Solution structure and mutational analysis of pituitary adenylate cyclase-activating polypeptide binding to the extracellular domain of PAC1-R_S

Abstract

The pituitary adenylate cyclase-activating polypeptide (PACAP) receptor is a class II G protein-coupled receptor that contributes to many different cellular functions including neurotransmission, neuronal survival, and synaptic plasticity. The solution structure of the potent antagonist PACAP (residues 6′–38′) complexed to the N-terminal extracellular (EC) domain of the human splice variant hPAC1-R-short (hPAC1-R_S) was determined by NMR. The PACAP peptide adopts a helical conformation when bound to hPAC1-R_S with a bend at residue A18′ and makes extensive hydrophobic and electrostatic interactions along the exposed β-sheet and interconnecting loops of the N-terminal EC domain. Mutagenesis data on both the peptide and the receptor delineate the critical interactions between the C terminus of the peptide and the C terminus of the EC domain that define the high affinity and specificity of hormone binding to hPAC1-R_S. These results present a structural basis for hPAC1-R_S selectivity for PACAP versus the vasoactive intestinal peptide and also differentiate PACAP residues involved in binding to the N-terminal extracellular domain versus other parts of the full-length hPAC1-R_S receptor. The structural, mutational, and binding data are consistent with a model for peptide binding in which the C terminus of the peptide hormone interacts almost exclusively with the N-terminal EC domain, whereas the central region makes contacts to both the N-terminal and other extracellular parts of the receptor, ultimately positioning the N terminus of the peptide to contact the transmembrane region and result in receptor activation.

Seven transmembrane domain G protein-coupled receptors (GPCRs) are cell surface proteins that transduce signals initiated by hormones or neurotransmitters into the cell (1). Class II GPCRs are a family of receptors that bind structurally related peptide hormones including glucagon, glucagon-like peptides, vasoactive intestinal peptide (VIP), corticotrophin-releasing factor (CRF), parathyroid hormone (PTH), and pituitary adenylate cyclase-activating polypeptide (PACAP) hormone. This family of receptors is capable of regulating intracellular concentrations of cAMP through activation of the adenylate cyclase pathway, and some can also modulate intracellular calcium levels through the phospholipase C pathway. Structurally, they have low homology with other GPCR families but are well conserved within the family. They all contain a relatively large amino-terminal extracellular (EC) domain that plays a critical role in ligand binding. The N-terminal EC hormone-binding domains have common features, typically containing six conserved cysteine residues, two conserved tryptophan residues, and an aspartate residue which has been suggested to be critical for ligand binding (2, 3). Many of the peptide ligands of these receptors have related sequences and can bind to more than one receptor subtype (4).

VIP and PACAP are two prototypical neuropeptides that modulate Class II GPCRs. They are widely expressed in the central and peripheral nervous system and modulate neurotransmission, secretagogue, neuroprotective, neurotrophic, and mitogenic functions (3, 5). Mature VIP exists as a 28-aa peptide, whereas PACAP has two active forms, PACAP-38 and PACAP-27. Mutagenesis studies indicate that deleting the first 5 aa of PACAP-38 converts PACAP from an agonist of the receptor to an antagonist with similar binding affinity. These observations are consistent with the model in which the N terminus of the peptide is involved in receptor signaling, whereas the remainder of the peptide is important for high affinity binding and receptor specificity (3, 5).

Two subtypes of receptors have been identified for PACAP and VIP (3, 5). The subtype I receptor (e.g., PAC1-R) is specific for PACAP over VIP, with a >1,000-fold difference in affinity. In contrast, subtype II receptors (e.g., VPAC1-R and VPAC2-R) bind PACAP and VIP with similar affinities. PAC1-R expression has been detected in the central and peripheral nervous system, the limbic system, and the adrenal gland (6) and has been proposed as a potential target for the treatment of epilepsy, neurodegeneration, and cognition disorders (e.g., Alzheimer's disease, schizophrenia, and anxiety) (7). VIP and PACAP also regulate the expression and release of proinflammatory cytokines and chemokines, and have the potential to be targeted for the treatment of inflammatory diseases such as asthma (8), Crohn's disease (9), rheumatoid arthritis (10), and multiple sclerosis (11). PACAP is also up-regulated in models of pain after insult or injury, and thus has been proposed as a potential target for pain modulation (12).

The sequence conservation in both endogenous peptide hormones and class II GPCRs suggests a similar mechanisms for signal transduction. However, the interactions between the peptides and the receptors are complex and not well understood. Some structural data exist to aid in understanding the complex biology of these systems. Earlier TRNOE studies of weakly binding peptides derived from PACAP indicate that the peptide hormone adopts a helical structure when bound to the receptor (13). Models of VIP/VPAC1-R based on photoaffinity labeling and homology modeling have been proposed (14) that define the general regions of interaction between the hormone and receptor. Recently, the solution structure of the N-terminal EC domain of the related corticotropin-releasing factor (CRF) receptor has been described (15). Although this study proposed a putative peptide-binding surface consistent with chemical shift perturbations, no structural data on the protein-peptide complex was obtained. Thus, additional structural data on the peptide receptor complex is necessary to provide a better understanding of the interactions that stabilize complex formation.

In this article, we report our results on the structure determination of the N-terminal EC domain of the hPAC1-R_S receptor in complex with the PACAP (6′–38′) peptide antagonist. The structure was used to define the binding interface between PACAP and the N-terminal EC domain, and a mutational analysis of the receptor and PACAP peptides was used to determine the energetic contribution from these residues to complex formation. A comparison of the binding affinity for the full-length receptor and the isolated N-terminal EC domain allowed a separation of the binding interactions contributed from the N-terminal EC domain and other extracellular regions of the receptor.

Results and Discussion

Structural Description.

The structure of the hPAC1-R_S N-terminal EC-domain (residues 21–122, C25G)/PACAP(6′–38′) complex was determined by using NMR-derived distance, residual dipolar coupling, and torsion angle restraints. (Throughout the text, peptide residues are always designated by using an apostrophe with the numbering of full length PACAP.) Fig. 1A depicts a superposition of the C^α trace of 10 low energy structures. Residues 25–120 are well defined except for the loop between β3-β4. For PACAP, the N-terminal residues 6′–18′ are less defined than the C terminus of the peptide. A summary of the structural statistics is given in supporting information (SI) Table 1. Analysis of the average-minimized structure with the program PROCHECK showed that 67% of the residues for the N-terminal EC domain of hPAC1-R_S lie in the most-favored region of the Ramachandran plot, whereas an additional 33% are in allowed regions.

Fig. 1.

Solution structure of hPAC1-R_s/PACAP complex. (A) Superposition of the C^α trace of 10 low-energy NMR-derived structures. Residues 25–120 of hPAC1-R_S and residues 10′–30′ of the PACAP peptide were used in the superposition. The atomic rmsd for the backbone atoms is 1.00 ± 0.17 Å and 1.40 ± 0.26 Å for all heavy atoms in the complex. Residues 27–118 of hPAC1-R_S and residues 6′–36′ of PACAP are displayed. (B) Ribbons depiction of the average minimized structure. Secondary structure elements and disulfides of the hPAC1-R_S N-terminal EC domains are labeled. The ribbon for PACAP is colored cyan.

A secondary structure representation (16) of the average minimized NMR structure of the N-terminal EC domain of the hPAC1-R_S /PACAP complex is shown in Fig. 1B. The structure of the N-terminal EC domain of PAC1-R_S consists of an N-terminal helix and four β-strands forming two antiparallel sheets. Three disulfide bonds lock these secondary structure elements together. The helix at the beginning of the domain is connected to β2 by a disulfide linkage between C34 and C63. Two antiparallel sheets are connected by a disulfide bond between C54 (just before the first strand) and C97, which is in a loop after strand β4. After β4 is an extended strand that is terminated by cis P107, leading to the final disulfide bond between C77- C113. The buried core of the protein contains the conserved tryptophans (W64 and W102) and aspartate residue (D59) that are characteristic of family members. The tryptophans span the region between the two antiparallel β-sheets, whereas D59 is in position to form a salt bridge to R95, which is also highly conserved. Mutation of these residues has been reported to diminish peptide binding (3, 5). From the structure, it is likely that these mutations cause a disruption of the core structure that leads to misfolded or less stable N-terminal EC domains, which would affect peptide binding.

The affinity of PACAP 6′–38′ for the soluble N-terminal EC domain of hPAC1-R_S is 350 nM. Under the experimental conditions used here, the protein is fully bound and is in slow exchange on the NMR time scale, resulting in only one set of peaks for the bound conformation. The protein resonances that shift upon peptide binding are localized to one face of the protein, suggesting that only limited structural changes are necessary to accommodate peptide binding.

Free in solution, PACAP 6′–38′ exhibits no detectable stable secondary structure as indicated by the low spectral dispersion observed in the ¹H-¹⁵N HSQC spectrum (SI Fig. 6). However, upon binding to hPAC1-R_S, the dispersion of the peptide amide resonances increases, indicative of the formation of defined structure. Our NMR structural studies of the complex indicate that residues 10′-30′ of PACAP are helical when bound to the hPAC1-R_S N-terminal EC domain. PACAP residues 29′–34′ show Nuclear Overhauser effect (NOE) contacts to Y118, E117, and E119 of the protein, which defines the orientation of the C terminus of the peptide. Residues 35′–38′ at the very C terminus of PACAP are disordered based on their line widths and near random coil chemical shifts. The N terminus of PACAP crosses β3 and contacts L74. To make this contact, the PACAP helix bends at A18′. This bend allows the N terminus of PACAP (residues 10′–17′) to run over β3–β4, whereas the C terminus of the peptide (residues 26′–34′) goes over the disulfide bond (C77–C113) and ends near the C terminus of the N-terminal EC domain. This binding mode places PACAP and the helix of the N-terminal EC domain on opposite sides of the central β-sheet core of the protein. PACAP, when bound, has good charge complementarity to the N-terminal EC domain (Fig. 3).

Fig. 3.

Charge density surface for hPAC1-R_S (A) and PACAP (B). Side chains of selected residues are rendered and labeled. PACAP mutations used the peptide (6′–38′) as template. The binding data for hPAC1-R_S/PACAP complex are given in nM units, with the standard deviation of triplicate measurements in parentheses.

Multiple splice variants have been identified for PAC1-R with different affinities for PACAP-27 and PACAP-38 as well as differential coupling to adenylate cyclase and phospholipase C pathways. There are three splice variants in the N-terminal EC domain, PAC1-R_null (wild type), PAC1-R_S (21-aa deletion) and PAC1-R_vs (45-aa deletion). hPAC1-R_null contains an additional 21 aa not present in the other proteins, as shown in the sequence alignment in Fig. 2A. This insert would be located in the loop between β3 and β4 and could be accommodated without disturbing the core structure of the N-terminal EC domain. These additional residues would most likely be involved in the binding of the C-terminal part of the PACAP peptide. The presence of this 21-aa insert could thus influence receptor selectivity for PACAP-27 and PACAP-38 as well as phospholipase C signaling pathways.

Fig. 2.

Sequence alignment of the N-terminal EC domain of several related class II GPCRs (A) and their corresponding peptide hormones (B). The secondary structure for hPAC1-R_S is indicated above the sequence. The results from the mutational data are summarized by using open, half-filled, and filled circles to indicate no effect, 3- to 5-fold, and >5-fold decrease in binding, respectively, upon mutation of the residue.

Comparison with Mouse CRF-R2β Receptor and VPAC1/VIP.

Recently, the structure of the N-terminal EC domain of the mouse CRF-R2β receptor has been reported (15). This structure has a similar β-sheet core, but hPAC1-R_S has an additional N-terminal helix and a short extended strand after β4. An N-terminal helix may not be observed in CRF-R2β because the domain has five fewer residues before the start of the first β-sheet than hPAC1-R_S, and this could destabilize any helical structure in CRF-R2β. When the two structures are compared, it can be seen that resonances that shift upon binding of the peptide Astressin B to the N-terminal EC domain of CRF-R2β map to a similar but not identical region as we find for PACAP binding to hPAC1-R_S (SI Fig. 7).

A model for VIP binding to VPAC1-R has also been reported based on photolabeling of the receptor and NMR studies of the VIP peptide in solution (14). The photolabeling studies indicated that substitution of benzoylbenzoyl-l-Lys at positions Y22′ and N24′ of VIP modified residues G116 and C122, respectively, of the N-terminal EC domain of VPAC1-R (14). This is consistent with our structure, where the corresponding residues Y22′ and A24′ of PACAP are close to the corresponding residues F106 and C113 in hPAC1-R_S (Fig. 2). In addition, they found that photolabeling F6′ of VIP modified residue D107 of VPAC1-R (which is homologous to residue E99 of hPAC1-Rs). To satisfy this constraint, they modeled the peptide as running parallel with β3–β4. However, in our structure, residues 6′–10′ of PACAP runs across (instead of parallel to) β3–β4, causing F6′ to be farther away from E99.

Mutational Studies.

Based on our structural data, mutational studies on the PACAP peptide were carried out to quantify the importance of individual residues for binding to the receptor (Fig. 3A). PACAP truncations (ending at 36′, 34′, 32′, and 30′) had only moderate effects (i.e., <3-fold) on affinity, consistent with the observation that these residues are unstructured in the complex. However, the PACAP (1′–28′) truncation resulted in a >10-fold loss in affinity. This decrease in affinity is consistent with the loss of the contacts between residues 29′–32′ of PACAP and E117-Y118 of the protein. From the structure, the central PACAP hydrophobic residues make significant contacts to the protein. Consistent with this, we found that single point mutations of the central hydrophobic residues V19′G, Y22′A, V26′G, and L27′A all had a significant (>20-fold) effect on affinity for binding to the N-terminal EC domain. These data indicate an extensive hydrophobic interaction between the N-terminal EC domain and residues 18′–27′ of PACAP. Mutations of residues in the N-terminal part of PACAP (e.g., Y10′A and R14′A) had a negligible effect on binding to the isolated N-terminal EC domain.

Mutations were also made in the soluble N-terminal EC domain and evaluated for binding to the PACAP peptide (Fig. 3B). The Y118A mutant had the greatest (>10-fold) effect on peptide binding. Mutation of the adjacent residue E117R exhibited an ≈3-fold loss in affinity for the peptide. The other protein mutants had smaller effects on affinity. From the sequence alignment in Fig. 2A, it can be seen that the contact region, residues 116–120 (DEYES), results in an acidic patch on hPAC1-R_S that complements the basic end of the peptide. However, in VPAC1-R, the corresponding residues are DDKAA, with a basic residue replacing the tyrosine in hPAC1-R_S. This lysine residue would likely reduce the affinity of PACAP-38 to VPAC1-R if its binding mode is similar to what we have found for hPAC1-R_S. This is consistent with the reported affinity of peptides binding to human VPAC1-R, where VIP∼PACAP-27 > PACAP-38 (5). These data support the role of the interaction between the C terminus of the peptide (residues 28′–32′) and the C terminus of the N-terminal EC domain in defining the high affinity and specificity of hormone binding to hPAC1-R_S. These data also help explain the selectivity of PAC1-R_S for PACAP-38 over VIP and PACAP-27.

Comparison of the Isolated N-Terminal Domain and the Full-Length Receptor.

The PACAP peptides were also evaluated for binding to the full-length receptor in a radioligand binding assay. Consistent with earlier reports on related receptors (17–19), the peptide hormone PACAP binds with ≈1,000-fold higher affinity to the full-length receptor when compared with the soluble N-terminal EC domain. Part of this increased affinity for the full-length receptor is likely due to the influence of the membrane environment (18), where the peptide could preform into a helical structure (13) and diffuse in two dimensions to interact with the receptor. Nonetheless, despite the absolute difference in affinity, there is a good correlation observed for mutant peptide binding to both the full-length receptor and the N-terminal EC domain (Fig. 4 and SI Table 2), suggesting that the peptide-binding mode observed for the isolated N-terminal EC domain is conserved in the full receptor. In particular, mutation and truncation of residues in the C-terminal part of the peptide (Y22′–K38′) have the same relative effect on binding to the full receptor and the isolated N-terminal EC domain, strongly indicating that these residues of the peptide hormone interact almost exclusively with the N-terminal EC domain of hPAC1-R_S even in the membrane environment. In contrast, several residues in the central region of the peptide (Y10′A, R14′A, and K21′E) exhibit a moderately higher effect on binding to the full receptor relative to the N-terminal EC domain (see Fig. 4). In our structure, these residues are aligned on the exposed face of the helical peptide when complexed to the N-terminal EC domain (see Fig. 3A) and thus are accessible for making additional binding interactions with other parts of the full-length receptor.

Fig. 4.

A comparison of ΔΔG for PACAP mutant binding to the isolated N-terminal domain and the full hPAC1-R_S receptor. Standard Gibbs free energy ΔG is calculated as ΔG = −RTlnK_I, where R is the gas constant (1.987 cal · K⁻¹ · mol⁻¹), T is the absolute temperature (298 K), and ΔΔG = ΔG (wild type) − ΔG (mutant).

Previous binding studies of PACAP peptides have delineated the critical role of the N-terminal residues 1′–10′ in receptor binding and activation (20). However, removal of the first five residues (1′–5′) of PACAP had no effect on binding to the N-terminal EC domain (Fig. 3A). In addition, whereas intrapeptide constraints indicated that residues 6′–9′ are in a helical conformation when bound to the N-terminal EC domain, no NOE contacts could be identified between these residues and the N-terminal EC domain. These data suggest that the N-terminal residues of PACAP (e.g., 1′–9′) interact with other parts of the receptor outside of the N-terminal EC domain, such as the extracellular loops or the transmembrane helices of the receptor.

Taken together, the structural, mutational, and binding data are consistent with a model for peptide hormone binding to class II GPCRs shown schematically in Fig. 5. This model has the PACAP peptide complexed to the N-terminal EC domain and lying along the extracellular region of the transmembrane domain. This orientation allows PACAP residues Y10′, R14′, and K21′, which we have shown are important for binding to full-length hPAC1-R_S (Fig. 4), to contact parts of the receptor outside of the N-terminal EC domain, such as the first EC loop (Fig. 5 colored green). Mutagenesis and photolabeling studies on VIP and PTH binding to their respective receptors have indicated the importance of the residues in the first EC loop for hormone binding and receptor activation (5, 21–23). The orientation of PACAP in this model is different from that proposed for Astressin B binding to the N-terminal EC domain of CRF-R2β based on chemical shift perturbations and electrostatic considerations, in which the peptide hormone was placed perpendicular to the membrane surface (15). Superposition of the receptor-bound structure of PACAP-21 (13) onto the model in Fig. 5 places the N-terminal residues of PACAP near the core of the helical bundle of the receptor, consistent with the role of these residues in receptor activation. The requirement for both the N-terminal and central residues of the peptide to contact regions of the receptor outside of the N-terminal EC domain positions the C terminus of the hormone distant from the transmembrane domain, in agreement with the structural and mutational data described in this work.

Fig. 5.

Model of PACAP binding to full-length PAC1-R. The N-terminal EC domain of PAC1-R is shown in pink, and the transmembrane segment of the receptor (modeled by using the rhodopsin structure, PDB ID code 1HZX) is shown in gray. The PACAP peptide used in this study is colored in cyan, with the residues that caused a larger change in binding to the full-length receptor (see Fig. 4) rendered and labeled. Shown in yellow is the structure of the receptor-bound PACAP-21 (13) (PDB code 1GEA) superimposed onto PACAP (6′–38′). The extracellular loop between TM helices 2–3 of rhodopsin is colored green.

Conclusions

The N-terminal EC domains of type II GPCR hormone receptors have high sequence homology and likely adopt similar binding modes for their peptide hormone ligands. Our structural and mutational studies have identified hormone residues that are critical for binding to the N-terminal EC domain of PAC1-R_S. These data aid in the understanding of how receptors differentiate between peptides of very similar sequence and may help in the design of specific antagonists that can be used to understand the activities of the receptor subtypes and ultimately lead to therapeutics for the treatment of a variety of human diseases.

Materials and Methods

Protein Preparation.

The expression of the N-terminal EC domain of hPAC1-R_S (human short variant with 21-aa deletion) residues 21–122 with a C25G point mutation was accomplished by using a thioredoxin fusion protein, followed by a His₆ tag. The C25G mutation was chosen because it has similar binding affinity to PACAP38 as the wild-type protein (Fig. 3B) and improved stability. The protein was expressed in E. coli BL21(DE3) cells and grown in M9 media. Uniformly ¹⁵N, ¹³C-labeled samples were prepared with media containing ¹⁵NH₄Cl and [U-¹³C]glucose. Deuterated samples were grown by using D₂O and [U-²H, ¹³C]glucose. Soluble protein was purified by Ni²⁺-affinity chromatography, followed by thrombin cleavage and a second Ni²⁺-affinity column to remove the fusion protein and His₆ tag. The protein was further purified with a gel filtration column (Superdex-75). Mutants were prepared by using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA) and purified as described above. All protein mutants were checked for proper folding by using [¹H, ¹³C]-HSQC spectra (24).

Labeled PACAP (6′–38′) was expressed as a thioredoxin fusion protein in E. coli BL21(DE3) cells. Uniformly ¹⁵N, ¹³C-labeled samples were prepared with media containing ¹⁵NH₄Cl and [U-¹³C]glucose. The fusion protein and His₆-tag were cleaved with enterokinase, giving the native N terminus for the peptide. The enterokinase enzyme was then removed with a commercial kit (Invitrogen, Carlsbad, CA). Peptides were purified by using HPLC. Unlabeled peptides were prepared in-house by using standard peptide chemistry on a peptide synthesizer and purified by using HPLC to >95% purity.

For structural studies, peptide was added at a slight excess to the protein, followed by removal of excess peptide by using a gel filtration (Superdex-75) column. Three samples were prepared for structural studies. One sample contained uniformly ¹⁵N, ¹³C, ¹H-labeled protein and unlabeled (i.e., ¹⁴N, ¹²C, ¹H) peptide. The second sample contained ¹⁵N, ¹³C, ¹H-labeled peptide and unlabeled protein. The final sample contained ¹⁵N, ¹³C, ¹H-labeled peptide and ²H-labeled protein.

NMR Spectroscopy.

NMR samples contained 0.5–1.0 mM hPAC1-R_S/PACAP complex in 50 mM phosphate (pH 7.0) and 100 mM ammonium sulfate. All NMR experiments were acquired at 298 K on a DRX500 or DRX800 NMR spectrometer (Bruker, Billerica, MA). Backbone ¹H, ¹³C, and ¹⁵N resonance and side chain assignments were obtained with samples containing either labeled protein/unlabeled peptide or deuterated protein/labeled peptide. Triple-resonance experiments [HNCA, HN(CO)CA, HN(CA)CB, HN(COCA)CB, HNCO, and HN(CA)CO] (25, 26) were obtained for both samples. The side chain ¹H and ¹³C NMR signals were assigned from the two samples by using H(CC-CO)NH, (H)C(C-CO)NH (27), or HCCH TOCSY experiments. Stereospecific assignments for a total of nine valine and leucine methyl groups of the N-terminal EC domain of hPAC1-R_S were obtained from an analysis of the ¹³C–¹³C coupling patterns observed for biosynthetically directed, fractionally ¹³C-labeled hPAC1-R_S (28). NOE distance restraints were obtained from three-dimensional ¹⁵N- and ¹³C-edited NOESY or three-dimensional ¹²C filtered ¹³C-edited NOESY spectra acquired with a mixing time of 80 ms (29, 30). To simplify analysis of the NOE experiments, these experiments were repeated for both the labeled protein/unlabeled peptide and the unlabeled protein/labeled peptide.

Residual Dipolar Couplings (RDC) (31) were measured by using a 0.5 mM ¹⁵N, ¹³C PACAP/¹⁵N-labeled hPAC1-R_S complex. Pf1-phage (ASLA Biotech, Riga, Latvia) was added to a concentration of 25 mg/ml (32). Dipolar couplings were measured from coupled ¹⁵N or ¹³C HSQC experiments.

Structure Calculations.

Initial structures of the hPAC1-R_S (residues 21–122,C25G)/PACAP complex were calculated by using a simulated annealing protocol with the program XPLORNIH (33). Three disulfide bonds (C34–C63), (C54–C97), and (C77–C113) were included in the calculations based on observed NOEs between the H^β protons of the two cysteines and other NOEs that were consistent with this disulfide-formation pattern. The final structures were calculated in CNX (34) by using a HADDOCK-type protocol (35) consisting of rigid body docking of hPAC1-R_S(21–122,C25G) and PACAP(6′–38′) starting from our initial structures of the complex and refined by using the full NMR restraints, followed by high-temperature (1,000 K) simulated annealing and molecular dynamics (2 ns, 300 K) using explicit water solvation and full electrostatics. Structures of the complex were calculated by using 1,023 intraprotein, 293 intrapeptide, 46 protein–peptide NOE constraints, 23 RDC, and three intrapeptide hydrogen bond restraints derived from an analysis of amide exchange rates and consistency with observed secondary structure. Because of peptide resonance degeneracy, several ambiguous NOE constraints were also used in the refinement. Torsion angle restraints, φ and ψ, were generated from an analysis of C′, C^α, and H^α chemical shifts by using the TALOS program (36). Forty-eight protein and 15 peptide torsional restraints were used from this analysis. A square-well potential (F_NOE = 50 kcal mol⁻¹) was used to constrain NOE-derived distances. Final structures contained no dihedral-angle violations >5° and no NOE violations >0.2 Å.

Peptide Binding.

Peptide binding to the N-terminal EC domain of hPAC1-R_S was determined by using a fluorescence polarization anisotropy (FPA) competition assay. Fluorescence polarization measurements were conducted as described (37) on an Analyst 96-well plate reader (LJL; Molecular Dynamics, Sunnyvale, CA) by using an Oregon green (OG)-labeled PACAP(6′–38′, K15-OG) peptide as the probe. Titrations were carried out in a buffer containing 120 mM sodium phosphate (pH 7.55), 0.01% bovine γ-globulin, and 0.1% sodium azide. The protein concentration was 600 nM with the probe concentration at 25 nM. Dissociation constants were determined from titration curves with in-house-written software by using the analytical expressions of Wang (38).

Radioligand Binding Assay.

The iodinated peptide, [¹²⁵I]PACAP-27 [specific activity 2,200 Ci/mmol (1 Ci = 37 GBq); PerkinElmer Life and Analytical Sciences, Boston, MA] was used in the binding assay. Crude membranes expressing hPAC1-R_S (5 μg per well) were incubated at 37°C for 20 min in a total assay volume of 0.25 ml in assay buffer (5 mM MgCl₂, 0.5% BSA, and 50 mM Tris·HCl, pH 7.4). For saturation experiments, membranes were incubated with increasing concentrations of radioligand (0.01–1 nM). Displacement studies used 0.2 nM [¹²⁵I]PACAP-27 with varying concentrations of unlabeled peptides. Nonspecific binding was defined by 300 nM unlabeled PACAP-38. The assay was terminated by rapid vacuum filtration through GF/B filters soaked overnight in 0.5% PEI at 4°C. Filters were washed three times with ice-cold harvest buffer (0.5 mM EDTA, 0.1% BSA, and 50 mM Tris·HCl, pH 7.4) and radioactivity bound to the filters was measured by TopCount (PerkinElmer). For the radioligand binding assay, the K_D of the ligand (equilibrium dissociation constant) and B_max (maximal receptor density) were analyzed by nonlinear regression of the saturation binding data. The K_i values were determined from the concentration inhibition curve by the method of Cheng and Prusoff (39). All values were calculated by using the Prism analysis package (GraphPad, San Diego, CA).

Abbreviations

PACAP

pituitary adenylate cyclase-activating polypeptide

VIP

vasoactive intestinal peptide

GPCR

G protein-coupled receptor

EC

extracellular

NMR

nuclear magnetic resonance

NOE

Nuclear Overhauser effect

TRNOE

transferred NOE

RDC

residual dipolar coupling.