INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

THROMBIN RECEPTOR activation in vascular smooth muscle cells, platelets, and endothelial cells plays a vital role in the induction of mitogenesis, hemostasis, and inflammation (6). The activation of the endothelial thrombin receptor also increases endothelial permeability to albumin (8, 15), an important early step in the induction of the inflammatory response. The thrombin receptor, consisting of seven transmembrane domains (i.e., the extracellular NH₂ terminus, 3 extracellular loops, 3 intracellular loops, and a COOH terminus), is coupled to heterotrimeric GTP-binding proteins (25). The receptor is activated after thrombin proteolytically cleaves the peptide bond linking Arg-41 and Ser-42 of the receptor's NH₂ terminus. This new NH₂ terminus functions as a "tethered ligand" that binds to specific sites on the thrombin receptor and activates the receptor as depicted in Fig. 1 (25). The observation that a 14-residue synthetic peptide (SFLLRNPNDKYEPF; TRP-14) corresponding to the NH₂ terminus of the tethered ligand mimics most of thrombin's actions supports this model of receptor activation (7, 18, 23, 25).

Bahou et al. (2) showed that antibodies (Abs) directed against a 20-amino acid peptide (residues 34-52) spanning the thrombin cleavage site of the NH₂ terminus prevented the thrombin-induced increase in cytosolic Ca²⁺ concentration ([Ca²⁺]_i) in human umbilical vein endothelial cells but did not affect the TRP-14-induced rise in [Ca²⁺]_i. However, an Ab directed against the entire NH₂-terminal extension plus the first transmembrane domain (residues 1-160) abolished both thrombin- and TRP-14-induced responses, suggesting that the NH₂ terminus contains the critical recognition sequences required for receptor activation (2). Gerszten et al. (9) studied receptor binding domains by making chimeras of the NH₂ terminus of the Xenopus thrombin receptor coupled to the extracellular loop domains of the human thrombin receptor. They concluded that both the NH₂-terminal exodomain and extracellular loop 2 (residues 244-265) were required for receptor activation (9).

Thrombin increases endothelial permeability by activating the thrombin receptor and stimulating protein kinase C (PKC) (16, 17) and Ca²⁺-signaling pathways (7, 14, 15). However, we (13) and others (2, 23, 24) have observed that the synthetic tethered ligands do not reproduce all the actions of thrombin, suggesting presence of multiple thrombin receptors and signaling pathways. The goals of the present study were to use site-specific Abs to determine the thrombin receptor domains responsible for receptor activation in endothelial cells and to address whether inhibiting receptor activation by this means would influence the increase in endothelial permeability. We used polyclonal Abs directed against peptide sequences corresponding to the postulated binding sites of thrombin on the receptor's extracellular NH₂ terminus (residues 70-99 and 1-160) and of the tethered ligand on loop 1 (residues 161-178) and loop 2 (residues 244-265).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Model of thrombin receptor activation. A: receptor in the unactivated state. Extracellular NH2 terminus contains the cleavage site for thrombin and the 3 extracellular loops (L1, L2, and L3). B: after thrombin cleavage of the NH2 terminus, the generated "tethered ligand" interacts with the receptor's extracellular loop domains, leading to receptor activation. The "?" indicates the uncertainty about binding sites of the NH2 terminus with domains of the receptor.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Materials. Endothelial cell culture medium MCDB 131, fetal bovine serum (FBS), Hanks' balanced salt solution (HBSS), and Dulbecco's modified Eagle's medium (DMEM) were obtained from GIBCO Laboratories (Grand Island, NY). Trichloroacetic acid (TCA), N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), bovine serum albumin (BSA), ethylene glycol-bis(-aminoethyl ether)-N,N,N',N'-tetraacetic acid (EGTA), m-maleimidobenzoyl-N-hydroxysuccinimide ester, Freund's adjuvant, hydrocortisone, endothelial growth supplement, type II calf gelatin, fibronectin, protein A-Sepharose, acetonitrile, trifluoroacetic acid, and ionomycin were purchased from Sigma Chemical (St. Louis, MO). The other materials were as follows: ¹²⁵I (Amersham, Arlington Heights, IL), fura 2-acetoxymethyl ester (AM; Molecular Probes, Eugene, OR), tris(hydroxymethyl)aminomethane (Tris) · HCl and glycine (Bio-Rad, Hercules, CA), polycarbonate micropore filters (Nuclepore, Pleasanton, CA), and plastic cylinders (Adaps, Dedham, MA). Human -thrombin (Enzyme Research Laboratories, South Bend, IN) purified from human plasma had an activity of 5,245 U/ml.

Endothelial cell culture. The human dermal microvessel endothelial cell line (HMEC) was used (1). These cells are endothelial cells in origin because they exhibited typical cobblestone morphology when grown in monolayer culture and presence of von Willebrand factor and angiotensin converting enzyme (1). HMEC were grown in complete medium consisting of MCDB 131 supplemented with 5% FBS, 0.10% hydrocortisone, and 0.01% endothelial growth factor. The cells were passaged every 7-8 days at confluency as visualized by phase microscopy and were used at passages 20-30.

Transendothelial ¹²⁵I-labeled albumin clearance rate. We determined the diffusive flux of ¹²⁵I-labeled albumin across endothelial monolayers as the index of endothelial permeability (5). The system consisted of abluminal and luminal compartments separated by a gelatin- and fibronectin-coated polycarbonate microporous filter (13 mm diameter, 10 µm thickness, 0.8 µm pore diameter), which was glued onto the base of a 9-mm (ID) plastic cylinder. The filters were sterilized under ultraviolet light for 24 h, seeded with HMEC (10⁵ cells/filter), and grown for 3-4 days until confluent. The luminal compartment, which contained 500 µl of DMEM with 0.5% BSA, 20 mM HEPES, and ~5 µCi/ml ¹²⁵I-albumin, was fitted with a styrofoam ring and "floated" in the abluminal medium to equalize the hydrostatic pressure difference. The abluminal compartment contained 25 ml of the same DMEM, which was kept at 37°C and continuously stirred to ensure rapid mixing. Thrombin was added to the luminal chamber at final concentrations that took into account both luminal and abluminal volumes of DMEM. Abluminal samples of 400 µl were obtained at 5-min intervals for 60 min and counted for radioactivity in a gamma counter (Packard Instruments, Downers Grove, IL). At the end of the experiment, 12% TCA precipitation was used to determine percentage of free ¹²⁵I, which was used to correct for the clearance of ¹²⁵I-albumin. Clearance values are reported as the clearance rate of ¹²⁵I-albumin determined as the volume from luminal chamber cleared into the abluminal chamber per min (µl/min) based on weighted least-squares nonlinear regression analysis (BMDP Statistical Software, Berkeley, CA). Transendothelial permeability of endothelial monolayers (P_ec) was calculated using the following equation: 1/P_ec = 1/(P_ec+f)  1/P_f, where P_ec+f is the permeability of monolayer plus filter (calculated from clearance rates) and P_f is the permeability of filter alone [determined to be 3.9 × 10⁵ cm/s on the basis of a filter surface area of 0.635 cm² (5)].

[Ca²⁺]_i measurements. HMEC were seeded at 5 × 10⁴ cells/glass coverslip (25 mm diameter) and grown for 2-3 days until confluent for [Ca²⁺]_i determinations as described previously (11, 14). The cells were loaded with 2 µM fura 2-AM for 60 min at room temperature. After the cells were loaded, they were washed two times with HBSS and pretreated with thrombin receptor peptide Abs for 30 min at room temperature. The cells were then placed into a Sykes-Moore chamber (Bellco, Vineland, NJ) and positioned onto the stage of an inverted Nikon Diaphot microscope that was coupled to a Deltascan microspectrofluorometric system [Photon Technology International (PTI), Princeton, NJ]. An optically isolated group of three to four cells was alternately excited at wavelengths of 340 and 380 nm, and the emitted light was collected with a photomultiplier at 510 nm. Background autofluorescence (in the absence of fura 2) was determined at the beginning of each day's experiment and was automatically subtracted while data were collected. At the end of each experiment, the fluorescence of Ca²⁺-saturated fura 2 was obtained by adding 10 µM ionomycin, and the fluorescence of free fura 2 was obtained by adding 0.1 M EGTA. Fluorescence ratios of excitation wavelengths 340 and 380 nm and [Ca²⁺]_i values were computed with PTI software that was based on a dissociation constant of 306 nM, which corrects for measurements made at room temperature (21).

Thrombin receptor peptide synthesis. Thrombin receptor peptides corresponding to potential binding domains for the tethered ligand (Table 1) were synthesized on an Applied Systems (Foster City, CA) automated peptide synthesizer (model 430 A), using Fmoc chemistry with FastMoc protocol as suggested by the manufacturer. All reactive Fmoc amino acids were protected on the side chains, and rink resin was used for amide at the COOH terminus. The products were obtained by cleavage and deprotection with 95% trifluoroacetic acid plus the appropriate scavengers. The peptides were purified to homogeneity by preparative reverse-phase high-performance liquid chromatography (HPLC; C18 column, using a gradient of 0-60% acetonitrile and 0.1% trifluoroacetic acid). Each peptide was checked and verified by its single peak in the analytical HPLC chromatogram, amino acid composition, and mass spectrum and by NH₂-terminal sequencing.

Ab production and purification. Polyclonal antisera were raised against thrombin receptor peptides in rabbits. Peptides were conjugated to keyhole limpet hemocyanin using the heterobifunctional coupling reagent m-maleimidobenzoyl-N-hydroxysuccinimide ester. After conjugation, 300-500 µg of peptide in 250 µl were mixed with an equal volume of complete Freund's adjuvant and injected intramuscularly at four different sites into the rabbit. Blood was collected before injection to obtain preimmune serum from each rabbit. Booster injections with incomplete adjuvant were made at 2-wk intervals for 4 wk.

Fluorescence-activated cell sorter analysis. Confluent monolayers of HMEC-1 were washed, incubated in serum-free medium for 2 h, and nonenzymatically dissociated with cell dissociation medium (Sigma Chemical). The cell suspension was incubated with 20% horse serum in PBS containing 0.1% NaN₃ on ice for 30 min to block nonspecific binding and then washed two times using PBS, and 5 × 10⁵ cells were incubated for 1 h on ice with 1 mg/ml preimmune control immunoglobulin G (IgG) or 1 mg/ml AbL1, AbL2, or AbDD in 3% horse serum-0.1% NaN₃-PBS. After incubation with primary Ab, the cells were washed two times and incubated with fluorescein isothiocyanate-labeled goat anti-rabbit IgG (Sigma Chemical) at a dilution of 1 mg/ml for 30 min on ice. The cells were then washed two times in horse serum-NaN₃-PBS and once in PBS, fixed in 1% paraformaldehyde, and analyzed by a Coulter fluorescence-activated cell sorter (FACS) analyzer (Flow Cytometry Laboratory, University of Illinois, Chicago, IL).

Statistics. Statistical analysis for transendothelial clearance of ¹²⁵I-albumin was carried out using the analysis of variance test, with level of significance set at P < 0.05. All experiments were performed three to four times with four to six monolayers per experiment. A multiple-range test for comparisons with a single control (Dunnett's test) was used for analysis of the [Ca²⁺]_i responses.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Thrombin increases albumin permeability and [Ca²⁺]_i in HMEC. Activation of confluent monolayers of HMEC with human -thrombin (1-8 nM) resulted in dose-dependent increases in transendothelial ¹²⁵I-albumin clearance rates (Fig. 2). Stimulation with 8 nM thrombin caused an approximately fourfold increase in the clearance rate over control values, similar to the increase seen with 6 nM thrombin (Lum, unpublished observations). Stimulation of HMEC with 0.5 and 1 nM thrombin increased [Ca²⁺]_i to 1.9 ± 0.3 and 1.75 ± 0.1 µM, respectively (Table 2), suggesting that these thrombin concentrations elicited the maximal rise in [Ca²⁺]_i.

View larger version (32K): [in this window] [in a new window]   Fig. 2.   Effects of different concentrations of thrombin (0, 1, 4, and 8 nM) on clearance rate of 125I-labeled albumin across confluent monolayers of human dermal microvascular endothelial cells (HMEC). Values are means ± SE in µl/min; each group contained 5-7 monolayers (n = 2 separate experiments). * Difference from control group (P < 0.01).

Effects of thrombin receptor Abs on increase in permeability. The purified thrombin receptor Abs used in this study are listed in Table 1. We have previously shown by immunoblotting that AbL1 (against residues 161-178) specifically binds to the thrombin receptor of endothelial cells (26). In the present study, FACS analysis showed that AbL1, AbL2, and AbDD caused a similar right shift in relative log fluorescence compared with preimmune IgG (Fig. 3), indicating the binding of the Abs to the HMEC thrombin receptor.

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Flow cytometric analysis of binding of anti-thrombin receptor antibodies (Abs) AbL1 (B), AbDD (C), and AbL2 (D) to cell surface of HMEC. See MATERIALS AND METHODS for details [binding of AbEE to endothelial thrombin receptor has been described by Bahou et al. (2)]. A shows effects of preimmune immunoglobulin G (IgG).

View larger version (17K): [in this window] [in a new window]   Fig. 4.   Effects of anti-thrombin receptor AbL1 (300 µg/ml, A) and AbL2 (300 µg/ml, B) on thrombin-induced increases in transendothelial clearance rate of 125I-albumin across HMEC monolayers. Cells were pretreated with Abs or IgG for 30 min and stimulated with 8 nM -thrombin (-Thr). Values are means ± SE; each group contained 5-6 monolayers (n = 3 separate experiments). * Difference compared with control group (P < 0.01). + Difference compared with -Thr group (P < 0.01).

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Effects of anti-thrombin receptor AbEE on -Thr-induced increase in the clearance rate of 125I-albumin in HMEC monolayers. Cells were pretreated with 300 µg/ml AbEE or preimmune IgG for 30 min and stimulated with 8 nM -Thr. Values are means ± SE; each group contained 5-6 monolayers (n = 3 separate determinations). * Difference compared with control group (P < 0.01). + Difference compared with -Thr group (P < 0.01).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Effects of anti-thrombin receptor AbDD on -Thr-induced increase in the clearance rate of 125I-albumin in HMEC monolayers. Cells were pretreated with 300 µg/ml AbDD or preimmune IgG for 30 min and stimulated with 8 nM -Thr. Values are means ± SE; each group contained 5-6 monolayers (n = 3 separate determinations). * Difference compared with control group (P < 0.01). + Difference compared with -Thr group (P < 0.01).

Effects of thrombin receptor Abs on increase in [Ca²⁺]_i. We investigated the effects of thrombin receptor Abs on the thrombin-induced increase in [Ca²⁺]_i (Table 2). HMEC were pretreated with 300 µg/ml AbL2 and stimulated with 0.5 or 1.0 nM thrombin. AbL2 reduced the thrombin-induced increases in [Ca²⁺]_i by ~70% at either concentration of thrombin (Table 2 and Fig. 7). Control preimmune IgG pretreatment did not significantly affect the thrombin-induced increase in [Ca²⁺]_i (Fig. 7). In contrast to AbL2, AbL1 did not significantly inhibit the increase in [Ca²⁺]_i in response to 0.5 nM thrombin (Table 2 and Fig. 8).

View larger version (22K): [in this window] [in a new window]   Fig. 7.   Effects of anti-thrombin receptor AbL2 on -Thr-induced increase in cytosolic Ca2+ concentration ([Ca2+]i) in HMEC. Cells were pretreated with 300 µg/ml AbL2 or IgG for 30 min and stimulated with 1.0 nM -Thr. Control indicates absence of Ab pretreatment. Short arrow indicates time of -Thr addition; long arrows indicate the groups. Results are representative of 5 separate determinations.

View larger version (20K): [in this window] [in a new window]   Fig. 8.   Effects of anti-thrombin receptor AbL1 on -Thr-induced increase in [Ca2+]i in HMEC. Cells were pretreated with 300 µg/ml AbL1 for 30 min and stimulated with 0.5 nM -Thr. Control indicates absence of Ab pretreatment. Short arrow indicates time of -Thr addition; long arrows indicate groups. Results are representative of 4 separate determinations.

View larger version (18K): [in this window] [in a new window]   View larger version (19K): [in this window] [in a new window]   Fig. 9.   Effects of anti-thrombin receptor AbEE (A) and AbDD (B) on the -Thr-induced increase in [Ca2+]i in HMEC. Cells were pretreated with 300 µg/ml AbEE or AbDD for 30 min and stimulated with 1 nM -Thr. Control indicates absence of Ab pretreatment. Short arrows indicate time of -Thr addition. Results are representative of 3 and 5 separate determinations, respectively.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Thrombin receptor activation involves the cleavage of its extracellular extension (between Arg-41 and Ser-42) by thrombin, thereby generating a new NH₂ terminus beginning at Ser-42 (25). The newly formed NH₂ terminus functions as a tethered ligand that binds to sites on the receptor to elicit receptor activation (Fig. 1). The goals of this study were to use receptor site-specific Abs to investigate the thrombin receptor domains in endothelial cells regulating receptor activation and to determine whether the Abs capable of blocking receptor activation would prevent the thrombin-induced increase in endothelial permeability.

We used HMEC, in which we showed that the basal permeability values of HMEC monolayers were in the range for cells that we have studied previously (bovine pulmonary microvessel endothelial cells; Ref. 22). Basal transendothelial ¹²⁵I-albumin clearance rates in both cell types ranged from 0.03 to 0.06 µl/min. The calculated basal endothelial permeability of albumin was 0.13 × 10⁵ cm/s (Table 3), a value 10-fold lower than the value of 1.2 × 10⁵ cm/s for bovine pulmonary artery endothelial cells (5). Activation of HMEC by thrombin resulted in permeability increases of two- to fourfold over baseline, which is typical of the response observed in other endothelial cell types (8, 13, 16).

We observed that the Ab directed against peptide sequences corresponding to the extracellular loop 2, but not the anti-loop 1 Ab, reduced by ~70% the thrombin-induced increases in [Ca²⁺]_i and endothelial permeability. This finding suggests that the binding of the Ab to loop 2 and the resulting steric hindrance may have prevented the activation of the thrombin receptor. Therefore, loop 2 contains the important recognition sequence for receptor activation that is responsible for increasing endothelial permeability. The control IgG and AbL1 had no effect on these responses; thus the results of the AbL2 cannot be ascribed to nonspecific effects of the Ab directed against loop 2. The importance of the loop 2 domain in mediating thrombin receptor activation is consistent with the work of Gerszten et al. (9) who investigated the thrombin receptor binding domains by making chimeras of human with Xenopus thrombin receptors and studied the activation of these modified receptors using the human thrombin receptor peptide SFLLRN. They found that the construct consisting of the human loop 2 sequence had a 50% effective concentration (EC₅₀) value 30-fold lower than the construct containing human loop 1 sequences (9), indicating the importance of loop 2 in activation of the receptor.

Although the present data indicate that loop 2 is a critical domain regulating thrombin receptor activation in endothelial cells, the results also indicate that the effect of AbL2 in inhibiting the thrombin-induced Ca²⁺ and permeability responses was partial. This incomplete inhibition was not due to the low Ab-to-receptor ratio, since a lower concentration of 200 µg/ml AbL2 showed a similar degree of inhibition as an Ab concentration of 300 µg/ml (data not shown). The effect also cannot be explained by activation of additional endothelial receptors such as the proteinase-activated receptor (PAR-2) (19) because AbEE directed against the entire NH₂ terminus of the thrombin receptor fully abrogated the responses. A logical explanation for this finding is that the NH₂ terminus may interact with other receptor sites besides loop 2 that also participate in activation of the thrombin receptor.

Because both a distal domain (residues 76-93) of the NH₂ terminus and the extracellular loop 2 have been proposed to be important sites required for thrombin receptor activation (9), we studied the effect of AbDD directed against residues 70-99 of the NH₂ terminus. This Ab reduced the thrombin-induced increases in permeability and [Ca²⁺]_i by 60-70%, respectively, consistent with chimera experiments (9). The finding supports the hypothesis that cooperativity between this NH₂ domain with other receptor sites (i.e., loop 2) is required for full receptor activation. The thrombin receptor chimera containing the entire NH₂ terminus showed a 10-fold lower EC₅₀ to stimulation by the thrombin receptor peptide than the construct containing a partial NH₂ terminus of residues up to 75 (9). The EC₅₀ was lowered by 100-fold in the chimera containing both the entire human NH₂ terminus plus loop 2 (9). These observations support our findings obtained in endothelial cells using site-specific Abs that the NH₂ terminus and loop 2 are both required for mediation of thrombin receptor activation and the thrombininduced increase in endothelial cell permeability.

We showed that AbEE (directed against residues 1-160 of the NH₂ terminus) was more effective than AbDD (directed against residues 70-99) in preventing the thrombin-induced increases in [Ca²⁺]_i and endothelial permeability. AbDD inhibited 60-70% of the thrombin-induced responses, whereas AbEE fully prevented both responses. On the basis of the model for thrombin receptor activation, AbEE may block the interaction of the NH₂ terminus at multiple sites of the thrombin receptor (2, 9), since it is directed against the entire extramembranous NH₂ region of the receptor including the thrombin cleavage site. In contrast, the less effective inhibition of AbDD observed may be attributed to the fact that this Ab is directed against a portion of the NH₂ terminus that excludes the thrombin cleavage site.

Interestingly, the recent report of a second human thrombin receptor (PAR-2) (4, 12) suggests that the permeability response may be subject to regulation by activation of two distinct thrombin receptors. Both PAR-2 and the first cloned thrombin receptor (PAR-1) mediate Ca²⁺ transients and phosphoinositide hydrolysis and appear to rely on the fibrinogen-binding exosite for receptor recognition; however, they are different in their tethered ligand sequences and cleavage sites (12). Although PAR-2 is distributed in a wide range of tissues, its occurrence in the endothelium is not clear and its role in contributing to loss of endothelial barrier is not known.

The present study provides convincing evidence that the thrombin-induced increase in endothelial permeability is the consequence of thrombin receptor activation and requires the generation of signaling molecules. The results argue against a nonspecific proteolytic action of thrombin as being responsible for increasing endothelial permeability. The thrombin-induced increase in endothelial permeability is dependent in part on the increase in [Ca²⁺]_i (15), the onset of which is rapid and transient, and precedes the onset of the increase in permeability (14). The significance for a Ca²⁺ requirement may be dependent on its function as a cofactor for activation of Ca²⁺-dependent PKC, a critical kinase regulating the increase in endothelial permeability (3, 17, 22). The rise in [Ca²⁺]_i may also contribute to the force generated by the endothelial actin-myosin contractile system, which induces endothelial contraction leading to increased endothelial permeability (10, 20).

In summary, we have shown, using site-specific Abs, that domain 70-99 of the NH₂ terminus and the extracellular loop 2 of the thrombin receptor are important determinants of thrombin receptor activation and the increase in endothelial permeability. The mechanism of the thrombin-induced increase in permeability may involve cleavage of the extracellular extension between Arg-41 and Ser-42, creating a new NH₂ terminus beginning at Ser-42. Residues 70-99 may undergo a conformational change, enabling the NH₂ terminus to bind to extracellular loop 2 and signaling the increase in permeability. The present data suggest that increased endothelial permeability involves in part the interaction of the NH₂ terminus with loop 2 of the thrombin receptor. Moreover, the present data suggest a novel means of preventing the thrombin-induced increase in vascular endothelial permeability using Abs directed against specific domains of the PAR-1 thrombin receptor.