A Polymorphic Protease-activated Receptor 2 (PAR2) Displaying Reduced Sensitivity to Trypsin and Differential Responses to PAR Agonists*

Protease-activated receptor 2 (PAR2) is a trypsin-activated member of a family of G-protein-coupled PARs. We have identified a polymorphic form of human PAR2 (PAR2F240S) characterized by a phenylalanine to serine mutation at residue 240 within extracellular loop 2, with allelic frequencies of 0.916 (Phe240) and 0.084 (Ser240) for the wild-type and mutant alleles, respectively. Elevations in intracellular calcium were measured in permanently transfected cell lines expressing the receptors. PAR2F240S displayed a significant reduction in sensitivity toward trypsin (∼3.7-fold) and the PAR2-activating peptides, SLIGKV-NH2 (∼2.5-fold) and SLIGRL-NH2(∼2.8-fold), but an increased sensitivity toward the selective PAR2 agonist,trans-cinnamoyl-LIGRLO-NH2(∼4-fold). Increased sensitivity was also observed toward the selective PAR-1 agonist, TFLLR-NH2 (∼7-fold), but not to other PAR-1 agonists tested. Furthermore, we found that TLIGRL-NH2 and a PAR4-derived peptide,trans-cinnamoyl-YPGKF-NH2, were selective PAR2F240S agonists. By introducing the F240S mutation into rat PAR2, we observed shifts in agonist potencies that mirrored the human PAR2F240S, suggesting that Phe240 is involved in determining agonist specificity of PAR2. Finally, differences in receptor signaling were paralleled in a cell growth assay. We suggest that the distinct pharmacological profile induced by this polymorphism will have important implications for the design of PAR-targeted agonists/antagonists and may contribute to, or be predictive of, an inflammatory disease.

PAR2 is expressed on a variety of cell types (5, 16 -23), where it has been reported to have an inflammatory role. For example, the inflammatory mediators, tumor necrosis factor, interleukin-1, and lipopolysaccharide, up-regulate endothelial PAR2 expression (24), while PAR2 activation stimulates inflammatory cytokine release from keratinocytes (21) and promotes endothelial rolling and adhesion of leukocytes (26). Moreover, the administration of the PAR2-AP, SLIGRL-NH 2 , in vivo, induces an inflammatory response (27). The resulting edema occurs via neuronal PAR2 activation (28). Given that PAR2 agonists can initiate an inflammatory response, polymorphisms within PAR2 that alter receptor signaling may have important implications for disease. The mechanism whereby the tethered ligand activates PAR2 remains to be clearly established. Elegant studies with the closely related thrombin receptor (PAR1) have suggested that the tethered ligand interacts with extracellular loop 2 (ECL2), indicating that this domain is important for governing agonist specificity and receptor signaling (29 -31). Recent evidence has also implicated ECL2 of PAR2 as a critical region for agonist specificity (32), and PAR2 function can be significantly altered by mutating the corresponding residues in ECL2 that were found to be of importance in PAR1 (33). Thus, natural mutations found within ECL2 may affect a PAR's ability to respond to its respective agonists.
In the course of cloning PAR2, we discovered a PAR2 variant (PAR 2 F240S) with a phenylalanine to serine mutation at position 240 of ECL2. We sought to determine 1) the frequency of the variant in a normal Caucasian population and 2) whether the functional characteristics of the variant receptor differed from the wild-type receptor in terms of its activation by trypsin and a variety of PAR agonists.

EXPERIMENTAL PROCEDURES
Peptides and Other Reagents-All peptides were synthesized by the Peptide Synthesis Facility (University of Calgary, Calgary, Alberta, Canada). Peptides were prepared in 25 mM HEPES buffer, pH 7.4, and were standardized by quantitative amino acid analysis to confirm peptide concentration and purity. The pGEM-T-Easy vector, Sau96I, Taq polymerase, dNTPs, MgCl 2 , and 10ϫ PCR buffer were purchased from Promega (Southampton, UK). All oligonucleotides were synthesized at Oswel Laboratories (Southampton, UK); FCS, Dulbecco's modified Eagle's medium, nonenzymatic cell dissociation fluid, penicillin, streptomycin, amphotericin, sodium pyruvate, Geneticin, and PBS (without calcium and magnesium) were from Life Technologies, Inc.; Porcine pancreatic type IX trypsin (13-20,000 units/mg; molar concentrations were calculated on the basis of 20,000 units/mg), sulfinpyrazone, and calcium ionophore (A23187) were from Sigma, and pcDNA3.1(ϩ) were from Invitrogen.
Cloning of PAR2 and the PAR2 Polymorphism-Oligonucleotides targeted to sequences at the beginning of (1U-sense, 5Ј-CCA GGA GGA TGC GGA GC-3Ј) and the end of the PAR2 reading frame (1D-antisense, 5Ј-GAG GAC CTG GAA AAC TCA ATA-3Ј) were used to amplify the cDNA from a human colonic cDNA library. The PCR product was cloned into a pGEM-T Easy vector and sequenced using the dideoxynucleotide sequencing method (34). The cDNA encoding human PAR2 was subcloned into pcDNA3.1(ϩ), and sequenced. To obtain a copy of PAR2 wild type cDNA (6), site-directed mutagenesis using the QuikChange kit (Stratagene, Cambridge, UK) was performed according to the manufacturer's instructions. To establish whether the Phe 240 residue was of importance for PAR2 activation in rat PAR2, we mutated T 719 to C 719 in a rat PAR2 clone (10). The human wild-type clone and rat PAR 2 F240S receptor clones were sequenced to confirm the engineered mutations.
Restriction Fragment Length Polymorphism Analysis-Genomic DNA was extracted from the blood of 125 normal Caucasian individuals by standard techniques (35). Consent was obtained from all subjects, and ethical approval was granted by the Southampton and SouthWest Hampshire Joint Ethics Committee. PAR2 primers corresponding to nucleotides 699 -718 in extracellular loop II (T2D-sense, 5Ј-GCT CTT GGT GGG AGA CAG GT-3Ј) and nucleotides 926 -949 in extracellular loop III (T4U-antisense, 5Ј-GGC TCT TAA TCA GAA AAT AAT GCA-3Ј) were used to generate a PCR product of 250 bp from the genomic DNA samples. A Sau96I site was designed into the T2D-sense primer and was accomplished by replacing nucleotide 716 (T) with G (see primer T2D above, wherein G (716) is underlined). A control Sau96I restriction site was present 100 bp from the 3Ј-end of the PCR product. The PCR was performed in a Gene Amp 2400 PCR System with 200 ng of DNA, 150 ng of each primer, 0.2 mM dNTPs, 2.5 mM MgCl 2 , and 1 unit of Taq polymerase in a 50-l reaction volume starting at 94°C for 5 min, 35 cycles of denaturing at 94°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 30 s, with a final extension at 72°C for 5 min. The 250-bp product was digested overnight at 37°C with 6 units/reaction of Sau96I and then run on an acrylamide gel, stained in ethidium bromide, and observed by ultraviolet light.
Genomic DNA Sequencing-To confirm data obtained from the restriction fragment length polymorphism analysis, oligonucleotides corresponding to nucleotides 389 -409 in extracellular loop I (S1D-sense, 5Ј-TGA AGA TTG CCT ATC ACA TAC-3Ј) and nucleotides 926 -949 in extracellular loop III (T4U-antisense, 5Ј-GGC TCT TAA TCA GAA AAT AAT GCA-3Ј) were used to amplify human genomic DNA. The 560-bp products were purified before fluorescence-based automated cycle sequencing was performed on an ABI 377 sequencer (ABI PRISM TM Dye Primer Cycle Sequencing-21M13 FS and M13REV FS Ready Reaction Kits).
Cell Culture and Transfection-Kirsten sarcoma-transformed rat kidney epithelial cells (KNRK; American Tissue Type Culture Collection, Manassas, VA) were selected, since they have previously been employed to generate PAR2-expressing cell lines (5, 10). Semiconfluent KNRK cells in 60-mm Petri dishes were transfected using the Lipo-fectAMINE method according to the manufacturer's protocol (Life Technologies, Inc.). Transfected cells were subcloned in Geneticin (0.6 mg/ ml)-containing medium (Dulbecco's modified Eagle's medium, 5% FCS, 100 M sodium pyruvate, 100 units/ml penicillin, 100 g/ml streptomycin, and 250 ng/ml amphotericin B), and clones were initially selected by their ability to produce a calcium signal in response to trypsin and the PAR2-selective peptide tc-NH 2 . To obtain permanent cell lines, cells expressing high levels of PAR2 were isolated by FACS using the B5 anti-PAR2 polyclonal rabbit antibody (10,36). All cell lines were propagated in Geneticin-containing medium without the use of trypsin in 95% air, 5% CO 2 at 37°C. The rat PAR2 cell line has been described elsewhere (10,33). Clones with matched expression, as assessed by FACS analysis, were selected for functional studies. Reverse transcriptase-PCR (as described previously (37)) was then performed using the primers S1D and T4U on the human wild-type and PAR 2 F240S cell lines, and the 560-bp products were sequenced to confirm the single nucleotide difference (T 719 to C 719 ) between the two cell lines.
Calcium Signaling Assay-The calcium cell-signaling assay was performed as described previously (33). Briefly, cells at 90% confluence in 80-cm 2 flasks (Life Technologies, Inc.) were rinsed with PBS, lifted with nonenzymatic cell dissociation fluid, and pelleted before resuspension in 1 ml of Dulbecco's modified Eagle's medium, 10% FCS, and 0.25 mM sulfinpyrazone. To the cells, 10 l of 2.5 mg/ml Fluo-3 acetoxymethyl ester (Molecular Probes, Inc., Eugene, OR) was added prior to gentle shaking for 35 min at room temperature. Cells were then washed in PBS and resuspended in calcium assay buffer (150 mM NaCl, 3 mM KCl, 1.5 mM CaCl 2 , 10 mM glucose, 20 mM HEPES, 0.25 mM sulfinpyrazone, pH 7.4). Fluorescence measurements were performed on a Perkin-Elmer fluorescence spectrometer 650-10S, with an excitation wavelength of 480 nm and emission recorded at 530 nm. Cell suspensions (2 ml of 3 ϫ 10 5 cells/ml) in 4-ml cuvettes were stirred with a magnetic flea bar and maintained at 24°C. To establish concentration-effect curves, the signal produced by the addition of a test agonist was measured as a percentage of the fluorescence peak height yielded by the addition of 2 M calcium ionophore (A23187, Sigma).
Cell Proliferation Assay-Human wild-type PAR2-or PAR 2 F240Stransfected cells were seeded in six-well culture plates at 100,000 cells/well in fresh medium and incubated for 24 h. Medium was aspirated, and cells were washed with PBS before the addition of 2 ml of low serum medium (0.2%). Following a 24-h incubation, test agents were introduced. The test agent trypsin, was added in the absence of serum for 1 h, before a final concentration of 0.2% FCS was added. At 72 h, cells were rinsed with PBS and harvested for counting with nonenzymatic cell dissociation fluid. Cells were counted on an improved Neubauer hemocytometer (American Optics, Buffalo, NY).
Statistics-Results were analyzed for statistical significance by the paired Student's t test, taking p Ͻ 0.05 as statistically significant.

Cloning of the Human PAR2 Polymorphic
Receptor-A comparison of the sequence obtained for the polymorphic receptor with the published sequence of human PAR2 (6) revealed a transition from a thymine to a cytosine at nucleotide position 719. Sequencing of genomic DNA from a selection of Caucasian individuals confirmed the existence of the mutation (Fig. 1,  upper). Individuals homozygous for wild type PAR2 were identified by a single thymine residue at position 719 ( Fig. 1, upper (a)), while individuals homozygous for polymorphic PAR2 were identified by the presence of a cytosine at residue 719 ( Fig. 1,  upper (b)). Finally, heterozygous individuals were identified by the presence of both a thymine and a cytosine at residue 719 ( Fig. 1, upper (c)). This mutation is predicted to result in a mutant receptor (PAR 2 F240S) with an amino acid change from Phe 240 to Ser 240 . The site of the mutation is approximately 6 amino acids from the fifth putative transmembrane domain, in ECL2 ( Fig. 1, lower).
Genotyping of the PAR2 Polymorphism-Restriction fragment length polymorphism analysis and automated DNA sequencing were employed to determine the frequency of the polymorphism in a Caucasian population. The frequency of the homozygous S/S genotype was found to be relatively uncommon ( Table I). The data were found to agree with the Hardy-Weinberg equilibrium (F 2 ϩ 2FS ϩ S 2 ϭ 1 where F 2 represents homozygote F/F frequency, S 2 represents homozygote S/S fre-quency, and 2FS represents heterozygote F/S frequency), indicating that the two alleles of the PAR2 gene are segregated in a Mendelian manner. PAR 2 F240S Is Less Sensitive to the PAR2 Agonists Trypsin, SLIGKV-NH 2 , and SLIGRL-NH 2 but More Sensitive to tc-LI-GRLO-NH 2 -The concentration-effect curves for trypsin and several selective PAR2-APs are shown in Fig. 2. For the wildtype receptor (solid lines), trypsin was 3 orders of magnitude more potent than the PAR2-APs, SLIGKV-NH 2 , tc-LIGRLO-NH 2 , and SLIGRL-NH 2 . The relative rank order of potencies was as follows: trypsin ϾϾϾϾ tc-LIGRLO-NH 2 Ն SLIGRL-NH 2 Ͼ SLIGKV-NH 2 , which is in keeping with a previous report (33). In the wild type receptor, trypsin stimulated a detectable calcium elevation at 2 nM, reaching a near maximum response at 100 nM. The effective response to the PAR2-AP, tc-LIGRLO-NH 2 , was between 1 and 100 M, for SLIGKV-NH 2 from 5 to 200 M, and for SLIGRL-NH 2 from 2 to 100 M (solid lines, Fig. 2a-d, respectively). For the PAR 2 F240S receptor (dashed lines), the potencies of all the PAR2 agonists were significantly different from those for the wild-type receptor. Trypsin was nearly 4-fold less potent ( Fig. 2a and Table II), activating the PAR 2 F240S receptor from 5 to 500 nM. The parent tethered ligand peptide, SLIGKV-NH 2 , was also less potent in this system by over 2-fold, activating the PAR 2 F240S receptor from 10 to 500 M ( Fig. 2b and Table II). However, in   both the wild-type and PAR 2 F240S receptor systems, the maximal response to SLIGKV-NH 2 appeared to be lower than the maximal response obtained with the other PAR2 agonists (ϳ50 versus ϳ60% relative to calcium ionophore). Strikingly, tc-LIGRLO-NH 2 was 4-fold more potent in the PAR 2 F240S receptor ( Fig. 2c and Table II), activating PAR 2 F240S from 0.2 to 20 M compared with 1-50 M in the wild-type receptor. The potency of SLIGRL-NH 2 was reduced by a similar degree to that of SLIGKV-NH 2 (ϳ2.5-fold), activating PAR 2 F240S from 5 to 200 M ( Fig. 2 and Table II). The relative order of potencies for these agonists in the PAR 2 F240S system, which differed quantitatively from those for the wild-type receptor, was as follows: trypsin Ͼ Ͼ tc-LIGRLO-NH 2 Ͼ Ͼ SLIGRL-NH 2 Ͼ SLIGKV-NH 2 .

PAR 2 F240S Displays Increased Sensitivity to the Selective PAR1-AP TFLLR-NH 2 but Not to SFLLR-NH 2 and Cit-NH 2 -
The concentration effect curves for PAR1-APs are shown in Fig.  2, e and f. For the wild-type receptor (solid lines), the relative rank order of potencies was as follows: SFLLR-NH 2 Ͼ Cit-NH 2 Ͼ TFLLR-NH 2 as previously reported (13). SFLLR-NH 2 , Cit-NH 2 , and TFLLR-NH 2 stimulated wild-type receptor activation from 5 to 200 M, 50 to 400 M, and 100 to 800 M, respectively. In the PAR 2 F240S receptor (dashed lines), SFLLR-NH 2 and Cit-NH 2 induced a response between 5 and 100 M and between 50 and 400 M, respectively, yielding concentrationeffect curves essentially the same as those for the wild-type receptor ( Fig. 2e and Table II). Surprisingly, TFLLR-NH 2 was over 7-fold more potent on the PAR 2 F240S receptor, provoking a response from 10 to 400 M (Fig. 2f and Table II). The relative rank order of potencies for these PAR1-APs in the PAR 2 F240S receptor was SFLLR-NH 2 Ͼ TFLLR-NH 2 Ͼ Cit-NH 2 , which differed qualitatively from those for the wild-type receptor (above). Following desensitization of PAR 2 F240S by the prior addition of the PAR2-selective peptide tc-LIGRLO-NH 2 (50 M), TFLLR-NH 2 (50 M) was no longer active, confirming its action on PAR 2 F240S (data not shown). Furthermore, thrombin at 5 units/ml (a concentration shown to be maximal for PAR1 activation (13)) failed to induce a calcium response in both the wild-type and PAR 2 F240S cell lines (data not shown), indicating an absence of functional PAR1 and PAR4 in the KNRK system.
Rat PAR 2 F240S Displays Differences in Agonist Sensitivity Similar to Human PAR 2 F240S-The relative potencies for the PAR agonists tested on the rat PAR 2 F240S receptor are dis-played in Table II. In the rat PAR 2 F240S receptor, trypsin and SLIGRL-NH 2 were ϳ2and ϳ3-fold less potent compared with the rat wild-type receptor, respectively. Compared with the wild-type receptor, tc-LIGRLO-NH 2 and TFLLR-NH 2 were ϳ3fold more potent in the rat PAR 2 F240S receptor.
Cell Growth Assay-The ability of selected PAR2 agonists to inhibit cell growth in the wild-type and PAR 2 F240S cell lines was assessed. Test concentrations of agonists were selected on the basis of the data shown for calcium signaling in Figs. 2 and 3. Results are shown in Fig. 4. Trypsin was significantly more effective at inhibiting cell growth in the wild-type (28.5 Ϯ 7.5%) cell line compared with the PAR 2 F240S (4.9 Ϯ 3.7%) cell line. tc-LIGRLO-NH 2 , tc-YPGKF-NH 2 , and TLIGRL-NH 2 inhibited growth of the PAR 2 F240S cell line by 32.7 Ϯ 5.6, 40.1 Ϯ 9.7, and 52.9 Ϯ 4.0% respectively but had little effect on the wildtype cell line (Ϫ4.9 Ϯ 2.8, Ϫ7.4 Ϯ 4.1, and 6.8 Ϯ 5.7%, respectively). SFLLR-NH 2 inhibited growth in the wild-type and PAR 2 F240S cell lines by 31.6 Ϯ 4.9 and 45.1 Ϯ 6.1%, respectively. SLIGKV-NH 2 inhibited growth by 24.1 Ϯ 6.3 and 25.8 Ϯ 11.3% in the wild-type and PAR 2 F240S cell lines, respectively. Cell numbers in untreated wells increased from 1 ϫ 10 5 to 6 ϫ 10 5 over the 96-h period of the experiment but were reduced in the presence of PAR2-APs, without any evidence of cell death as assessed by trypan blue exclusion. DISCUSSION The main finding of this study was the discovery of a human PAR2 genetic polymorphism that displays differential activation in response to trypsin and other PAR agonists. To our knowledge, this is the first report of a functionally distinct polymorphic form of a protease-activated receptor. The polymorphic receptor displayed reduced sensitivity to trypsin, SLIGKV-NH 2 , and SLIGRL-NH 2 but an increase in sensitivity to the PAR2-selective agonist tc-LIGRLO-NH 2 and surprisingly also to the PAR1-selective agonist TFLLR-NH 2 . In addition, we report that TLIGRL-NH 2 and tc-YPGKF-NH 2 could be utilized as selective agonists for the polymorphic receptor. Constructing the same Phe 240 to Ser 240 mutation in rat PAR2 led to comparable findings, suggesting that Phe 240 may participate directly in regulating agonist specificity in PAR2 from a variety of species. Furthermore, differences in receptor signaling between the wild-type and polymorphic cell lines, in terms of and wild-type receptor systems The activity of each agonist in the PAR 2 F240S receptor system was expressed relative to the activity of the same agonist in the wild-type receptor. The concentration of each agonist causing a response in the PAR 2 F240S receptor was divided by the concentration of that agonist required to cause the same calcium signal (relative to A23187) in the wild-type receptor. Values are averages obtained from four points along the parallel region of the concentration-effect curves shown in Figs. 3 and 4 (rat not shown). Values greater than 1.0 designate a sensitivity that is lower in the PAR 2 F240S receptor than in the wild-type receptor. ND, not determined. elevations in intracellular calcium, were paralleled by differences observed in a cell growth assay wherein PAR2 agonists inhibited cell growth, as has been previously reported (5). We propose that the changes in responsiveness of the polymorphic receptor toward trypsin and a spectrum of other agonists will have important implications for the development of PAR2 antagonists and protease inhibitors and for the possibility that this polymorphism may play a role in disease. We confirmed that the Phe 240 to Ser 240 mutation in PAR2 was a true polymorphism by direct DNA sequencing and restriction fragment length polymorphism analysis. In addition, we characterized a population of Caucasian individuals to determine the frequency of the polymorphism. Analysis of the data using the Hardy-Weinberg equation indicated that the alleles are normally distributed and are likely to be segregated in a Mendelian fashion. Furthermore, the data suggest that individuals possessing both alleles are uncommon, representing less than 2% of the population tested. Unfortunately, due to subject confidentiality, we were unable to contact specific individuals of potential interest to explore their possible phenotypes or association with disease. Clearly, in future work, this issue deserves systematic investigation with new populations, including those with a defined disease.
Using intracellular elevations of calcium as an indicator of receptor activation, we observed that trypsin was substantially less potent at provoking a response in the polymorphic com-pared with the wild-type receptor. In agreement with the apparent reduction of receptor sensitivity toward the tethered ligand revealed by trypsin was the finding that SLIGKV-NH 2 and SLIGRL-NH 2 were also less potent in activating the polymorphic receptor. In stark contrast to the reduced sensitivity of PAR 2 F240S toward trypsin, SLIGKV-NH 2 , and SLIGRL-NH 2 was the finding that tc-LIGRLO-NH 2 displayed a marked increase in potency for activating PAR 2 F240S compared with the wild-type receptor. The main structural difference between the tc-LIGRLO-NH 2 compound and the other PAR2-APs (including the tethered ligand) is the large N-terminal aromatic transcinnamoyl group. The loss of the aromatic side chain in the Phe 240 to Ser 240 mutation may allow the trans-cinnamoyl group to dock more efficiently, thereby inducing greater receptor activation. It is important to note that the N-terminal trans-cinnamoyl group has played a key role in the development of PAR1 antagonists (38). Consequently, the data obtained with tc-LIGRLO-NH 2 may provide a potential insight into the reactivity of the polymorphic receptor toward other PAR antagonists, which also possess aromatic side chains (39).
We extended our studies to look for potential shifts in potencies for PAR1-APs in the polymorphic receptor. It is well established that the PAR1 tethered ligand sequence, SFLLR . . . can activate PAR2, while the PAR2 peptide sequence SLIGRL-NH 2 is without effect on PAR1 (11,12). Of considerable interest was the finding that TFLLR-NH 2 , originally developed as a PAR1selective agonist (11)(12)(13), showed a marked increase in potency in activating the polymorphic receptor, resulting in a shift in the rank order of relative potencies between TFLLR-NH 2 and Cit-NH 2 . However, we noted that the PAR1-APs SFLLR-NH 2 and Cit-NH 2 possessed equal potency in both the wild-type and polymorphic cell lines. These results would suggest that the PAR1-APs SFLLR-NH 2 and Cit-NH 2 probably interact with ECL2 of PAR2 in a different manner compared with the PAR2-APs. This conclusion is in keeping with previous observations where a mutant rat PAR2, containing a 15-residue sequence in ECL2, that was identical to PAR1, displayed reduced sensitivity to SFLLR-NH 2 compared with the wild-type receptor (33).
The reason for the apparent shift in the potency of TFLLR-NH 2 and not SFLLR-NH 2 in the polymorphic receptor is intriguing. In light of previous data obtained with the TFLLR-NH 2 peptide, showing a reduced potency in activating PAR2 (12,13), one would have predicted SFLLR-NH 2 to be more potent at activating the polymorphic receptor. Indeed, substitutions of the serine residue at position 1 of SLIGKV-NH 2 and SFLLR-NH 2 are not well tolerated for successful activation of PAR2 (11,12). Interestingly, substituting the serine in SLI-GRL-NH 2 with a threonine generates a peptide (TLIGRL-NH 2 ) that is relatively selective for the polymorphic receptor, having FIG. 4. Inhibition of cell growth of human wild-type and PAR 2 F240S cell lines in response to PAR agonists. Semiconfluent wild-type or PAR 2 F240S expressing cell lines were incubated in 0.2% serum-containing medium for 24 h before agonists were introduced and incubated for a further 48 h (except for trypsin, which was incubated with cells in serum-free medium for 1 h before the addition of a final concentration of 0.2% FCS). Cells were lifted and counted on a hemocytometer. The inhibition of growth caused by PAR agonists was expressed as a percentage increase of cells in the treated cultures (⌬Nt) relative to the number of cells in the control untreated cultures (⌬Nc). Inhibition (%) ϭ (1 Ϫ (⌬Nt/⌬Nc)) ϫ 100. KV-NH 2 , SLIGKV-NH 2 ; tc-NH 2 , tc-LIGRLO-NH 2 ; SF-NH 2 , SFLLR-NH 2 ; tc-KF-NH 2 , tc-YGPKF-NH 2 ; TL-NH 2 , TLIGRL-NH 2 . Results are expressed as the mean Ϯ S.E. of three or four separate experiments, each performed in triplicate. *, p Ͻ 0.05 compared with wild-type response to respective agonist. A and C, responses of human wild-type PAR2 and PAR 2 F240S to TLIGRL-NH 2 (A) and to tc-YGPKF-NH 2 (C). B and D, responses of rat wild-type PAR2 and PAR 2 F240S to TLIGRL-NH 2 (B) and to tc-YGPKF-NH 2 (D). The control exposure to SFLLR-NH 2 (ࡗ) is shown to the right of each tracing. Cells were lifted with nonenzymatic cell dissociation fluid and loaded with Fluo-3 (22 M) prior to incubation for 35 min at room temperature. Cells were challenged with the agonists shown, and responses were monitored by fluorescence spectrophotometry (excitation, 480 nm; emission, 530 nm). Responses were compared with the peak height obtained with SFLLR-NH 2 (50 M).
little effect on wild type PAR2 and no action on PAR1 (12). It would appear that the polymorphic receptor can tolerate conservative changes at position 1 of the activating peptide. Equally striking was the activity of the PAR4-derived peptide, tc-YPGKF-NH 2 , which appeared to be entirely selective for the polymorphic receptor, compared with the wild-type PAR2. These marked differences in PAR-AP-mediated receptor activation between the wild-type and polymorphic receptor confirm that these two receptors are functionally very distinct, at least in the way they recognize agonists (and by extension, probably antagonists). It is interesting to note that the rat PAR 2 F240S mutant echoed the results obtained with the human polymorph, implying that the Phe 240 residue not only serves as an important residue in agonist specificity but could also be involved in the underlying mechanism used for recognizing specific agonists, including the tethered ligand. Furthermore, the remarkable ability of the human PAR2 polymorphic receptor to be activated by peptides that have little resemblance to its own tethered ligand (e.g. tc-YGPKF-NH 2 and TFLLR-NH 2 ) raises the possibility that other small peptides found in vivo may activate this receptor, resulting in an inappropriate tissue response.
Finally, we sought to determine whether the differences in receptor activation observed in the calcium signaling assay might also be reflected in the ability of the polymorphic receptor to inhibit cell growth (5). In this regard, trypsin was less active in the polymorphic cell line, in keeping with the data obtained with the calcium signaling assay. In addition, tc-YGPKF-NH 2 , tc-LIGRLO-NH 2 , and TLIGRL-NH 2 failed to inhibit growth of the wild-type receptor cell line but had a significant effect on slowing the growth of the polymorphic receptor cell line. These data were entirely in accord with the structure-activity relationships for the calcium signaling by these agents. However, SLIGKV-NH 2 appeared to inhibit cell growth to similar degrees in both cell lines, suggesting that the relatively small differences in receptor activation between the two cell lines observed with the calcium assay are such that they cannot be observed by the less sensitive cell growth assay. Nevertheless, the data obtained from the cell growth assay demonstrated that this polymorphism significantly reduces the ability of PAR2 agonists to regulate cell growth and presumably to affect other PAR2-mediated cell responses.
Accumulating evidence suggests that PAR2 plays a key role in the inflammatory process and a protective role in the airways (25,27,28,40). Currently, little is known about the pathophysiological roles of PAR2, and a better understanding will depend on the development of selective PAR2 antagonists. To date, only one receptor-selective nonpeptide antagonist has been developed for PAR1 (39), and as yet PAR2 antagonists are still unavailable. Considering the differential activation of the polymorphic PAR2 receptor by a variety of PAR agonists, we suggest that this receptor will need to be considered when designing antagonists for any of the PARs cloned to date. Whether this PAR2 polymorphism plays a role in disease has yet to be determined and merits further study.
In conclusion, we have found a polymorphic variation of the human PAR2 receptor at amino acid 240, where Phe (wild type) or Ser can be found. This variation results in significant differential receptor activation by trypsin and PAR agonists. Such variation may represent a genetic basis for interindividual differences in disease susceptibility, phenotype, or response to therapeutic agents targeting PARs.