Importance of the Amino Terminus in Secretin Family G Protein-coupled Receptors INTRINSIC PHOTOAFFINITY LABELING ESTABLISHES INITIAL DOCKING CONSTRAINTS FOR THE CALCITONIN RECEPTOR*

The calcitonin receptor is a member of the class B family of G protein-coupled receptors, closely related to secretin and parathyroid hormone receptors. Although mechanisms of ligand binding have been directly ex-plored for those receptors, current knowledge of the molecular basis of calcitonin binding to its receptor is based only on receptor mutagenesis. In this work we have utilized the more direct approach of photoaffinity labeling to explore spatial approximations between distinct residues within calcitonin and its receptor. For this we have developed two human calcitonin analogues incorporating a photolabile p -benzoyl- L -phenylalanine residue in the mid-region and carboxyl-terminal half of the peptide in positions 16 and 26, respectively. Both probes specifically bound to the human calcitonin receptor with high affinity and were potent stimulants of cAMP accumulation in calcitonin receptor-bearing human embryonic kidney 293 cells. They covalently labeled the calcitonin receptor in a saturable and specific manner. Further purification, deglycosylation, specific chemical and enzymatic cleavage, and sequencing of labeled wild type and mutant calcitonin receptors identified the sites of labeling for the position 16 and 26 probes as receptor residues Phe 137 and Thr 30 , respectively. Both were within the extracellular amino terminus of the calcitonin receptor, with the former adjacent to the first transmembrane segment exposure to x-ray film with intensifying screens at (cid:3) 80 °C. Aliquots of affinity-labeled receptor and relevant receptor fragments were deglycosylated with endoglycosidase F, as described previously (23). Radiochemical Sequencing— For this, the purified fragment from CNBr cleavage of the wild type human calcitonin receptor labeled with the Bpa 16 or from cleavage of the S27M/M48I/M49I mutant receptor labeled with the Bpa 26 probe was coupled to N -(2-aminoethyl-1)-3- aminopropyl glass beads (Sigma) through the sulfhydryl side chain of Cys residues. Cycles of Edman degradation were repeated manually in a manner that has been previously reported in detail (26), and the radioactivity released in each cycle was quantified in a (cid:3) -spectrometer. Statistical Analysis— All observations were repeated at least three times in independent experiments and are expressed as the means (cid:4) S.E. Binding curves were analyzed and plotted using the nonlinear regression analysis routine for radioligand binding in the Prism soft- ware package (GraphPad Software, San Diego, CA). Binding kinetics was determined by analysis with the LIGAND program of Munson and Rodbard (27). transmembrane domains of the receptor, and (iii) the carboxyl-terminal region of the (residues interacts other extracellular loop domains. The covalent attachment of calcitonin residue 26 to receptor residue Thr 30 within the ami-no-terminal tail of the calcitonin receptor may suggest some differences from this model. To further test this hypothesis by photoaffinity labeling studies, development of novel probes incorporating photolabile residues at their amino termini is also needed.

The calcitonin receptor is a member of the class B family of G protein-coupled receptors, closely related to secretin and parathyroid hormone receptors. Although mechanisms of ligand binding have been directly explored for those receptors, current knowledge of the molecular basis of calcitonin binding to its receptor is based only on receptor mutagenesis. In this work we have utilized the more direct approach of photoaffinity labeling to explore spatial approximations between distinct residues within calcitonin and its receptor. For this we have developed two human calcitonin analogues incorporating a photolabile p-benzoyl-L-phenylalanine residue in the mid-region and carboxyl-terminal half of the peptide in positions 16 and 26, respectively. Both probes specifically bound to the human calcitonin receptor with high affinity and were potent stimulants of cAMP accumulation in calcitonin receptor-bearing human embryonic kidney 293 cells. They covalently labeled the calcitonin receptor in a saturable and specific manner. Further purification, deglycosylation, specific chemical and enzymatic cleavage, and sequencing of labeled wild type and mutant calcitonin receptors identified the sites of labeling for the position 16 and 26 probes as receptor residues Phe 137 and Thr 30 , respectively. Both were within the extracellular amino terminus of the calcitonin receptor, with the former adjacent to the first transmembrane segment and the latter within the distal amino-terminal tail of the receptor. These data are consistent with affinity labeling of other members of the class B G protein-coupled receptors using analogous probes and may suggest a common ligand binding mechanism for this family.
Calcitonin, a hypocalcemic peptide hormone, is secreted from the thyroid gland in response to elevations in serum calcium levels. Its hypocalcemic effect is mediated by inhibition of bone resorption by osteoclasts and enhancement of renal calcium excretion. These actions are important for its widespread clinical use for treatment of bone disorders, including Paget's disease, osteoporosis, and hypercalcemia of malignancy (1,2).
The calcitonin peptide consists of 32 amino acids, with a disulfide bond between residues 1 and 7 which is conserved among all species and that is believed to be critical for its agonist activity. The amino acid sequence and biological potency of calcitonin vary considerably from species to species, but the integrity of the disulfide bond and the carboxyl-terminal proline-amide are necessary for full biological activity (1)(2)(3). The disulfide bond can be replaced by other covalent bonds that lead to improved biological stability while retaining full potency (1). The sequence within the amino-terminal loop region is highly conserved in a variety of species but demonstrates divergence in the rest of the sequence (4). Like secretin and peptides for other members of class B G protein-coupled receptor family, the amino-terminal region of calcitonin contains key determinants for receptor agonist selectivity, whereas the carboxyl-terminal region contains determinants for high affinity binding. Progressive truncation of residues in the disulfide bond-looped domain leads first to partial and then to antagonist peptides (1,2,5,6).  tend to form an amphiphilic ␣-helical structure that is important for high affinity binding (1,2,7). The calcitonin receptor is closely related to the secretin and parathyroid hormone (PTH) 1 receptors in the class B family of the G protein-coupled receptor superfamily, also having a long structurally unique amino-terminal domain that contains six conserved Cys residues. It shares ϳ30% identity with the secretin receptor and 32% with the PTH 1 receptor. The human calcitonin receptor has three isoforms resulting from alternative mRNA splicing. Isoform I has 490 amino acids, including a 22-residue signal sequence and a 16-residue insert in the predicted first intracellular loop domain (residue 175-190) that is absent in isoforms II and III. Apart from this 16-amino acid insert, isoform III also has the first 47 residues missing at the receptor amino terminus.
Knowledge of the molecular basis of ligand binding is important for structure-based drug design. At the present time, our understanding of the mechanism of calcitonin binding to its receptor is based predominantly on limited chimeric receptor studies (8 -10). However, currently there is no working model to predict how the two molecules might interact. In this work, we attempt to establish initial constraints that will contribute to the development of a model for the interaction of calcitonin with its receptor. With our success with the secretin receptor (11)(12)(13)(14)(15)(16)(17), we use the more direct and powerful approach of photoaffinity labeling. For this we have developed two photolabile radioiodinatable agonist probes by incorporating a photolabile residue, p-benzoyl-L-phenylalanine (Bpa), into the midregion of the human calcitonin peptide in position 16 and into the carboxyl-terminal half of the ligand in position 26. Both probes bound to the human calcitonin receptor specifically and with high affinity and efficiently covalently labeled the receptor. By sequential targeted enzymatic and chemical fragmentation reactions, the ligand binding region for the position 16 probe was localized to a domain within the amino terminus of its receptor adjacent to the first transmembrane domain, whereas that for the position 26 probe was localized within the distal amino terminus of the receptor. Using radiochemical Edman degradation sequencing, the specific residues labeled by these probes were identified as Phe 137 and Thr 30 , respectively. These represent the first experimentally derived residue-residue approximations between calcitonin-like agonists and this receptor and should be very helpful for docking this ligand in a molecular model.

EXPERIMENTAL PROCEDURES
Materials-Human calcitonin was purchased from Bachem (Torrance, CA). Cyanogen bromide (CNBr) and solid phase oxidant N-chlorobenzenesulfonamide (IODO-BEAD) were purchased from Pierce. Endoproteinase Lys-C was from Calbiochem. Endoproteinase F (Endo F) was produced in our laboratory (18). All other reagents were analytical grade.
Peptide Synthesis-The probes, human (Ile 8 ,Bpa 16 ,Arg 18 )calcitonin (Bpa 16 analogue or probe) and (Ile 8 ,Arg 18 ,Bpa 26 )calcitonin (Bpa 26 analogue or probe), were designed to contain a photolabile Bpa in position 16 or 26, respectively, for covalent labeling of the calcitonin receptor. Both probes contained a naturally occurring Tyr residue in position 12 as the site for radioiodination and an Ile residue in the position of Met 8 to eliminate a site for potential oxidative damage during radiolabeling. Additionally, Lys 18 was replaced with an Arg to facilitate the specific digestion of the labeled receptor without the cleavage of the probes themselves (Fig. 1). These changes were demonstrated to be well tolerated without resulting in substantial loss of biological activity of the ligand (19). Both probes were synthesized by manual solid-phase techniques and purified to homogeneity by reversed-phase HPLC using techniques that were previously described (20). The expected molecular masses of the probes were verified by matrix-assisted laser desorption/ ionization-time of flight mass spectrometry.
Radioiodination-Fifteen g of the synthetic probes ((Ile 8 ,Bpa 16 , Arg 18 )calcitonin or (Ile 8 ,Arg 18 ,Bpa 26 )calcitonin) were radioiodinated oxidatively with Na 125 I upon exposure to an IODO-BEAD for 15 s and purified by reversed-phase HPLC to yield specific radioactivities of 2000 Ci/mmol (20). In the same way radioiodination of the natural human calcitonin was performed to produce a radioligand for calcitonin receptor binding.
Receptor Preparations-The receptor-bearing human embryonic kidney 293 (HEK293) cell line stably expressing the human calcitonin isoform II receptor (HEK293-CTR) was provided by GlaxoSmithKline and was used as the source of receptors for the current study. Cells were cultured at 37°C in a 5% CO 2 environment on Falcon tissue culture plasticware in Dulbecco's modified Eagle's medium supplemented with 5% fetal clone-2 (Hyclone Laboratories, Logan, UT). Cells were passaged twice a week and lifted mechanically before use.
Development of new calcitonin receptor mutants was necessary for the current study. One of these incorporated an additional site for CNBr cleavage in a key position for localization of the site of labeling for the Bpa 26 probe. This represented mutation of Ser 27 of the calcitonin recep-tor to Met (S27M). A triple-mutant calcitonin receptor construct was developed to introduce a new CNBr cleavage site while simultaneously eliminating a pair of naturally occurring Met residues, representing Ser 27 to Met, Met 48 to Ile, and Met 49 to Ile (S27M/M48I/M49I) receptor mutant. This was used for radiochemical sequencing to identify the site of labeling with the Bpa 26 probe. In addition, two more calcitonin receptor constructs were generated that included Phe 137 to Ala (F137A) and Thr 30 to Ala (T30A), each representing mutation of the site of labeling by the Bpa 16 and Bpa 26 probes, respectively. All above constructs were prepared using an oligonucleotide-directed approach with the QuikChange TM site-directed mutagenesis kit from Stratagene (La Jolla, CA). They were subcloned into the eukaryotic expression vector, pcDNA3 (Invitrogen), and their sequences were verified by direct DNA sequencing (21). All mutant calcitonin receptor constructs were expressed transiently in COS cells (American Type Cell Collection, Manassas, VA) after transfection using a modification of the DEAEdextran method (22). These cells were harvested mechanically 72 h after transfection. Receptor-enriched plasma membranes were prepared from the above stable HEK293 cell line and transiently transfected COS cells using methods that we previously reported (23).
Ligand Binding-Receptor binding of calcitonin, the Bpa 16 analogue, and the Bpa 26 analogue was characterized in a standard assay using membranes from the HEK293-CTR cell line. Membranes (ϳ10 g) were incubated with a constant amount of radioligand, 125 I-calcitonin (3-5 pM), in the presence of increasing concentrations of non-radiolabeled calcitonin or the Bpa 16 analogue or the Bpa 26 analogue (0 -1 M) for 1 h at room temperature in Krebs-Ringer-HEPES medium (25 mM HEPES, pH 7.4, 104 mM NaCl, 5 mM KCl, 1 mM KH 2 PO 4 , 1.2 mM MgSO 4 , 2 mM CaCl 2 , 1 mM phenylmethylsulfonyl fluoride, 0.01% soybean trypsin inhibitor) containing 0.2% bovine serum albumin. Bound and free radioligand were separated using a Skatron cell harvester (Molecular Devices, Sunnyvale, CA) with glass fiber filtermats that had been soaked in 0.3% Polybrene for 1 h, and bound radioactivity was quantified in a ␥-spectrometer. Nonspecific binding was determined in the presence of 1 M calcitonin and represented Ͻ20% of total binding.
Biological Activity Assay-The agonist activities of the Bpa 16 and Bpa 26 analogues were studied for stimulation of cAMP in the HEK293-CTR cells using a competition binding assay (Diagnostic Products Corp., Los Angeles, CA). Cells were stimulated with increasing concentrations of calcitonin or the Bpa 16 or Bpa 26 analogue at 37°C for 30 min, and the reactions were stopped by adding ice-cold perchloric acid. After adjusting the pH to 6 with KHCO 3 , cell lysates were cleared by centrifugation at 3000 rpm for 10 min, and the supernatants were used in the assay as previously described (24). Radioactivity was quantified by scintillation counting in a Beckman LS6000.
Photoaffinity Labeling Studies-For covalent labeling studies, plasma membranes from receptor-bearing HEK293-CTR cells containing ϳ50 g of protein were incubated with ϳ0.1 nM 125 I-(Ile 8 ,Bpa 16 ,Arg 18 )calcitonin or 125 I-(Ile 8 ,Arg 18 ,Bpa 26 )calcitonin in the presence of increasing concentrations of calcitonin (0 -1 M) for 1 h at room temperature before photolysis for 30 min at 4°C in a Rayonet photochemical reactor (Southern New England Ultraviolet, Hamden, CT) equipped with 3500-Å lamps. To scale up receptor purification, a larger amount of membranes (ϳ150 -200 g) was incubated with each radiolabeled probe (ϳ0.5 nM) in the absence of competing calcitonin. After photolysis, membranes were washed, pelleted, solubilized in SDS sample buffer, and applied to a 10% SDS-polyacrylamide gel for electrophoresis (25). Radiolabeled bands were detected by autoradiography.
Peptide Mapping-Radioactive receptor bands were cut out from the gel and homogenized in a Dounce homogenizer in water followed by lyophilization and ethanol precipitation. Purified materials were used for chemical or enzymatic cleavage experiments. CNBr and endoproteinase Lys-C were used to separately or sequentially cleave the labeled receptor using procedures previously described (16). The products of cleavage were resolved on 10% NuPAGE gels using MES running buffer (Invitrogen). After electrophoresis, labeled bands were identified by exposure to x-ray film with intensifying screens at Ϫ80°C. Aliquots of affinity-labeled receptor and relevant receptor fragments were deglycosylated with endoglycosidase F, as described previously (23).
Radiochemical Sequencing-For this, the purified fragment from CNBr cleavage of the wild type human calcitonin receptor labeled with the Bpa 16 or from cleavage of the S27M/M48I/M49I mutant receptor labeled with the Bpa 26 probe was coupled to N-(2-aminoethyl-1)-3aminopropyl glass beads (Sigma) through the sulfhydryl side chain of Cys residues. Cycles of Edman degradation were repeated manually in a manner that has been previously reported in detail (26), and the radioactivity released in each cycle was quantified in a ␥-spectrometer.
Statistical Analysis-All observations were repeated at least three times in independent experiments and are expressed as the means Ϯ S.E. Binding curves were analyzed and plotted using the nonlinear regression analysis routine for radioligand binding in the Prism software package (GraphPad Software, San Diego, CA). Binding kinetics was determined by analysis with the LIGAND program of Munson and Rodbard (27).

Characterization of Photolabile Calcitonin
Probes-Both the Bpa 16 and Bpa 26 probes were synthesized by manual solid phase techniques and purified by reversed-phase HPLC, and their identities were verified by mass spectrometry. They both bound saturably and specifically to calcitonin receptor-bearing HEK293-CTR membranes. As shown in Fig. 2, the Bpa 16 probe bound to its receptor with similar affinity to that of natural calcitonin (calcitonin, K i ϭ 83 Ϯ 6 pM; Bpa 16 probe, K i ϭ 168 Ϯ 18 pM), whereas the Bpa 26 probe had affinity more than an order of magnitude lower (K i ϭ 3.1 Ϯ 0.3 nM). They both represented full agonists, simulating cAMP accumulation in HEK293-CTR cells in a concentration-dependent manner, with the Bpa 16 probe having higher potency than the Bpa 26 probe (calcitonin, EC 50 ϭ 30 Ϯ 7 pM; Bpa 16 probe, EC 50 ϭ 28 Ϯ 6 pM; Bpa 26 probe, EC 50 ϭ 97 Ϯ 14 pM; Fig. 2).
Photoaffinity Labeling of the Calcitonin Receptor-Both the Bpa 16 and Bpa 26 probes were used to explore their ability to covalently label the calcitonin receptor. As shown in Fig. 3, they labeled the calcitonin receptor specifically and saturably. The protein band labeled with each probe migrated on a 10% SDS-PAGE gel at approximate M r ϭ 97,000 that shifted to M r ϭ 52,000 after deglycosylation with endoglycosidase F. The differential migration of the labeled glycosylated receptor from earlier studies may relate to species differences and/or to different degrees of receptor glycosylation in distinct cell lines used (19, 28 -31). As expected, the labeling was inhibited by increasing concentrations of calcitonin (Bpa 16 probe, IC 50 ϭ 81 Ϯ 9 nM; Bpa 26 probe, IC 50 ϭ 5.0 Ϯ 1.2 nM). No radioactive band was present in the affinity-labeled non-receptor bearing HEK293 cell membranes.
Identification of Domains of Labeling by Peptide Mapping-We have successfully used CNBr for identification of ligand binding sites of the cholecystokinin receptor (26,32,33), the secretin receptor (11, 12, 14 -17), and the motilin receptor (34). Here again, we used CNBr as the first indication of domain of labeling for the calcitonin receptor. Theoretically, CNBr cleavage of the calcitonin receptor would yield 16 fragments ranging in molecular mass from 0.1 to 11 kDa, with 2 of the fragments also containing potential sites of N-linked glycosylation (Fig. 4). As shown in Fig. 4, CNBr cleavage of the calcitonin receptor labeled with the Bpa 16 probe resulted in a band that migrated on a 10% NuPAGE gel at approximate M r ϭ 9,500 and did not further shift after deglycosylation with endoglycosidase F. Given the molecular mass of the radioiodinated Bpa 16 probe (3657 Da) and the absence of glycosylation, there was only one candidate fragment matching these data. This represents the fragment spanning the amino terminus, the first transmembrane domain, the first intracellular loop, and the second transmembrane domain (Cys 134 -Met 187 , molecular mass ϭ 6138 Da). As also shown in Fig. 4, CNBr cleavage of the calcitonin receptor labeled with the Bpa 26 probe yielded a band migrating at approximate M r ϭ 20,000 and shifted to approximate M r ϭ 6,500 after deglycosylation. Taking into account the molecular mass of the radioiodinated Bpa 26 probe (3732 Da) and clear evidence of glycosylation, the first CNBr fragment at the distal amino terminus of the calcitonin receptor is the only candidate that matches these data.
Endoproteinase Lys-C, which specifically cleaves at Lys residues, was used either separately or sequentially with CNBr to further refine the labeled receptor domain for the Bpa 16 probe. As shown in Fig. 5, endoproteinase Lys-C cleavage of the intact calcitonin receptor labeled with the Bpa 16 probe yielded a glycosylated fragment band (M r ϭ 26,000, top right panel, third lane) that migrated at approximate M r ϭ 5,500 (top right panel, fourth lane) after deglycosylation. This represents the fragment His 121 -Lys 141 . Taken together with the above CNBr data, the labeling domain for the Bpa 16 probe was now narrowed to the segment Cys 134 -Lys 141 . This conclusion was further supported by endoproteinase Lys-C cleavage of the labeled M r ϭ 9,500 fragment resulting from the CNBr digestion of the labeled intact calcitonin receptor (Figs. 4 and 5). This sequential digestion yielded a labeled fragment that migrated at approximate M r ϭ 4,500 (Fig. 5, top right panel, second lane, the labeled segment Cys 134 -Lys 141 ).
Endoproteinase Lys-C was also used sequentially with CNBr to refine the labeled receptor domain for the Bpa 26 probe. As also shown in Fig. 5 (bottom right panel), endoproteinase Lys-C cleavage of the labeled M r ϭ 20,000 fragment resulting from CNBr digestion of the labeled intact calcitonin receptor yielded a radioactive band migrating on a 10% NuPAGE gel at approximate M r ϭ 19,000 (Fig. 5, bottom right panel, fourth lane). Moreover, endoproteinase Lys-C cleavage of the deglycosylated M r ϭ 6,500 CNBr fragment (Fig. 5, bottom right panel, second  lane) from the intact calcitonin receptor labeled with the Bpa 26 probe yielded a radioactive band shifting to approximate M r ϭ 5,500 (Fig. 5, bottom right panel, third lane). These data, sug-gesting that the 22-amino acid signal sequence was cleaved in the mature calcitonin receptor, clearly demonstrate that the first endoproteinase Lys-C fragment Leu 23 -Lys 37 at the distal amino terminus of the calcitonin receptor represented the domain of labeling for the Bpa 26 probe.
To further refine the region of labeling for the Bpa 26 probe, a receptor mutant was developed to introduce an additional site for CNBr cleavage, representing the Ser 27 to Met (S27M) calcitonin receptor mutant, and was transiently expressed in COS cells. The S27M construct bound calcitonin with high affinity (K i ϭ 4.0 Ϯ 1.1 nM) and had a normal cAMP response to calcitonin (EC 50 ϭ 96 Ϯ 15 pM). It was also labeled by the Bpa 26 probe saturably and specifically (Fig. 6, left panel). CNBr cleavage of the labeled S27M receptor construct resulted in a band migrating at approximate M r ϭ 19,500, migrating slightly faster than the labeled glycosylated M r ϭ 20,000 band from CNBr cleavage of the labeled wild type receptor (Fig. 6, middle  panel). After deglycosylation, the M r ϭ 6,500 CNBr fragment from the labeled wild type receptor (Fig. 6, right panel, second lane) shifted to approximate M r ϭ 6,000 in the labeled S27M receptor construct (Fig. 6, right panel, first lane). Together with the above data from the endoproteinase Lys-C cleavage, these data clearly indicated that segment Asn 28 -Lys 37 within the distal amino terminus of the calcitonin receptor contained the site of labeling for the Bpa 26 probe.
Site Identification by Radiochemical Sequencing-Manual Edman degradation sequencing of the purified labeled fragment (Cys 134 -Met 187 , see Fig. 4) resulting from CNBr cleavage of the calcitonin receptor was performed to identify the specific residue labeled by the Bpa 16 probe. A radioactive peak eluted consistently in cycle 4 as shown in Fig. 7, left panel. This result indicates that the site of labeling for the Bpa 16 probe was residue Phe 137 , which is located in the extracellular amino terminus of the calcitonin receptor adjacent to the first transmembrane domain.
To identify the specific residue labeled by the Bpa 26 probe by radiochemical sequencing, an additional Met calcitonin receptor mutant was generated that represented the S27M/M48I/ M49I receptor construct. It was designed to couple the CNBr fragment Asn 28 -Met 59 through Cys 55 to N-(2-aminoethyl-1)-3aminopropyl glass beads (26). This receptor mutant bound calcitonin with high affinity (K i ϭ 3.4 Ϯ 1.4 nM) and had similar cAMP responses to calcitonin stimulation as the wild type receptor (EC 50 ϭ 109 Ϯ 29 pM). The S27M/M48I/M49I calcitonin receptor mutant was saturably and specifically labeled by the Bpa 26 probe (Fig. 6, left panel). CNBr cleavage of the deglycosylated S27M/M48I/M49I calcitonin receptor mutant labeled with the Bpa 26 probe yielded a labeled fragment migrating at approximate M r ϭ 7,000 (Fig. 6, right panel), representing the fragment Asn 28 -Met 59 , distinct in migration from the M r ϭ 6,000 and M r ϭ 6,500 CNBr fragments resulting from cleavage of the labeled S27M mutant and wild type receptor, respectively (Fig. 6, right panel). Radiochemical sequencing of the labeled fragment Asn 28 -Met 59 from the S27M/M48I/M49I calcitonin receptor mutant identified Thr 30 as the site of labeling by the Bpa 26 probe (Fig. 7, right panel).
Characterization of Calcitonin Receptor Site Mutants-The F137A and T30A calcitonin receptor mutants were expressed transiently in COS cells and studied in that cell system to explore the potential impact on the binding and biological activity of calcitonin. Mutation of the residues that had been covalently labeled in the photoaffinity labeling studies were found to not interfere with the normal binding of calcitonin (F137A, K i ϭ 344 Ϯ 28 pM; T30A, K i ϭ 263 Ϯ 70 pM) or with its ability to elicit a full biological response with normal potency (F137A, EC 50 ϭ 117 Ϯ 29 pM; T30A, EC 50 ϭ 157 Ϯ 20 pM). These observations confirm that there is adequate space between the relevant residues in calcitonin and its receptor when normally docked to accommodate the photoprobes. Because the benzoylphenylalanine residues in the probes are larger than the natural calcitonin residues in those locations, such space is important so as not to interfere with normal binding and activation. These observations are fully consistent with the photoaffinity labeling data described above. DISCUSSION G protein-coupled receptors represent the largest group of drug targets in the body. Efforts in identification of the structural basis of ligand binding of receptors have long been a focus for developing receptor-active drugs. However, because of the sparsity and physicochemical nature of these molecules, such efforts have been hindered because of the inability to employ high resolution methods. Our current understanding of the molecular basis of calcitonin binding to its receptor is largely limited to analysis of receptor chimeras (8 -10). Photoaffinity labeling is a more direct approach to identify interactions between a ligand and its receptor. In this work we have successfully developed two agonist probes with photolabile residues in the mid-region and carboxyl-terminal half of calcitonin analogues and identified specific residues within the amino-terminal domain of the calcitonin receptor that are in approximation to these residues when the agonist probes were docked. Photoaffinity labeling has been used for labeling calcitonin receptors from cultured cell lines (19, 28, 30, 31, 35), kidney plasma membranes (36), and transfected cell lines (19,29).  Fig. 4). The right panel also shows the CNBr fragment from cleavage of the labeled S27M/M48I/M49I receptor mutant migrated at M r ϭ 7,000, running slightly slower than that from the labeled wild type receptor and the S27M receptor construct.
Among these studies many used aryl azide-containing moieties such as N-(␤-aminoethyl)-4-azido-2-nitroaniline (30,31,35) and N-hydroxysuccinimide-4-azidobenzoate (36). However, the labeling efficiency through these photoreactive cross-linkers was poor, likely due to the fact that this group of photolabile residues generates highly reactive electrophilic species, leading to low yield photo-insertion reactions (37). Benzophenones have been the preferred chemical moiety for higher yield photoinsertion (37). Suva et al. (19) have successfully incorporated an (⑀-p-benzoylbenzoyl)lysine into a series of salmon calcitonin analogues and demonstrated high efficiency labeling of the calcitonin receptor expressed endogenously in cultured cell lines and transiently in transfected COS cells. However, whether these benzophenone-containing calcitonin analogues could be useful for exploring the ligand binding domains of the calcitonin receptor is not clear. In this work we not only developed two high efficiency photolabile human calcitonin probes incorporating a Bpa but also were able to use them for further mapping the domains of labeling within their receptor. The Bpa 16 and Bpa 26 probes used in this study incorporated the photolabile residue Bpa in the mid-region and carboxyl-terminal half in positions 16 and 26 of the ligand, respectively, both within regions that are important for high affinity binding (2,7).
It is noteworthy that the intrinsic photoaffinity labeling approach identified two residues within the amino-terminal domain of the calcitonin receptor as the sites of covalent attachment to these probes. In fact this domain has been shown to be critical for ligand binding by analysis of calcitonin-glucagon (9, 10) and calcitonin-PTH (8) receptor chimeras. The importance of the amino-terminal domain in ligand binding has been consistent for other members in the class B G protein-coupled receptor family, including receptors for secretin (22, 38 -41), vasoactive intestinal polypeptide (VIP) (38,40), PTH (42), and pituitary adenylate cyclase-activating polypeptide (43,44). This is also the domain labeled in analogous photoaffinity labeling studies for mapping of the binding domains of the secretin receptor (11, 12, 14 -17) (49).
Of particular interest, this work demonstrated spatial approximation between the mid-region of calcitonin (position 16) and receptor residue Phe 137 , a position within the extracellular amino terminus adjacent to the first transmembrane domain of the calcitonin receptor. Sexton and co-workers recently showed that this region was the domain of labeling for a salmon calcitonin probe incorporating a Bpa at position 19 of the peptide. 2 As summarized in Table I, this identification is quite similar to the localization of the analogous region of the secretin receptor interacting with a photolabile residue in the mid-region, in position 13 of secretin (14), and that of the VIP receptor interacting with a carboxyl-terminal residue of the VIP ligand (49). It is also the domain of the PTH 1 receptor interacting with a photolabile residue in the mid-region, in position 13 of PTH (45,50), and in the carboxyl-terminal region, in position 33 of PTH-related peptide (PTHrP) (46) (Table I). In this work we also demonstrated the proximity between the carboxyl-terminal half (position 26) and receptor Thr 30 , a residue within the distal amino-terminal tail of the calcitonin receptor. The identification of this domain is similar to covalent labeling of the distal amino-terminal tail of the secretin receptor by secretin probes incorporating photolabile residues in positions 6, 12, 14, 18, 22, and 26 (11, 12, 14 -17) (Table I). It is also similar to covalent labeling of the distal amino-terminal tail of the PTH 1 receptor by PTH/PTHrP probes, incorporating a photolabile residue in positions 23 (47) and 28 (46) (Table I). The identification of the amino-terminal domain of the calcitonin receptor as the labeling domain for both Bpa 16 and Bpa 26 probes is distinct from photoaffinity labeling of the first extracellular loop domain of the PTH 1 receptor by a position 27 probe (51) and from that of the top of the sixth transmembrane domains of the PTH 1 receptor by PTH/PTHrP probes incorporating a photolabile residue at their amino termini (52,53) (Table I).
The covalent attachment of calcitonin residue 16 to receptor residue Phe 137 within the amino terminus is also consistent with chimeric calcitonin-glucagon receptor studies that suggested the helical portion of the hormone within residues 8 -22 of calcitonin as the principal determinant for binding to the receptor amino terminus (10). In that work, it was also demonstrated that residues 2-6 of calcitonin interact with the receptor transmembrane loop region and are critical for activation of adenylate cyclase (10). It was based on this study that transmembrane domains of the receptor, and (iii) the carboxylterminal region of the peptide (residues 22-32) interacts with other extracellular loop domains. The covalent attachment of calcitonin residue 26 to receptor residue Thr 30 within the amino-terminal tail of the calcitonin receptor may suggest some differences from this model. To further test this hypothesis by photoaffinity labeling studies, development of novel probes incorporating photolabile residues at their amino termini is also needed.
Like that of the secretin receptor and all other members of the class B G protein-coupled receptors, the amino-terminal extracellular domain of the calcitonin receptor contains six conserved Cys residues, differing from members of the class A receptors in the rhodopsin/␤-adrenergic receptor family. These Cys residues are predicted to form disulfide bonds that are thought to be important for ligand binding. Although there are no direct data to demonstrate their involvement in forming intra-domain disulfide bonds in the calcitonin receptor as in receptors for secretin (54), PTH (55), glucagon-like peptide 1 (56), and corticotropin-releasing factor (57), progressive truncation of this domain resulted in loss of ligand binding and receptor activation. 3 Clearly, these conserved Cys residues that probably all involve forming disulfide bonds within the aminoterminal extracellular domain are important to constrain the conformation of the calcitonin receptor. Such constraints should be complementary to those coming from photoaffinity labeling studies for the elucidation of the molecular basis of ligand binding.
It should be noted that the mutation of the residues that were covalently labeled in the current photoaffinity labeling studies (Phe 137 and Thr 30 ) did not interfere with the normal binding and biological activity of calcitonin. This confirms the presence of adequate space in those positions when the natural agonist peptide is normally docked, thus permitting the siting of the photolabile benzoylphenylalanine residue in the photoprobes in the positions of smaller natural residues in calcitonin. The photoaffinity labeling studies provide the constraints of spatial approximation but not necessarily positions of direct residue-residue interactions. In fact, if such interactions were present and critical, the modifications in the photoprobes would likely interfere with their use.
In conclusion, having identified two specific receptor residues that are proximate to two specific residues within a cal-citonin agonist ligand, we have provided two valuable constraints for the molecular modeling of the agonist-bound calcitonin receptor. As the number of pairs of approximated ligand-receptor residues grows and as other key constraints such as disulfide bonding patterns become available, a meaningful model can then be proposed. Such a model can provide insights into whether a common ligand binding mechanism exists for all members of the class B G protein-coupled receptors. Distal N-ECD Val 4 15 12 Distal N-ECD Val 6 17 14 Distal N-ECD Pro 38 17 18 Distal N-ECD Arg 14 12 22 Distal N-ECD Leu 17 11, 16 26 Distal N-ECD Leu 36