Identification of Residues Important for Ligand Binding of Thromboxane A2 Receptor in the Second Extracellular Loop Using the NMR Experiment-guided Mutagenesis Approach*

The second extracellular loop (eLP2) of the thromboxane A2 receptor (TP) had been proposed to be involved in ligand binding. Through two-dimensional1H NMR experiments, the overall three-dimensional structure of a constrained synthetic peptide mimicking the eLP2 had been determined by our group (Ruan, K.-H., So, S.-P., Wu, J., Li, D., Huang, A., and Kung, J. (2001) Biochemistry 40, 275–280). To further identify the residues involved in ligand binding, a TP receptor antagonist, SQ29,548 was used to interact with the synthetic peptide. High resolution two-dimensional 1H NMR experiments, NOESY, and TOCSY were performed for the peptide, SQ29,548, and peptide with SQ29,548, respectively. Through completed 1H NMR assignment and by comparing the different spectra, extra peaks were observed on the NOESY spectrum of the peptide with SQ29,548, which implied the contacts between residues of eLP2 at Val176, Leu185, Thr186, and Leu187 with SQ29,548 at position H2, H7, and H8. Site-directed mutagenesis was used to confirm the possible ligand-binding sites on native human TP receptor. Each of the four residues was mutated to the residues either in the same group, with different structure or different charged. The mutated receptors were then tested for their ligand binding activity. The receptor with V176L mutant retained binding activity to SQ29,548. All other mutations resulted in decreased or lost binding activity to SQ29,548. These mutagenesis results supported the prediction from NMR experiments in which Val176, Leu185, Thr186, and Leu187 are the possible residues involved in ligand binding. This information facilitates the understanding of the molecular mechanism of thromboxane A2 binding to the important receptor and its signal transduction.

Thromboxane A 2 (TXA 2 ) 1 is a potent platelet aggregatory and vasoconstrictive mediator (2). The function of TXA 2 is mediated by specific cell surface receptor, thromboxane A 2 receptor (TP) (3). The understanding of the structure and func-tion of TP receptor can greatly explain how the ligand binds to its receptor and initiates the following cell signaling.
TP receptor was first purified from platelet in 1989, and the cDNA of TP receptor was cloned from placenta in (4,5). Other human prostanoid receptor cDNAs have also been cloned by homology screening. All of the prostanoid receptors belong to the G-protein-coupled receptor family that share a basic seven transmembrane segments and couple to different signal transduction systems to play diverse physiological and pathological roles (6 -14). TXA 2 binds to TP receptor and triggers an increase of intracellular calcium. There were two TP receptor isoforms with different C-terminal tails, resulting from alternative splicing that the last 15 amino acids of the C terminus were replaced by 79 amino acids (15,16). The two TP receptor isoforms coupled to the same signal transduction, but endothelium expressed only the spliced form and placenta expressed both types of the TP receptors (15)(16)(17).
Based on the sequence alignment, the second extracellular loop (eLP2) and the third and seventh transmembrane domains of the prostanoid receptors are highly conserved and are proposed to be involved in ligand binding (18).  in the eLP2 of the EP3 receptor have been reported as an essential determinant of ligand selectivity (19). These results suggest that the extracellular domains of other prostanoid receptors are involved in the initial specific ligand interaction. The residues responsible for specific ligand recognition within eLP2 of the TP receptor have not been thoroughly examined. The mutations based on alignment only are controversial and will need structural information to support. The structures of the transmembrane domains of prostanoid receptors may be similar, but the specific recognition sites on extracellular domains will be different because the ligand structures are different. Thus, structural characterization of the extracellular functional domains of prostanoid receptors could help in understanding the specificities of ligand binding. In our current study, the structure of the highly conserved eLP2 has been characterized by high resolution NMR using a synthetic eLP2 peptide with constrained loop ends (1). To identify which residues make up the ligand recognition site of the receptor, SQ29,548 was added to the peptide to determine the interaction using high resolution two-dimensional 1 H NMR technique. The residues identified from the NMR approach for the interaction between the ligand and the eLP2 were further confirmed by site-directed mutagenesis. Results from the studies provided an approach of using NMR experiments guided site-directed mutagenesis for identification of the important residues of other prostanoid receptors and other G-protein-coupled receptors, which is more reasonable and close to the fact than those mutations performed only based on alignment.

EXPERIMENTAL PROCEDURES
Materials-Ethanol-d 6  Peptide Synthesis-A constrained loop peptide mimicking the sequence of the second extracellular loop of TP receptor (residues 173-193) with homocysteine added at both ends ( Fig. 1) was synthesized for NMR study using fluorenylmethoxycarbonyl-polyamide solid phase method and cyclized by the formation of disulfide bound as described previously (1, 20 -22). Briefly, the peptide was purified to homogeneity by HPLC. For the cyclization, the peptide was dissolved in 1 ml of dimethyl sulfoxide (Me 2 SO) and added to H 2 O at a final concentration of 0.02 mg/ml with pH 8.5 adjusted by triethylamine, and stirred overnight at room temperature. The cyclic peptide was then lyophilized and purified by HPLC on the C4 column.
NMR Sample Preparation-The HPLC-purified constrained loop peptide was dissolved in 20 mM sodium phosphate buffer, pH 6.0, at a final concentration of 5 mM. 1 mg of SQ29,548 was dissolved in 50 l of ethanol-d 6 and then added to 0.45 ml of sodium phosphate buffer (20 mM) containing 10% D 2 O (1). Any insoluble ligand was removed by centrifugalization. The concentration for the mixture of peptide and SQ29,548 was the same as above.
NMR Experiments-Proton NMR experiments were carried out on a VARIAN Unity Plus 500 spectrometer, which was equipped with Z-pulsed field gradient. Two-dimensional NMR experiments (DQF-COSY, TOCSY, and NOESY) were performed for eLP2 only (1), SQ29,548 only, and their mixture at 298 K. The WATERGATE method was used to suppress the signal of water. NOESY spectra were recorded with mixing time of 200 ms. TOCSY spectra was carried out with decoupling in the presence of scalar interactions spin-lock sequence with a total mixing time of 50 ms. 512 t1 increments were used in F1 with 32 scans per t1 increment and composed of 2048 complex points in F2 in all experiments. Quadrature detection was achieved in F1 by the statestime proportional phase increment method. The NMR data were processed using Felix program. All free induction decays were zero-filled to 2048 ϫ 2048 before Fourier transformation, and 0°(for DQF-COSY), 70°(for TOCSY), or 90°(for NOESY) shifted sinbell 2 window function was used in both dimensions.
PCR Cloning of the TP Receptor-PCR cloning was used to isolate the full-length cDNA of the TP receptor from human lung cDNA obtained from Invitrogen. The PCR primers were designed based on human TP cDNA with some modifications (5,23). The primer sequences were: 5Ј-CGGAATTCATGTGGCCCAACGGCAGTTC-3Ј (forward) and 5Ј-GA-AGATCTCGCTCTGTCCACTTCCTACTG-3Ј (reverse), with EcoRI and BglII sites on the ends. The full-length cDNA of the TP receptor was obtained from standard PCR amplification that was performed in 50 l of reaction mixture containing 1 unit of Vent polymerase and buffer (New England Biolabs, Beverly, MA), a 0.4 M concentration of each primer, 2 l of human lung cDNA for 30 cycles of 98°C for 1 min, 60°C for 1 min, and 72°C for 1 min. The amplified products were isolated from agarose gel and subcloned into the EcoRI/BglII sites of pAcSG. Correct cDNA sequence of the receptor was confirmed by restriction enzyme digestions and DNA sequencing analysis using Sanger dideoxy chain termination method (24).
Site-directed Mutagenesis-pAcSG-TP wild-type cDNA was first subcloned into EcoRI/XbaI sites of pcDNA3.1(ϩ) expression vector. The TP receptor mutants were then constructed using standard PCR. The procedure utilized pcDNA3.1(ϩ) vector with wild-type TP receptor as template and two synthetic oligonucleotide primers containing the desired mutation for the reaction. The primers, which were complementary to opposite strands of the template, extended during the temperature cycling of 95°C for 30 s, 53°C for 1 min 30 s, and 68°C for 13 min for a total of 25 cycles with an additional extension cycle of 68°C for 10 min using Pfu DNA polymerase from Stratagene (La Jolla, CA). The mutant products were treated with DpnI endonuclease (Stratagene) to digest the parental DNA template and confirmed by DNA sequencing. The plasmids were then prepared using Midiprep kit (Qiagen) for the transfection into COS-7 cells for expression.
Expression of TP Receptor Wild-type and Mutants in COS-7 Cells-COS-7 cells were cultured at 37°C in a humidified 5% CO 2 atmosphere in high glucose Dulbecco's modified Eagle's medium containing 10% fetal bovine serum, antibiotics, and antimycotics. The cells placed on 100-mm dishes at a density of 1.0 ϫ 10 6 were cultured overnight and then transfected with 10 g of purified cDNA of pcDNA3.1(ϩ)/TP wildtype or mutants by DEAE-dextran method. Approximately 48 h after transfection, the cells were harvested in ice-cold phosphate-buffered saline buffer and collected by centrifuge for further protein determination.
Ligand Binding Assay-Ligand binding assay for TP receptor was performed using the method as described by Tai's group (25). The cell pellets of 800 g in 25 mM Tris-HCl buffer, pH 7.4, containing 5 mM CaCl 2 were incubated with 3 nM [ 3 H]SQ29,548 (30,000 cpm, 30 Ci/mol, PerkinElmer Life Sciences) in the presence or absence of 5 M of unlabeled (cold) SQ29,548 in the 0.1-ml reaction volume with vigorous shaking at room temperature for 60 min. The reaction was then terminated by adding 1 ml of ice-cold washing buffer (25 mM Tris-HCl, pH 7.4). The unbound ligand was filtered through an ice-cold washing buffer presoaked Whatman GF/C glass filter (Whatman, Clifton, NJ) under vacuum. The radioactivity of the TP receptor-bound [ 3 H]SQ29,548 remained on the glass filter was counted in 4 ml of scintillation mixture using a Beckman ␤ counter (Fullerton, CA).

NMR Study of TP eLP2
Interacted with SQ29,548 -The constrained synthetic eLP2 peptide (Fig. 1) mimicking the second extracellular loop of the native human TP receptor, which showed the conformational change upon the interaction with the receptor antagonist SQ29,548 in the circular dichroism (CD) and fluorescent spectroscopic studies (1), was used for the two-dimensional 1 H NMR experiments. To observe the interaction between the constrained peptide and SQ29,548 at the atomic level, NOESY spectra for the eLP2 peptide, SQ29,548, and the mixture of the peptide with SQ29,548 were recorded separately under the same conditions as described under "Experimental Procedures." Resonance assignments were made using a standard approach (Table I) (26). To determine which residues of the constrained TP eLP2 interact with the ligand SQ29,548, the NOESY spectra were used to identify the intermolecular contact between the peptide and SQ29,548 (Fig. 2). The results indicate that Val 176 , Leu 185 , Thr 186 , and Leu 187 interacted with SQ29,548 and predict that the residues are involved in the TP receptor initial ligand recognition.

Recombinant TP Receptors-To test whether the residues
Val 176 , Leu 185 , Thr 186 , and Leu 187 identified by the NMR experiments using the constrained eLP2 peptide are involved in the ligand recognition for the native TP receptor, a series of recombinant protein of the human TP receptor with point mutation at the four residues were constructed. These four residues Val 176 , Leu 185 , Thr 186 , and Leu 187 were first replaced with glycine to eliminate the side chains of the residues. After transfection of the cDNA of the recombinant TP receptors into COS-7 cells, the similar expression level of the TP receptors were confirmed by Western blot (Fig. 3A). The binding of the recombinant receptors to its ligand was then performed using [ 3 H]SQ29,548, and unlabeled (cold) SQ29,548 was used as a competitive ligand (Fig. 3B). All of the mutants with glycine replacement showed decreased or lost binding activity to the receptor antagonist, SQ29,548, as compared with the TP wild type (Fig. 3B). These data indicate that the side chains of the residues Val 176 , Leu 185 , Leu 187 , and Thr 186 of the native TP receptor are important to the ligand binding via a direct contact in the ligand-binding site or an indirect induced structural effect. These results also support the conclusion based on the NMR experiments in which the four residues of the TP receptor play important roles on the receptor ligand interaction.
To further identify what determines the ligand binding, the four residues were then mutated to either the same type residues, residues with different structures, or residues with different charged (Fig. 4). Val 176 was mutated to residue Asp, Leu, or Arg. Leu 185 and Leu 187 were mutated to residue Ala, Asp, or Arg. Thr 186 was mutated to residue Ala, Arg, or Ser. The cDNAs of the mutated receptors were obtained using standard PCR approach and then transfected into COS-7 cells. The expression of the recombinant TP receptors was confirmed by Western blot (Fig. 5A). The binding of the mutated TP receptors to SQ29,548 was shown in Fig. 5. Only one recombinant TP receptor with a V176L mutation retained the binding activity to SQ29,548 (Fig. 5B). All other mutants showed significantly decreased or lost binding activity (Fig. 5, B-E). In contrast, the control mutants of the TP receptor, Y178W and S181T, which are highly conserved in all the prostanoid receptors (Fig. 6), remained full binding activities to SQ29,548 as compared with the wild-type TP receptor (Fig. 7). These results indicate that the hydrophobic side chain of Val 176 is important for the interaction with SQ29,548. For the residues Leu 185 , Thr 186 , and Leu 187 , any structural changes to the side chain will affect TP receptor binding to its antagonist.

DISCUSSION
Several alignment-based mutageneses supported the proposition that the third and seventh transmembrane domains of TP receptor were involved in ligand binding for the prostanoid receptors (18,23). However, most of the residues located on the transmembrane domain are covered by the extrarcellular domains of the receptor and the cell membrane molecules. The initial specific ligand docking residues shall be on the surface of the receptor molecule. Based on our molecular modeling studies, the TP receptor ligand, SQ29,548 (about 20 Å long), must contact the extracellular domains of the receptor before binding to the residues on the third and seventh transmembrane domains. It is also believed that the initial contact residues on the molecular surface shall be specific. This hypothesis comes from our previous study (1). The synthetic constrained eLP2 peptide of the TP receptor has been shown to change the conformation upon the addition of the receptor antagonist SQ29,548 using fluorescence spectroscopic studies (1). The interaction was also supported by the separated CD spectroscopic studies (1). These results provide the evidences that the second extracellular loop of the TP receptor is involved in the receptor ligand recognition. Studies from other groups also provided evidence to support the hypothesis of the second extracellular loop involving ligand recognition. Tai's group and Dorn II's group suggested that extracellular loops one and two were involved in ligand binding (25,27), with point mutations of several cysteine residues at these loops exhibiting no binding activity. Residues 198 -205 in the second extracellular loop of the EP3 receptor have been reported as an essential determinant of ligand selectivity (19). However, the residues responsible for the ligand recognition have not been thoroughly examined. The mutations based on alignment only are controversial and need structural information to support them. But, x-ray structure is not yet available for any mammalian G-protein-coupled receptor due to the difficulties in crystallizing the membrane-bound proteins. The structural bases of the ligand-specific recognition with the extracellular parts of the prostanoid receptors are not well known. So, characterization of the ligand recognition sites on any prostanoid receptor at the three-dimensional structural level represents a key step to reveal the specific recognition of the different prostaglandins and thromboxane by their receptors.
In this paper, to further identify the interaction and localize the residues within the eLP2 region responsible for the important ligand recognition, two-dimensional NMR spectroscopy was successfully carried out and the results showed the detail contacts between the eLP2 peptide and the receptor antagonist (Fig. 2C). The observed ligand recognition residues on the eLP2 peptide were further confirmed by the site-directed mutagenesis approach for the native TP receptor (Figs. 3 and 5). The combination of the two-dimensional NMR experiments and the NMR experiment-guided mutagenesis methods provided a quicker way to identify the important ligand recognition site of the TP receptor. This approach can be used to characterize the ligand binding to other domains of the receptor.
During our preparation of this manuscript, Le Breton's group reported a mapping of the ligand-binding site of the human TP receptor using photoaffinity labeling and site-specific antibody probes (28). The antibody screening revealed that inhibition of the amino acid region Cys 183 -Asp 193 was critical for radioligand binding and platelet aggregation. The studies provided evidences that the ligand interacts with amino acids within the second extracellular loop of the TP receptor (28). It further supported our conclusion described in this paper in which the four residues Val 176 , Leu 185 , Thr 186 , and Leu 187 within the second extracellular loop are identified as important residues for the receptor ligand recognition. In comparison, the combination of the two-dimensional NMR experiments and the NMR experiment-guided mutagenesis approach could give detailed structural information about the interaction of the receptor and ligand, which could not be achieved by other approaches, including general mutation approach, photoaffinity labeling, and site-specific antibody screening. The agreement among our conclusion with the photoaffinity labeling and sitespecific antibody investigation has further supported the reliability of the NMR experiment-based mutagenisis approach used for the identification of the ligand recognition site of the TP receptor. One of the key factors in these studies is to design a synthetic peptide with biological function. By using a constrained peptide to mimic the extramembrane loops of TP receptor, we successfully identified the ligand recognition site for the receptor.
Our identification of the important residues of TP eLP2 responsible for the contact with TP receptor ligand reported here does not exclude the other possible ligand-binding sites reported by other groups. We suspected that the ligand-docking site might differ from the final ligand-binding site, because the residues important to TP receptor ligand binding located within the transmembrane domains are conserved. The initial docking residues of the prostanoid receptors with their ligand shall be specific. The traditional alignment-based mutagenesis FIG. 6. Sequence alignment of the eLP2 of different prostanoid receptors. The highlighted letters are conserved among all of the receptors. The mutation of Y178 and S181 to W and T, which were conserved in other prostanoid receptors, were used as control mutants.
FIG. 7. Comparison of the binding activities for the control mutants of the TP receptors. As compared with the wild-type TP receptor, Y178W and S181T retained all the binding activity to its ligand, SQ29,548. approach may pick up some residues, which may not be involved in direct ligand contact, but which indirectly affect the protein activity through the change of protein conformation distantly. Nevertheless, our proton level information for identification of the TP receptor ligand recognition site on the extracellular domain will serve as a very valuable tool to characterize the structure of the TP receptor ligand docking site and understand the biological mechanism of TXA 2 binding to its receptor. In addition, it also provides great reference information to determine the ligand-docking sites for other prostanoid receptors, and understand the specific recognition among the eight different prostanoid receptors. In general, the NMR experiment-based mutagenesis approach is also suitable for identification of the ligand recognition sites for other G-protein-coupled receptors.