl ester was dissolved in dimethyl sulfoxide to a concentration of 1 mM. Purified peritoneal mast cells (10⁶ cells/ml) or HEK293 (10⁶ cells/ml) were incubated with a final concentration of 1 µM fura-2 in bovine serum albumin (0.1%)-HEPES buffer for 30 min at 37°C as described previously (Kuno et al., 1993). The cells were washed twice in HEPES buffer by centrifugation (150g, 4°C, 5 min), to remove the fura-2 that had not been taken up into the cell. The cells were resuspended in HEPES buffer at 10⁶ cells/ml and kept at room temperature. Five hundred microliters (0.5 × 10⁶) of cells was then transferred to a quartz cuvette, placed in a CAF-100 Ca²⁺ analyzer (Japan Spectroscopic Co., Inc., Tokyo, Japan), and kept suspended by a magnetic stirrer. Data were obtained using Virtual Bench (National Instruments Corporation, Austin, TX) and analyzed using Excel software (Microsoft, Redmond, WA).

Statistics. All values are given as mean ± S.E.M. for the numbers of experiments noted, and statistical analyses were performed using analysis of variance and the Student's t test.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Peritoneal Mast Cells Transcribe Both PAR-1 and PAR-2 Genes. Using solution phase RT-PCR (Fig. 1A) and RT-in situ PCR (Fig. 1B), we showed that rat peritoneal mast cells express mRNA for PAR-1 and PAR-2. Figure 1A shows the expression of PAR-1 (lane 1) and PAR-2 mRNA (lane 2) by rat kidney homogenate as a positive control. Bands corresponding to PAR-1 (732 bp) and PAR-2 (543 bp) mRNA were also observed from peritoneal mast cells isolated from unsensitized rats or from rats that were sensitized with N. brasiliensis. Cloning and sequencing of the RT-PCR-amplified products confirmed that they were PAR-1 and PAR-2.

View larger version (62K): [in this window] [in a new window]   Fig. 1.   Expression of PAR-1 and PAR-2 mRNA by rat peritoneal mast cells. A, PAR-1 and PAR-2 mRNA expression by RT-PCR. Lanes 1 and 2, expression of PAR-1 and PAR-2 mRNA, respectively, in rat kidney (positive control); lanes 3 and 4, expression of PAR-1 and PAR-2 mRNA, respectively, by unsensitized peritoneal mast cells; and lanes 5 and 6, expression of PAR-1 and PAR-2 mRNA, respectively, by sensitized peritoneal mast cells. B, PAR-1 and PAR-2 mRNA expression by RT-PCR in situ. PAR-1 (panel 1) and PAR-2 (panel 4) mRNA is seen in the perinuclear cytoplasm of peritoneal mast cells. Positive controls (panels 2 and 5) and negative controls (panels 3 and 6) for PAR-1 (panels 2 and 3) and PAR-2 (panels 5 and 6) are also shown. Data are representative of three independent experiments.

PAR-AP Modulates PAR mRNA. We incubated peritoneal mast cells for 2 h with PAR-AP and with interferon- (IFN-) (800 U/ml was optimal in other systems for altering gene expression) to modulate the expression of PAR mRNA. When fewer PCR cycles (30 instead of 35) were used, no mRNA for PAR-1 or PAR-2 was detected in sham-treated peritoneal mast cells (Fig. 2). Treatment of peritoneal mast cells with PAR-1AP and PAR-1 reverse sequence control peptide for 2 h up-regulated the expression of PAR-1 mRNA to detectable levels (Fig. 2A). We quantified the up-regulation of PAR-1 mRNA by calculating the optical density ratio of the PAR-1/-actin bands and expressed this ratio as a percentage. ApfFRchACY-NH₂ (Cit) and the PAR-1 reverse sequence control peptide induced the expression of PAR-1 bands whose optical densities were approximately 38 and 55% that of the respective -actin bands. The PAR-2AP and PAR-2 reverse sequence control did not affect the expression of PAR-1 (Fig. 2A). PAR-2AP treatment for 2 h with tc-LIGRLO-NH₂, but not SLIGRL-NH₂, up-regulated PAR-2 mRNA expression, producing an optical density that was approximately 50% that of the -actin band (Fig. 2B). Moreover, the PAR-1AP and the PAR-1 reverse sequence peptide control also up-regulated PAR-2 mRNA expression, producing bands with 52 and 31% the optical intensity of the respective -actin bands (Fig. 2B). Under optimum PCR conditions, PAR-1 and PAR-2 mRNA (35 cycles, 56°C) bands were detected with 56 and 42% of the optical density of their respective -actin bands (30 cycles, 64°C) (Fig. 2C). After IFN- treatment (800 U/ml, 24 h), PAR-1 and PAR-2 expression was down-regulated. The optical densities of the PAR-1 and PAR-2 bands were reduced to approximately 2 and 11% of the -actin optical densities, respectively (Fig. 2C).

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Modulation of PAR-1 and PAR-2 mRNA expression. A, examination of FSLLRY-NH2- (40 µM), Cit- (40 µM), LSIGRL-NH2- (10 µM), SLIGRL-NH2- (10 µM), and tc-LIGRLO-NH2-mediated (10 µM) modulation of PAR-1 mRNA expression by peritoneal mast cells using RT-PCR. Cit is not significantly different from FSLLRY-NH2. B, examination of FSLLRY-NH2- (40 µM), Cit- (40 µM), LSIGRL-NH2- (10 µM), SLIGRL-NH2- (10 µM), and tc-LIGRLO-NH2-mediated (10 µM) modulation of PAR-2 mRNA expression by peritoneal mast cells using RT-PCR. C, examination of IFN--mediated modulation of PAR-1 and PAR-2 mRNA expression by peritoneal mast cells using RT-PCR. Peritoneal mast cells were treated with buffer () or with 800 U/ml IFN- (+).

Effects of Serine Proteinases and PAR-Activating Peptides on Peritoneal Mast Cell Function. In vitro treatment of peritoneal mast cells with thrombin (0.1-10 U/ml), trypsin (40 U/ml), or tryptase (0.1 to 1.5 µg/ml) failed to induce -hexosaminidase release from peritoneal mast cells (n = 4; data not shown). Therefore, we tested the effects of PAR-AP on -hexosaminidase release from peritoneal mast cells of unsensitized rats. Treatment of peritoneal mast cells with 10 µM of the PAR-1AP Cit (10 min, 37°C) significantly induced -hexosaminidase release (11.7 ± 3.7%) and this increased to about 40% at 40 µM Cit. However, another PAR-1AP, TFLLR-NH₂ (0.1-40 µM) and the reverse sequence control (inactive) peptide (FSLLRY-NH₂, 0.1-10 µM) failed to induce significant release of -hexosaminidase (Fig. 3A, n = 4-16). At 40 µM, the control peptide FSLLRY-NH₂ released low levels of -hexosaminidase, but this was significantly less potent than Cit.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   PAR-AP-mediated -hexosaminidase release. A, PAR-1AP-mediated -hexosaminidase release from peritoneal mast cells. Cit and TFLLR-NH2 were identified in other biological systems as PAR-1 AP, with FSLLRY-NH2 used as a reverse sequence control. -Hexosaminidase release is represented as the percentage of total (mean ± S.E.M.; n = 4-16; *, p < 0.05 versus spontaneous -hexosaminidase release). Spontaneous -hexosaminidase release from untreated peritoneal mast cells was 5.1 ± 0.8%; data shown corrected for spontaneous -hexosaminidase release. B, PAR-2AP-mediated -hexosaminidase release from peritoneal mast cells. tc-LIGRLO-NH2 and SLIGRL-NH2 were identified in other biological systems as PAR-2 AP, with tc-OLRGIL-NH2 used as a reverse sequence control. -Hexosaminidase release is represented as the percentage of total (mean ± S.E.M.; n = 4-16, *; p < 0.05 versus spontaneous -hexosaminidase release). Spontaneous -hexosaminidase release from untreated peritoneal mast cells was 5.1 ± 0.8%; data shown corrected for spontaneous -hexosaminidase release.

PAR-AP Treatment of Peritoneal Mast Cells Induces Dose-Dependent Changes in Free Intracellular Calcium. When peritoneal mast cells were stimulated with PAR-1AP (10-40 µM; Fig. 4A) and PAR-2AP (1-10 µM; Fig. 4B), there was a concentration-dependent increase in [Ca²⁺]_i. Similar to the -hexosaminidase release data, the PAR-2AP tc-LIGRLO-NH₂ was more potent at increasing [Ca²⁺]_i than the PAR-1AP Cit.

View larger version (23K): [in this window] [in a new window]   Fig. 4.   PAR-AP-mediate increased free [Ca2+]i levels in a dose-dependent manner. A, PAR-1AP (Cit)-mediated dose-dependent increase in [Ca2+]i as determined by fura-2 fluorescence. Peritoneal mast cells (1 × 106/ml) were loaded with 1 µM fura-2 (30 min, 37°C), and 0.5 × 106 cells were used per measurement (representative of four independent experiments). B, PAR-2AP (tc-LIGRLO-NH2)-mediated dose-dependent increase in [Ca2+]i as determined by fura-2 fluorescence. Peritoneal mast cells (1 × 106/ml) were loaded with 1 µM fura-2 (30 min, 37°C), and 0.5 × 106 cells were used per measurement (representative of four independent experiments).

Do PAR-AP and Compound 48/80 Share Signaling Pathways? Cross-desensitization of increases in [Ca²⁺]_i indicates that two stimuli activate cells via similar pathways. We used cross-desensitization to determine whether the PAR-AP activate peritoneal mast cells via pathways similar to that of compound 48/80. Stimulation of peritoneal mast cells with 48/80 (0.75 µg/ml) triggered an increase in [Ca²⁺]_i (Fig. 5A, panel 1). Subsequent stimulation with 48/80 produced no increase in [Ca²⁺]_i, indicating pathway desensitization. Similarly, stimulation with PAR-1AP (Cit, 10 µM) increased [Ca²⁺]_i and rendered peritoneal mast cells unresponsive to subsequent PAR-1AP stimulation (Fig. 5A, panel 2). Cross-desensitization experiments revealed that stimulation of peritoneal mast cells with 48/80 rendered the mast cell unresponsive to stimulation with PAR-1AP (Fig. 5A, panel 1). Moreover, stimulation of peritoneal mast cells with PAR-1AP consistently suppressed the peritoneal mast cells response to stimulation with 48/80 (Fig. 5A, panel 2). 48/80 induced an increase in [Ca²⁺]_i of approximately 200 nM (panel 1). After activation with the PAR-1AP, 48/80 consistently increased the [Ca²⁺]_i by approximately 50 nM (panel 2). Thus, peritoneal mast cells cross-desensitize to 48/80- and Cit-mediated activation.

View larger version (35K): [in this window] [in a new window]   Fig. 5.   Cross-desensitization of PAR-AP- and compound 48/80-mediated increases in [Ca2+]i levels in peritoneal mast cells. A, compound 48/80 desensitizes peritoneal mast cells to subsequent compound 48/80-mediated activation (panel 1) as determined by fura-2 fluorescence. Peritoneal mast cells (1 × 106/ml) were loaded with 1 µM fura-2 (30 min, 37°C), and 0.5 × 106 cells were used per measurement (representative of four independent experiments). B, despite desensitizing peritoneal mast cells to subsequent stimulation by itself (panel 1), compound 48/80 does not cross-desensitize peritoneal mast cells to subsequent activation by the PAR-2AP tc-LIGRLO-NH2. Peritoneal mast cells (1 × 106/ml) were loaded with 1 µM fura-2 (30 min, 37°C), and 0.5 × 106 cells were used per measurement (representative of four independent experiments). C, trypsin desensitizes peritoneal mast cells to subsequent activation by the PAR-2AP tc-LIGRLO-NH2, but does not desensitize peritoneal mast cells to subsequent compound 48/80-mediated activation (panel 3). Peritoneal mast cells (1 × 106/ml) were loaded with 1 µM fura-2 (30 min, 37°C), and 0.5 × 106 cells were used per measurement (representative of three independent experiments).

Do PAR-1AP and PAR-2AP Share Signaling Pathways? We investigated whether stimulation of peritoneal mast cells with PAR-1AP (Cit) cross-desensitized peritoneal mast cells to stimulation with PAR-2AP (tc-LIGRLO-NH₂). Stimulation of peritoneal mast cells with the PAR-1AP only partially inhibited the PAR-2AP-mediated increase in [Ca²⁺]_i by approximately 50 nM (Fig. 6, A and B). However, stimulation of peritoneal mast cells with PAR-2AP seemed to desensitize completely the PAR-1AP-mediated increase in [Ca²⁺]_i (Fig. 6, A and B).

View larger version (28K): [in this window] [in a new window]   Fig. 6.   Cross-desensitization of PAR-1AP- and PAR-2AP-mediated increases in the [Ca2+]i levels in peritoneal mast cells. A, PAR-1AP (Cit) partially cross-desensitizes peritoneal mast cells to subsequent PAR-2AP (tc-LIGRLO-NH2)-mediated activation. B, PAR-2AP (tc-LIGRLO-NH2) cross-desensitizes peritoneal mast cells to subsequent PAR-1AP (Cit)-mediated activation. Peritoneal mast cells (1 × 106 /ml) were loaded with 1 µM fura-2 (30 min, 37°C), and 0.5 × 106 cells were used per measurement (representative of four independent experiments).

Effects of Benzalkonium Chloride and Neuraminidase on PAR-AP-Mediated -Hexosaminidase Release. We studied the signaling mechanisms involved in PAR-AP stimulation of peritoneal mast cells further. Because cationic compounds can activate peritoneal mast cells in a 48/80-like manner, we studied the effects of two inhibitors of 48/80-mediated peritoneal mast cells activation on PAR-AP-mediated -hexosaminidase release. Treatment of peritoneal mast cells with benzalkonium chloride (3 µg/ml, 15 min), which is a partial agonist for the G protein-activated by 48/80 and can inhibit the 48/80-activated G_i protein, suppressed 48/80-mediated -hexosaminidase release by 88% (Fig. 7A). Moreover, benzalkonium chloride suppressed PAR-2AP-mediated (0.1 and 1 µM tc-LIGRLO-NH₂) -hexosaminidase release by 62 and 61%, respectively. We could not determine whether benzalkonium chloride treatment significantly affected PAR-1AP-mediated (10 µM Cit) -hexosaminidase release because benzalkonium chloride treatment itself induced peritoneal mast cells to release 4% -hexosaminidase (Fig. 7A). Benzalkonium chloride did not affect spontaneous or reverse sequence control peptide-mediated (10 µM) -hexosaminidase release.

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Effects of benzalkonium chloride (BAC) and neuraminidase (NA) treatment on PAR-1AP- and PAR-2AP-mediated -hexosaminidase release. A, PAR-2AP (tc-LIGRLO-NH2)-mediated, but not PAR-1AP (Cit)-mediated -hexosaminidase release was significantly inhibited after treatment of peritoneal mast cells with the Gi protein inhibitor benzalkonium chloride (3 µg/ml, 15 min, 37°C). -Hexosaminidase release is represented as the percentage of total (mean ± S.E.M.; n = 4; *, p < 0.05 versus absence of benzalkonium chloride). Spontaneous -hexosaminidase release from untreated peritoneal mast cells was 4.2 ± 1.2%; data corrected for spontaneous -hexosaminidase release. B, PAR-1AP (Cit)- and PAR-2AP (tc-LIGRLO-NH2)-mediated -hexosaminidase release was significantly inhibited after treatment of peritoneal mast cells with NA (2 U/ml, 1 h, 37°C). -Hexosaminidase release is represented as the percentage of total (mean ± S.E.M.; n = 4; *, p < 0.05 versus absence of NA). Spontaneous -hexosaminidase release from untreated peritoneal mast cells was 6.1 ± 2.6%; data corrected for spontaneous -hexosaminidase release.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Recently, it was demonstrated that PAR-1AP-mediated and thrombin-induced rat paw edema was suppressed after chronic in vivo treatment with 48/80 (Vergnolle et al., 1998). Nevertheless, the exact mechanism underlying the mast cell dependence of thrombin-mediated inflammation remains unknown. Whether mast cell activation in vivo by thrombin and PAR-1AP is a result of stimulation of PAR-1 on mast cells, or an indirect effect after PAR-1 stimulation of other cells is not known. We therefore investigated the expression of PAR and the effects of PAR-AP on rat peritoneal mast cells.

We have shown that rat peritoneal mast cells express mRNA for PAR-1 and PAR-2 (Fig. 1). Our data differ from recent findings (Nishikawa et al., 2000) that reported PAR-1, but not PAR-2 mRNA could be detected in rat mast cells. We used Sprague-Dawley rats and 35 PCR cycles, whereas Nishikawa et al. (2000) used Wistar rats, 30 PCR cycles, and a different set of primers. Because heparin contained in peritoneal mast cells inhibits the RT-PCR process (Gilchrist et al., 1997), we heparinase-treated peritoneal mast cells RNA before the RT-PCR, which Nishikawa et al. (2000) do not report doing. Furthermore, with RT-in situ PCR we confirmed that peritoneal mast cells and not the contaminating cells contained PAR-2 mRNA.

Despite the expression of PAR-1 and PAR-2 mRNA by rat peritoneal mast cells and human mast cells (D'Andrea et al., 2000), we were unable to detect PAR-2 protein, using B5 antibody. This antibody readily detects PAR-2 in tissues from a variety of species (Kong et al., 1997; Corvera et al., 1999). Our negative result could reflect the level of receptor expression, because HEK cells, which respond (calcium signaling) to PAR-1AP and PAR-2AP, have a level of receptor expression below that required for detection by B5 (Kawabata et al., 1999). Unfortunately, we were unable to acquire a suitable rat PAR-1-targeted antibody to explore in depth the presence or absence of PAR-1 protein in the mast cell population.

Consistent with the findings of others (Vergnolle et al., 1999) and a recent report that a PAR-1AP (TRAP₁₄) induces interleukin-6 release from murine peritoneal mast cells (Gordon et al., 2000), our data indicate that rat peritoneal mast cells release -hexosaminidase after activation with PAR-1AP (Cit) and PAR-2AP (tc-LIGRLO-NH₂). However, Cit only weakly stimulated -hexosaminidase release at 10 µM, a concentration thought to be optimal for activating PAR-1 specifically (Kawabata et al., 1999). We did not detect a response of peritoneal mast cells (either calcium signaling or release of -hexosaminidase) to thrombin (activates PAR-1) or trypsin and tryptase (activate PAR-2). Moreover, the structure-activity relationships of PAR-AP for either mast cell calcium signaling or mast cell -hexosaminidase release were discordant with the structure-activity relationship for these peptides in other systems (e.g., human HEK cells; Kawabata et al., 1999) or rat PAR-2-expressing KNRK cells; Al-Ani et al., 1999). Specifically, reverse-sequence peptides FSLLRY-NH₂ and tc-OLRGIL-NH₂ that only weakly activate either PAR-1 or PAR-2 in other systems (Hollenberg et al., 1997; Al-Ani et al., 1999) activated mast cell -hexosaminidase release. Furthermore, the simple tethered ligand-derived activating peptides TFLLR-NH₂ (for PAR-1) and SLIGRL-NH₂ (for PAR-2) are potent PAR-APs in other assays (Al-Ani et al., 1999; Kawabata et al., 1999) but did not activate peritoneal mast cells. Even SLIGRL in the presence of the protease inhibitor amastatin was also inactive on mast cells. This is consistent with the report that PAR-1AP (SFLLR-NH₂) and thrombin failed to induce histamine release from peritoneal mast cells (Nishikawa et al., 2000). Our data thus point to activation mechanisms for Cit and tc-LIGRLO-NH₂ in mast cells that are different from either PAR-1 or PAR-2. Thus, the role of PAR-1 and PAR-2 in regulation of mast cell function remains unclear.

Because several cationic compounds activate peritoneal mast cells by a mechanism similar to that of 48/80, we compared the mechanisms involved for activating peritoneal mast cells by PAR-APs with those of 48/80. A cross-desensitization protocol was used to compare PAR-APs with 48/80. Cross-desensitization of a calcium signal can indicate that the activating agents share a common mechanism of action (Kawabata et al., 1999). 48/80 suppressed the calcium response of peritoneal mast cells to subsequent activation with the PAR-1AP (Cit) and vice versa. Therefore, PAR-1AP (Cit) activation of peritoneal mast cells seems to occur by a 48/80-related mechanism (Fig. 8). Similarly, tc-LIGRLO-NH₂ (PAR-2AP) treatment abolished the peritoneal mast cells [Ca²⁺]_i response to subsequent 48/80 stimulation (Fig. 8). However, prior treatment with 48/80 only partially suppressed the subsequent increase in [Ca²⁺]_i mediated by tc-LIGRLO-NH₂. These results suggest not only that tc-LIGRLO-NH₂ activates a pathway similar to that of 48/80 but also that it has another distinct mechanism of action (Fig. 8). It is important to note that pretreatment of mast cells with trypsin, desensitized them to subsequent stimulation with tc-LIGRLO-NH₂, but had no effect on subsequent 48/80-mediated increases in [Ca²⁺]_i. Thus, in addition to activating a mechanism in common with 48/80, tc-LIGRLO-NH₂ activated a trypsin-sensitive pathway distinct from that activated by 48/80 (Fig. 8). The nature of this pathway remains to be determined.

View larger version (30K): [in this window] [in a new window]   Fig. 8.   Schematic representation of proposed mechanisms of action of Cit and tc-LIGRLO-NH2. GPCR, G protein-coupled receptor.

We next used cross-desensitization to assess the relationship between the putative receptor activated by tc-LIGRLO-NH₂ and the one activated by Cit. The PAR-2AP, tc-LIGRLO-NH₂, cross-desensitized the calcium response of peritoneal mast cells to subsequent activation with the PAR-1AP (Cit) and partially vice versa. Activation of peritoneal mast cells mediated via tc-LIGRLO-NH₂ and Cit therefore seems to be regulated in part by a common mechanism, distinct from either PAR-1 or PAR-2 (Fig. 8).

48/80 is believed to interact with negatively charged sialic acid residues on the mast cell, insert its hydrophobic tail into the cellular membrane, and directly activate a G protein, resulting in mast cell activation (Mousli et al., 1989; Aridor et al., 1993). Thus, we tested whether PAR-1AP (Cit) and the PAR-2 AP (tc-LIGRLO-NH₂), both of which are positively charged peptides, would interact with sialic acid residues on mast cells. Removal of sialic acid from peritoneal mast cells with neuraminidase suppressed, but did not abolish the Cit- and tc-LIGRLO-NH₂-mediated release of -hexosaminidase, indicating that sialic acid residues play a role in their action. Thus, we examined sialic acid dependence of PAR-1 and PAR-2 activation in HEK293 cells known to express these receptors. Treatment of HEK293 cells with neuraminidase did not inhibit PAR-1- or PAR-2-mediated increase in [Ca²⁺]_i, indicating that signaling by PAR-1 and PAR-2 in HEK293 cells is sialic acid-independent. This supports our hypothesis that Cit- and tc-LIGRLO-NH₂-mediated activation pathways in mast cells are due to a receptor/pathway other than PAR-1 or PAR-2 that involves sialic acid.

We investigated signaling by PAR-AP further and showed that tc-LIGRLO-NH₂ and 48/80-mediated activation of peritoneal mast cells seems to involve the activation of the same G protein. Benzalkonium chloride, which has a long hydrophobic domain similar to that of 48/80, antagonizes the effects of 48/80 by competing for and inhibiting the G protein, which otherwise is activated by 48/80 (Bueb et al., 1990; Grundemar et al., 1994). Benzalkonium chloride also suppressed -hexosaminidase release caused by tc-LIGRLO-NH₂, suggesting that tc-LIGRLO-NH₂- and 48/80-mediated signaling occurs via the same G protein (Fig. 8). However, because benzalkonium chloride induced low levels of -hexosaminidase release from peritoneal mast cells, we could not determine whether benzalkonium chloride affected Cit-mediated (PAR-1AP) -hexosaminidase release from peritoneal mast cells.

In conclusion, our data show that although rat peritoneal mast cells possess mRNA for both PAR-1 and PAR-2, and that the mRNA abundance can be regulated by several stimuli, we have no evidence for functional PAR-1 or PAR-2 on peritoneal mast cells (calcium signaling; -hexosaminidase release) either in terms of activation by proteinases or the activation profile for selective PAR-1AP and PAR-2AP. Thus, the notion that PAR plays a functional role in mast cell activation requires further investigation. Indeed, the structure-activity profiles for PAR-AP and cross-desensitization data (Cit, tc-LIGRLO-NH₂, and 48/80) suggest that mast cells possess functional PAR-AP recognition sites that are distinct from either PAR-1 or PAR-2. The sensitivity of mast cells to these PAR-AP via a non-PAR-1/non-PAR-2 mechanism poses a challenge to researchers in this field, who make use of such reagents to characterize the consequences of activating PAR in vivo. Studies are needed not only to identify the distinct PAR-AP receptor systems present on mast cells but also to develop new PAR-1AP and PAR-2AP that do not stimulate mast cells in a non-PAR-1/non-PAR-2 manner.