JBC Focus on PI3-Kinase with Echelon

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Originally published In Press as doi:10.1074/jbc.M704563200 on September 7, 2007

J. Biol. Chem., Vol. 282, Issue 45, 32834-32843, November 9, 2007
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
282/45/32834    most recent
M704563200v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Harikumar, K. G.
Right arrow Articles by Miller, L. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Harikumar, K. G.
Right arrow Articles by Miller, L. J.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Fluorescence Resonance Energy Transfer Analysis of Secretin Docking to Its Receptor

MAPPING DISTANCES BETWEEN RESIDUES DISTRIBUTED THROUGHOUT THE LIGAND PHARMACOPHORE AND DISTINCT RECEPTOR RESIDUES*

Kaleeckal G. Harikumar{ddagger}, Polo C.-H. Lam§, Maoqing Dong{ddagger}, Patrick M. Sexton1, Ruben Abagyan§, and Laurence J. Miller{ddagger}2

From the {ddagger}Department of Molecular Pharmacology and Experimental Therapeutics, Mayo Clinic, Scottsdale, Arizona 85259, the §Department of Molecular Biology, Scripps Research Institute and Molsoft LLC, La Jolla, California 92037, and the Department of Pharmacology, Monash University, Clayton, Victoria 3800, Australia

Received for publication, June 4, 2007 , and in revised form, August 21, 2007.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Full structural characterization of G protein-coupled receptors has been limited to rhodopsin, with its uniquely stable structure and ability to be crystallized. For other members of this important superfamily, direct structural insights have been limited to NMR structures of soluble domains. Two members of the Class II family have recently had the structures of their isolated amino-terminal regions solved by NMR, yet it remains unclear how that domain is aligned with the heptahelical transmembrane bundle domain of those receptors. Indeed, three distinct orientations have been suggested for different members of this family. In the current work, we have utilized fluorescence resonance energy transfer to establish the distances between four residues distributed throughout fully biologically active, high affinity analogues of secretin and distinct residues in each of four extracellular regions of the intact secretin receptor. These 16 distance constraints were utilized along with nine photoaffinity labeling spatial approximation constraints to study the three proposed orientations of the peptide-binding amino terminus and helical bundle domains of this receptor. In the best model, the carboxyl terminus of secretin was found to bind in a groove above the beta-hairpin region of the receptor amino terminus, with its amino-terminal end adjacent to the third extracellular loop and top of transmembrane segment VI. This refined model of the intact receptor was also fully consistent with the spatial approximation of the Trp48-Asp49-Asn50 endogenous agonist segment with the third extracellular loop region that it has been shown to photolabel. This provides strong evidence for the orientation of peptide-binding and signaling domains of a prototypic Class II G protein-coupled receptor.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 
Understanding receptor structure provides critical insights for the development and refinement of drugs targeting these important molecules. Class II guanine nucleotide-binding protein (G protein)-coupled receptors include several potentially important drug targets, such as receptors for glucagon-like peptide, glucagon, calcitonin, parathyroid hormone, and vasoactive intestinal polypeptide. The first receptor in this family to be cloned was the secretin receptor (1) that is expressed on ductular cells of the pancreatic and biliary tree, where it mediates the secretion of bicarbonate-rich fluid. This receptor has become prototypic of Class II G protein-coupled receptors, sharing primary sequence signatures with the other receptors and key structural motifs, such as three conserved amino-terminal intradomain disulfide bonds (2, 3). However, there is little known about the full three-dimensional structure of any of the receptors in this family.

A substantial advance in our insight into the structure of Class II G protein-coupled receptors came recently with the solution of an NMR structure of the isolated amino terminus of the CRF2beta receptor (4). Based on homology of the primary sequence of the VPAC1 (5) and secretin receptors (6), homologous amino-terminal structures have been proposed. However, the molecular mechanisms proposed for natural peptide ligand binding to these receptors are quite distinct (4-6). Further, the proposed orientations of the amino-terminal domains of these receptors relative to their heptahelical transmembrane bundle regions are all distinct from each other (4-6). Additional experimentally derived constraints beyond the integrity of the peptide backbone are required to determine the correct orientation between these two receptor domains.

In the current work, we have generated such constraints by utilizing quantitative fluorescence resonance energy transfer (FRET)3 measurements between four residues distributed throughout secretin as it is docked with this receptor and distinct extracellular residues located within the amino terminus and within each of the loops of this receptor. The secretin-like fluorescent probes have previously been fully characterized as full agonist ligands (7). Further, the fluorescence properties of these probes established adequate mobility to satisfy the criteria necessary for valid FRET measurements (7). These 16 experimentally derived distances were utilized to refine three series of molecular models of secretin docked to its receptor that were constrained to accommodate the proposed orientations between the amino terminus and helical bundle domains of receptors in this family (4-6).

Secretin docking with the amino-terminal domain of the secretin receptor in each of these series of molecular models was guided by nine spatial approximation constraints coming from photoaffinity labeling studies with secretin-like peptide agonist probes incorporating photolabile sites of covalent attachment throughout the pharmacophore (6, 8-11). It is notable that all three distinct orientations accommodated each of these photoaffinity labeling constraints reasonably well. There was, however, a gradient of accommodation of the FRET data, with the best series of models being oriented based on spatial approximation of the Trp48-Asp49-Asn50 region of the amino terminus with the top of transmembrane helix VI. This was based on WDN peptide photoaffinity labeling studies (12). The most marked difference in the three series of models was the location of this structural motif, found to be far away from the receptor core in the models analogous to the orientations proposed for the CRF2beta and VPAC1 receptors (4, 5).


   EXPERIMENTAL PROCEDURES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
RESULTS
 DISCUSSION
 REFERENCES
 
Materials—Reagents used for receptor mutagenesis were obtained from Stratagene (La Jolla, CA) and Bio-Rad. Ham's F-12 cell culture medium, antibiotic supplements, and Lipofectamine-plus reagent were purchased from Invitrogen; Fetal Clone II was from Hyclone Laboratories (Logan, UT), and nonenzymatic cell dissociation solution was from Sigma. [Tyr10]secretin-27 and natural rat secretin were synthesized in our laboratory (13). The Tyr10 analogue of secretin was labeled by oxidative radioiodination to form the [125I-Tyr10]secretin-27 radioligand we described previously (13). This was purified to homogeneity (approximate specific radioactivity of 2000 Ci/mmol) by reversed phase high performance liquid chromatography. All other reagents were analytical grade.

Secretin-like Fluorescent Probes—A series of secretin analogues that incorporated fluorescent Alexa488 into the amino terminus, midregion, and carboxyl terminus, in positions -1, 13, 22, and 29 (or +2 relative to natural secretin-27), were utilized as fluorescence donors for this work. These probes were previously prepared and validated as full agonists that bind saturably and specifically to the secretin receptor with high affinity (7). Their fluorescence characteristics, both when free in solution and when bound to the secretin receptor, have been described previously (7).

Secretin Receptor Constructs—The cDNA encoding the wild type rat secretin receptor in the eukaryotic expression vector pcDNA3.0 (Invitrogen) was used as template for preparation of the receptor constructs used in this work (Table 1). Based on the experimentally determined disulfide-bonding pattern (2, 3), the free cysteine residues not involved in disulfide bonds and available for reaction with sulfhydryl-reactive reagents were mutated to serine. These included replacement of residues in the amino-terminal region of the receptor (Cys11 and Cys186) with serine, both individually (C11S and C186S) and in groups (null-reactive pseudo-wild type secretin receptor (C11S,C186S)), as well as selective insertion of cysteine residues into the second (T261C) and third (V346C) extracellular loop domains of the null-reactive pseudo-wild type secretin receptor. Single codon mutations were introduced into the parental construct using appropriate sets of complementary forward and reverse oligonucleotide primers with the QuikChange mutagenesis method (Stratagene). Each mutated construct had its sequence confirmed by direct DNA sequencing.


View this table:
[in this window]
[in a new window]

  		TABLE 1
Details of secretin receptor mutant constructs

 
Secretin Receptor-bearing Cell Lines—Chinese hamster ovary (CHO) cells stably expressing cysteine-reactive secretin receptor constructs were created for FRET studies. Parental CHO cells that do not express endogenous secretin receptor and that have no saturable secretin radioligand binding and no cAMP response to secretin were used to make stable cell lines expressing these mutants. Approximately one million parental CHO cells were seeded in a 10-cm Petri dish in Ham's F-12 medium and were transfected with 10 µg of the relevant cDNA construct using Lipofectamine plus reagent (Invitrogen), according to the manufacturer's instructions. 48 h after transfection, the cells were treated with trypsin and lifted from the dishes and transferred into flasks for selection using G-418. The receptor-bearing cells were enriched by sequential rounds of limiting dilution cloning. Selected receptor-bearing cells were confirmed by secretin binding and receptor activation.

Receptor Binding Studies—CHO cells stably expressing wild type or mutant secretin receptor constructs were plated in 24-well Costar tissue culture plates at a density of ~50,000 cells/well 3 days prior to receptor-binding assays. The cells were incubated with ~5 pM [125I-Tyr10]secretin-27 in the absence or presence of increasing concentrations of unlabeled secretin in 0.5 ml of Krebs-Ringer-Hepes (KRH) medium containing 25 mM HEPES, pH 7.4, 104 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM KH2PO4, 1.2 mM MgSO4, 0.2% bovine serum albumin, and 0.01% soybean trypsin inhibitor at room temperature for 1 h. Nonspecific binding was defined as radioligand bound in the presence of 1 µM unlabeled secretin and was typically less than 15% of total cpm bound. After termination of the binding reaction by washing the cells twice with iced KRH medium, cells were lysed with 0.5 M NaOH, and the residual radioactivity in the lysate was quantified by a {gamma} counter. Data were analyzed using the LIGAND program of Munson and Rodbard (14) and were graphed using the nonlinear least squares curve-fitting routine in the Prism 3.0 suite of programs by GraphPad (San Diego, CA).

Biological Activity Assays—The ability of secretin receptor constructs to signal when stimulated with secretin was assessed by measuring cAMP responses in CHO cells expressing the relevant receptor constructs. For this, ~8,000 cells were plated in each well of a 96-well plate 72 h prior to the assay. On the day of assay, cells were washed twice with PBS before being stimulated for 30 min at 37 °C with increasing concentrations of secretin (ranging from 10-12 to 10-6 M) in KRH buffer containing 0.2% bovine serum albumin, 0.01% soybean trypsin inhibitor, 0.1% bacitracin, and 1 mM 3-isobutyl-1-methylxanthine. Reactions were stopped by adding 6% ice-cold perchloric acid. The supernatants were adjusted to pH 6, and cAMP levels were assayed in a 384-well white Optiplate using a LANCE kit from PerkinElmer Life Sciences. This assay uses a homogenous time-resolved FRET immunoassay based on the competition between a europium-labeled cAMP tracer complex and sample cAMP for binding sites on cAMP-specific antibodies labeled with Alexa Fluor® 647. Briefly, samples (6 µl of supernatants from each well) and equal volumes of Alexa Fluor® 647-cAMP antibodies in KRH buffer containing 0.1% bovine serum albumin (1:100 dilution) were incubated for 30 min at room temperature. 12 µl of detection mix containing the europium-labeled streptavidin and biotin-cAMP was then added. After incubation for an additional 1 h, the time-resolved FRET signals at 615- and 665-nm wavelengths upon excitation at 340 nm were measured using the 2103 Envision fluorescence plate reader (PerkinElmer Life Sciences). The signals obtained at 665 nm were used directly for data analysis after subtraction of the readings from blank wells. The cAMP levels were calculated by interpolation using a cAMP standard curve.

Design and Synthesis of Donor and Acceptor Groups for FRET Studies—Fluorescence donor and acceptor groups were chosen that were suitable for FRET studies. For donors, we used Alexa488-derivatized secretin peptides, and for acceptors, we utilized Alexa568-labeled secretin receptor constructs. This pair of fluorophores were found to have appropriate spectral overlap, with 50% energy transfer at 62 Å, similar to the analogous fluorophores used in our previous report of FRET at the ligand-occupied cholecystokinin receptor (15, 16).

For the current study, the fluorescence donors were designed to situate the Alexa488 in a series of distinct positions throughout the secretin peptide, at its amino terminus, at its carboxyl terminus, and in two positions within the middle region of the peptide. We previously prepared and characterized these probes, establishing their high affinity binding and their full biological activity and characterizing their fluorescence properties both in solution and bound to secretin receptor (7).

The cysteine-reactive Alexa568-tagging reagent, Alexa568-MTS, was prepared as previously described (15, 16). This reagent was used to derivatize the accessible cysteine residues on the ectodomain of CHO cell lines stably expressing secretin receptor constructs having a single reactive cysteine residue. These included constructs with the reactive cysteine within the receptor amino terminus (Cys11), first exoloop (Cys186), second exoloop (Cys261), and third exoloop (Cys346) (illustrated in Fig. 1).

Labeling of Cells Expressing Mutant Secretin Receptors—Derivatization of secretin receptors with Alexa568-MTS was carried out as described previously (15). Briefly, cells expressing the monoreactive secretin receptor constructs were detached from tissue culture flasks using nonenzymatic cell dissociation medium and were incubated with 1 µM cell-impermeant Alexa568-MTS reagent for 20 min at room temperature in KRH medium. The unreacted MTS reagent was removed by washing and centrifugation. The Alexa568-labeled secretin receptorbearing cells were then incubated with 50 nM Alexa488-secretin fluorescent probes for 1 h at 4 °C. Labeled cells were washed with ice-cold buffer and resuspended in cold buffer for FRET studies. Control studies included the labeling of untransfected parental CHO cells with fluorescence donor and acceptor as well as performance of the experiment in the presence of a saturating concentration of nonfluorescent secretin.

Fluorescence Spectroscopy—Steady-state fluorescence was recorded using a Fluoromax-3 fluorometer (SPEX industries, Edison, NJ) at 25 °C with a 1-ml quartz cuvette. Fluorescence resonance energy transfer measurements were made using intact cells. After excitation at 481 nm, multiple emission spectra were accumulated between 500 and 700 nm (scanned at 0.3 s/nm). Unlabeled cells were utilized as a control to correct for background fluorescence and light scatter.

Fluorescence Resonance Energy Transfer—The critical distance (Ro) between the Alexa488-secretin probes (donor) and the Alexa568-secretin receptor constructs (acceptor) was calculated using the Förster equation, as described previously (15). The critical distance between the donor and the acceptor at 50% transfer efficiency was calculated based on the equation, Ro = 9786 (Jn-4{kappa}2QD)1/6 Å, where n is the refractive index of the medium (1.4 for aqueous medium) and {kappa}2 is a geometric factor describing the relative orientation of the transition dipole of the donor and acceptor fluorophores. The orientation factor was considered to be 2/3, a condition that is valid if donor and/or acceptor can exhibit isotropic, dynamic orientation within the time scale of the fluorescence lifetime of the probes. Our previous steady state measurements of anisotropy of the receptor-bound probes (0.02-0.05) support this assignment, based on sufficient rotational freedom (7). QD is the fluorescence quantum yield of the donor in the absence of acceptor. For the Alexa488-probes (donor), the quantum yield was determined based on the comparison of emission intensities with that of sodium fluorescein in 0.1 N NaOH, using a quantum yield value 0.92 for fluorescein. For the Alexa568 receptors (acceptor), we compared fluorescence intensity with that of rhodamine-B in water, using the standard value of 0.70. Quantum yields of donor and acceptor bound to the secretin receptor were similar to those described earlier for analogous experiments with the cholecystokinin receptor (15) (quantum yield of 0.60 for donor and 0.68 for acceptor). J is the spectral overlap integral between donor emission (F({lambda})) and acceptor absorbance (isin({lambda})) spectra. It was calculated by the following: J ={int}F({lambda})isin({lambda}){lambda}4d{lambda}/{int}F({lambda})d{lambda}, where {lambda} is the wavelength in cm, F({lambda}) is the normalized and integrated fluorescence of the donor at wavelength {lambda}, and isin({lambda}) is acceptor absorption in absorbance units at wavelength {lambda} for the overlapping area. The efficiency of energy transfer (E) was calculated using the following equation based on the fluorescence intensity E = 1 - FDA/FD, where FDA and FD are the fluorescence intensities of the donor (Alexa488-ligand) in the presence and absence of the acceptor (Alexa568-secretin receptor), respectively. Corrected proximal distance is the 6th inverse power of the distance between donor and acceptor as noted in the equation, R = Ro((1 - E)/E)1/6 Å, where Ro is the distance between donor and acceptor with an efficiency of energy transfer of 50%.

Molecular Modeling—All molecular modeling activities were conducted using a stochastic global energy optimization procedure implemented in internal coordinate mechanics (17). This procedure consisted of the following iterative steps: (a) random conformational change of a dihedral angle according to the biased probability Monte Carlo method (18); (b) local minimization of all free dihedral angles; (c) acceptance or rejection of the new conformation based on the Metropolis criterion at the simulation temperature, usually at 600 K (19). This procedure can generate and search through diverse sets of conformations by actively sampling a selected set of dihedral angles. All calculations were carried out on 3.4-GHz Intel XEON-EMT processors.

Three series of molecular models of the rat secretin receptor were generated, each based on one of the distinct orientations between the amino-terminal and core transmembrane helical bundle domains proposed for receptors in this family (4-6). In each of these, the amino-terminal and core domains of the secretin receptor represented homology models that we previously developed (6). For the amino terminus, we utilized the 20 NMR models of the mouse CRF2beta receptor amino-terminal domain as template (6). For each of the 20 resultant models, Cys11 was modified to Alexa568-labeled cysteine, and its side chain was sampled (illustrated in Fig. 1). The helical bundle reported in our previous study was also utilized (6). The first, second, and the third extracellular loops were built and attached to the relevant helices after searching through the Protein Data Bank for suitable loop templates, followed by side chain sampling and backbone minimization. Cys186, Thr261, and Val346 of the three extracellular loops were modified to Alexa568-labeled cysteine, with their side chains further sampled as also illustrated in Fig. 1.

In all models, a distance restraint was imposed between the carboxyl-terminal end of the amino-terminal domain and the amino-terminal end of transmembrane I. Each series of molecular models utilized a distinct spatial orientation of these two receptor domains as a constraint. The first series utilized spatial approximation between the WDN epitope and the top of the sixth transmembrane segment (6), whereas the second and third series utilized the orientations proposed for the CRF2beta and VPAC1 receptors, respectively (4, 5). The amino-terminal domains were then docked onto the helical bundle by Monte Carlo sampling of the six positional variables of the aminoterminal domain and the side chain dihedrals of the four Alexa568-labeled residues. The contacts between the two domains were further optimized by side chain sampling to generate preliminary models of the secretin receptor incorporating both the amino terminus and receptor core (20).

Docking of Secretin to the Full Models of the Secretin Receptor—The initial conformation of secretin to be utilized in docking studies was taken from a previous solution-phase NMR determination of the porcine secretin structure, and Gln14 was modified back to arginine, the residue in rat secretin (21). The amino terminus was extended by an Alexa488 label. Leu13 and Leu22 were modified to Alexa488-labeled lysine. The carboxyl terminus was extended by Gly28-Cys29-Alexa488.

Distance restraints were imposed between residues Phe6, Arg12, L14K, Arg14, Arg18, Arg21, L22K, Leu23, and Leu26 of secretin and residues Val4, Val6, Val103, Pro38, Arg14, Arg15, Leu17, Arg21, and Leu36 of the secretin receptor, respectively. A general distance restraint was imposed between residue His1 of secretin and the top of transmembrane helix VI, as supported by previous photoaffinity labeling studies (22). In addition, 16 distance restraints were imposed between the four Alexa488-labeled secretin peptide residues and four Alexa568-labeled secretin receptor residues using the experimentally determined target. Hybrid receptor models were constructed in which the four Alexa568-labeled secretin receptor residues were represented by explicit atoms, whereas the rest of the receptor was represented by a grid potential. The secretin peptide was first docked to these hybrid models in 10 independent runs for each series. During the initial docking, the six positional variables of secretin as well as the side chains of the eight Alexa-labeled residues were actively sampled with rigid backbone and flexible side chains during cycles of local minimization. Each of the 10 independent runs in each series converged to a single solution after 20 h of docking. Subsequently, the secretin-secretin receptor complexes were refined in full atomic models for each series. During refinement, the side chains of both secretin and the secretin receptor were actively sampled, whereas the backbone of secretin residue Asp15 and residues 1-43 and 102-110 of the secretin receptor were allowed to be more flexible during cycles of local minimization. The refinement process typically lasted for 90 h, with the lowest energy conformation among the 10 independent runs in each series retained. The health of the models was established by PROCHECK and WHAT_CHECK evaluations (23, 24).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
DISCUSSION
 REFERENCES
 
Characterization of Secretin Receptor Constructs—The secretin receptor has been carefully characterized to establish that it has three highly conserved disulfide bonds, involving six conserved amino-terminal cysteine residues (2, 3), and has two additional free cysteine residues in its ectodomain, in positions 11 (amino terminus) and 186 (first loop) (3) (Fig. 1). The latter are available for derivatization using the cell-impermeant, sulfhydryl-reactive fluorescence-tagging reagent, Alexa568-MTS. In the current series of studies, we have mutated Cys11 and Cys186 to serine residues separately or together to prepare monoreactive and null-reactive pseudo-wild type secretin receptors (Table 1). In the setting of the null-reactive receptor, we also introduced new cysteine residues into positions believed to accommodate such mutations (T261C in the second loop and V346C in the third loop), based on analysis of the level of conservation of these residues in family alignments (Table 1). Each of the mutant secretin receptor constructs that were employed for the current series of studies listed in Table 1 was characterized for secretin binding and biological responses (Table 2). Each of these constructs was able to bind secretin with equivalent high affinity and responded to secretin stimulation with a fully efficacious cAMP response. Of note, the second loop construct (T261C) had a substantial negative impact (~6.5-fold) on the potency of stimulation of cAMP production with secretin.


View this table:
[in this window]
[in a new window]

  		TABLE 2
Binding and biological characteristics of secretin receptor constructs Values are expressed as means ± S.E. of data from at least three independent experiments. Basal levels of cAMP were 1.7 ± 0.4 pmol/million cells, and maximal values reached 199 ± 35 pmol/million cells, with no significant differences for any of the experimental groups.

 
FRET Studies—It was important to fully characterize and validate fluorescence donors and acceptors. The donors for FRET were Alexa488-labeled secretin peptides, in which the fluorophore was incorporated at either end of the ligand or into two positions in its midregion (positions 13 and 22). The fluorescence properties of these ligand probes have been previously described (7). It is critical to note that the mobility of the fluorophore in each of these probes while bound to the secretin receptor was within the range of 0.02-0.05, reflecting an adequate degree of mobility to be valid for use in FRET studies (7).

The Alexa568-MTS reagent was utilized to derivatize the monoreactive accessible cysteine residues in the secretin receptor constructs to act as fluorescence acceptors. It is important to note that other membrane proteins also have accessible free cysteine residues that get derivatized by this approach. The specificity of the FRET signal is dependent on those residues on other membrane proteins that were derivatized with fluorescence acceptor being too far away from the fluorescence donor that is selectively occupying the secretin receptor for them to contribute to the fluorescence transfer signal. To establish this, controls were incorporated that selectively eliminated the fluorescence derivatization of the receptor (null-reactive construct) and that selectively eliminated the fluorescent ligand occupation of the receptor (nonfluorescent ligand competition control and non-receptor-bearing CHO cell control). The fluorescence donor probes each had its emission peak at 521 nm after excitation at 481 nm. The fluorescence acceptors had minimal emission when excited at 481 nm, but when excited by light at 578 nm they emitted well at 603 nm.


Figure 1
View larger version (59K):
[in this window]
[in a new window]

  		FIGURE 1.
Secretin receptor constructs. Shown is a schematic diagram of the primary sequence and proposed topology of the rat secretin receptor. The pseudo-wild type null-reactive background receptor converted each of the free cysteine residues in positions 11 and 186 to serine residues. Highlighted are the four sites of fluorescence labeling of the receptor in various monoreactive constructs. These include sites within the amino terminus (position 11) and the first (position 186), second (position 261), and third (position 346) extracellular loops.

 
Fluorescence resonance energy transfer was measured between each of the four Alexa488-secretin probes bound to CHO cells expressing each of the Alexa568-labeled monoreactive secretin receptor constructs. Significant energy transfer was observed between each of the probes and each of these constructs. Shown in Fig. 2 are data illustrating significant energy transfer in a representative experiment using one of these pairs of donor and acceptor (Alexa488-secretin bound to monoreactive V346C mutant secretin receptor derivatized with Alexa568) as well as the absence of energy transfer when the null-reactive receptor construct was expressed under the same conditions. Additionally, no significant energy transfer was observed for any of the other controls that eliminated donor occupation of the secretin receptor by using competition with nonfluorescent secretin or working with non-receptor-bearing cells (Fig. 2).

The degrees of spectral overlap between each fluorescent ligand occupying the secretin receptor and the receptor constructs fluorescently labeled in distinct positions within the amino terminus (position 11), first loop (position 186), second loop (position 261), and third loop (position 346) were determined and are illustrated in Fig. 3. The degrees of spectral overlap varied widely among the pairs of donor and acceptor reflecting distinct differences in their spacing, given that the efficiency of fluorescence transfer was also measured for each pair and was found to be relatively constant (Fig. 3). These data were subsequently analyzed using the Förster equation to calculate the proximal distances between the fluorescence Alexa488 donors and the Alexa568 acceptors. Measured distances between donor and acceptor pairs are illustrated in Table 3, whereas the corrected distances are illustrated in Table 4. The donor at the amino terminus of secretin was found to be closest to acceptors situated within the second and third loops of the receptor. The donor at the carboxyl terminus of secretin was found to be closest to the acceptor within the amino-terminal tail of the receptor. The donors within the midregion of secretin were found to be closest to the second loop position of the acceptor.


View this table:
[in this window]
[in a new window]

  		TABLE 3
Measured distances between donor Alexa488-ligands and acceptor Alexa568-secretin receptor constructs Values are expressed as means ± S.E. of data from at least five independent experiments.

 


View this table:
[in this window]
[in a new window]

  		TABLE 4
Proximal distances between donor Alexa488-ligands and acceptor Alexa568-secretin receptor constructs Values are expressed as means ± S.E. of data from at least five independent experiments. Proximal distances were calculated from the experimentally derived values in Table 3, utilizing the measured efficiencies of transfer between donor and acceptor for each pair of constructs (Fig. 3).

 


Figure 2
View larger version (26K):
[in this window]
[in a new window]

  		FIGURE 2.
Fluorescence emission spectra. Shown are representative fluorescence emission spectra collected between 500 and 700 nm when the noted parental and genetically engineered CHO cells were allowed to bind Alexa488-secretin, had their sulfhydryl groups derivatized with Alexa568, and were excited with light at 481 nm. Shown in the top panel are data from representative studies with the V346C monoreactive secretin receptor construct and the null-reactive receptor construct. Significant energy transfer (a FRET emission peak at 603 nm) was apparent with the monoreactive construct and absent with the control null-reactive construct. The lower panel includes two additional control experiments in which nonfluorescent secretin competed with the Alexa488-secretin for receptor binding and in which non-receptor-bearing parental CHO cells were utilized. Neither gave any evidence of a FRET signal.

 
Molecular Modeling—In the construction of the current models, we incorporated nine distance restraints previously reported in photoaffinity labeling studies and the 16 distance restraints identified in the current FRET studies. Key features of the best models from each of the series of molecular modeling studies are illustrated in Fig. 4. Each of these illustrates very distinct orientations between the amino-terminal domain and the transmembrane helical bundle domain of the secretin receptor.


Figure 3
View larger version (28K):
[in this window]
[in a new window]

  		FIGURE 3.
Degree of spectral overlap and efficiency of energy transfer. Shown are bar graphs illustrating the degree of spectral overlap between donor Alexa488-ligands and acceptor Alexa568-secretin receptor constructs (A) and the efficiency of energy transfer between donors and acceptors in intact cells (B). Data are expressed as means ± S.E. of a minimum of five independent experiments. The degree of spectral overlap varied considerably with distinct pairs of donor and acceptor, supporting different distances for different pairs. The efficiency of energy transfer was more consistent. These efficiencies are critical for the calculation of absolute distances between the donor and acceptor, as illustrated in Tables 3 and 4.

 
The stereochemistry and packing environment of the best models from each series were acceptable. When evaluated by PROCHECK, the best model in the first series had 83.7, 13.5, 2.9, and 0% of its backbone dihedral angles in the most favored, additionally allowed, generously allowed, and disallowed regions of a Ramachandran plot, respectively. The overall average G-factor was -0.10. Its second generation packing quality determined by WHAT_CHECK was -0.174. For the best model using the CRF2beta receptor orientation series (series two), 81.1, 14.1, 4.5, and 0.3% of its backbone dihedral fell in the most favored, additionally allowed, generously allowed, and disallowed region of the Ramachandran plot, respectively. The overall average G-factor was 0.00. Its second generation packing quality was -0.293. For the best model using the VPAC1 receptor orientation series (series three), 83.0, 14.4, 1.6, and 1.0% of its backbone dihedral fell in the most favored, additionally allowed, generously allowed, and disallowed region of Ramachandran plot, respectively. The overall average G-factor was 0.06. Its second generation packing quality was -0.417.


Figure 4
View larger version (23K):
[in this window]
[in a new window]

  		FIGURE 4.
Secretin receptor models with different orientations of the amino-terminal domain. Shown are the best models of the full secretin receptor-secretin peptide complex using three different orientations of the amino-terminal and core domains. Only residues Cys24-Lys379 of the receptor are shown as a ribbon; the Trp48-Asp49-Asn50 sequence is shown in both xstick and cpk representations, with the direction of orientation highlighted with arrows. The secretin peptide is colored blue-red from the amino terminus to the carboxyl terminus. Left, the receptor generated from the orientation of the secretin receptor proposed by Dong et al. (6) in series one is represented as a yellow ribbon. Middle panel, the receptor generated from the orientation of the CRF2beta receptor proposed by Grace et al. (4) in series two is represented as a green ribbon. Right, the receptor generated from the orientation of the VPAC1 receptor proposed by Tan et al. (5) in series three is represented as a purple ribbon.

 
Tables 5 and 6 show the final distances in the best models of each series. In the best model in series one, the maximum C{alpha}-C{alpha} distances between cross-linked residues were well within the range that can be reached by the photoaffinity labeling group. The final distances between the four positions within the secretin receptor for Alexa568 and the four positions of Alexa488 within the secretin peptide in the best model in series one show little violation from the experimental distances (Table 6). Therefore, the best model in series one was able to satisfactorily accommodate all of the experimental data.


View this table:
[in this window]
[in a new window]

  		TABLE 5
C{alpha} distances between cross-linked residues in secretin and the secretin receptor after grid docking and flexible backbone refinement, in the best models from each of the three series, the first based on the secretin receptor structure proposed by Dong et al. (6), the second based on the CRF2beta receptor structure proposed by Grace et al. (4), and the third based on the VPAC1 receptor structure proposed by Tan et al. (5)

 


View this table:
[in this window]
[in a new window]

  		TABLE 6
Distances between donor Alexa488-ligands and acceptor Alexa568-secretin receptor in the best models from each of the three series, the first based on the secretin receptor structure proposed by Dong et al. (6), the second based on the CRF2beta receptor structure proposed by Grace et al. (4), and the third based on the VPAC1 receptor structure proposed by Tan et al. (5)

 
The best models generated from the two alternative orientations were of a quality similar to that of the best model from the proposed orientation in the first series. The final distances between photoaffinity-labeled residue pairs in these two models were slightly longer than that of the orientation proposed in series one (Table 5). More importantly, these two alternative orientations resulted in larger distance violations of the FRET data than in the orientation proposed in series one (2.9 and 3.1 Å for CRF2beta and VPAC1 receptor orientations versus 1.0 Å for the orientation proposed in series one; Table 6). Fig. 5 shows more detailed views of the best model of any of these series, representing the best model in series one.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
REFERENCES
 
FRET measurements provide potentially powerful tools to explore molecular conformations of molecules and complexes of interacting molecules. In the current work, we have applied this quantitative technique to establish spatial relationships between Alexa488 in distinct positions spanning the secretin peptide (amino terminus, positions 13 and 22 within the midregion, and carboxyl terminus) when docked at the secretin receptor and distinct sites within each of the extracellular domains of this receptor (amino-terminal tail in position 11, first loop in position 186, second loop in position 261, and third loop in position 346). This provides a set of 16 unique constraints that provide new data on the spatial orientation of secretin as docked to the amino-terminal domain of this receptor and the receptor transmembrane helical bundle domain. Inadequate experimentally derived constraints currently exist to be confident of the orientation between these two domains for any receptor in this important Class II G protein-coupled receptor family. As such, there is little consensus across the structures that have been proposed in the literature to date, with all differing substantially in regard to the relative orientation of these two receptor domains (4-6, 25, 26). There are also widely divergent proposals for the site of peptide docking to receptors in this family. NMR studies of antagonist peptide binding to isolated amino-terminal domains of the CRF2beta receptor (4, 25) and of the PAC1 receptor (26) demonstrate peptide binding to opposite sides of their highly similar aminoterminal domains.

Although high resolution refinement of the structure of the entire receptor is not yet possible, the current study provides constraints across sites throughout the secretin pharmacophore and independent sites in each of the extracellular domains of the receptor. This allowed a global estimate of the positioning of these sites in the secretin-docked receptor that also serves as a guide to the likely orientation of the amino-terminal domain relative to the receptor core. However, to allow the FRET approach to provide meaningful data, strict controls were established that included studies that established normal high affinity and functional docking of secretin to the monoreactive cysteine mutants of the secretin receptor that were later derivatized with fluorescence acceptor. This minimized any steric interference by the Alexa moiety and ensured that distances to that acceptor were internally consistent. Furthermore, no nonspecific FRET signal occurred in cells that lacked the acceptor-derivatized receptor or where saturable binding of the donor-labeled peptide probe was competitively eliminated using unlabeled secretin.

Experimentally utilizing the probes individually and establishing their high affinity binding and full biological activity helped to ensure meaningful measurements from each position. These individual distance restraints were incorporated via two different modeling procedures. In the first method that was most similar to the experimental approach, four secretin ligands, each containing one Alexa488 label, were docked to four distinct secretin receptor constructs, each containing one Alexa568 label to generate 16 models. The advantage of this method is its close approximation to the experimental procedure. Practically, however, each of these 16 models utilizes only one distance restraint and is therefore often incompatible with the other 15 models and distance restraints. Combining these 16 models into one unified model that would be fully compatible with all of the 16 distance restraints was not feasible. Therefore, to arrive at a unified model, we used an alternative method, in which all of the four Alexa488 and four Alexa568 labels were incorporated into the same model simultaneously. The disadvantage of this approach is the possible steric hindrance of the eight Alexa moieties in the same model. However, the ability to utilize all 16 distance restraints in a single unified model seems to outweigh this possible drawback. This approach was followed for each of three distinct spatial orientations between the receptor domains that have been described for members of the Class II G protein-coupled receptor family (4-6).


Figure 5
View larger version (38K):
[in this window]
[in a new window]

  		FIGURE 5.
Secretin receptor model. Shown is the best model of the full secretin receptor-secretin peptide complex. This represents the model from series one that accommodates nine photoaffinity labeling constraints, three disulfide bonds shown to exist in the secretin receptor, and 16 FRET constraints. Left panel, the full receptor is represented as a yellow ribbon, with specific transmembrane segments and extracellular loop three identified. Surface representation of residues Cys24-Asn110 of the amino-terminal domain is colored according to the electrostatic potential. Blue, positive charge; red, negative charge. The secretin peptide is colored blue-red from the amino terminus to the carboxyl terminus. The nine photoaffinity labeling constraints are represented as dotted lines. Right, receptor residues Cys24-Lys379 are represented as a yellow ribbon. The secretin peptide is colored blue-red from the amino terminus to the carboxyl terminus. The nine photoaffinity labeling constraints are represented as dotted lines.

 
Indeed, the best models in each series were able to accommodate all 16 distance restraints. It is also notable that PROCHECK and WHAT_CHECK evaluations of these models indicated little difference in their structural quality or packing characteristics. Similarly, all three models accommodated all of the photoaffinity labeling data for residues scattered throughout the pharmacophoric region of secretin. The distances between the photolabile residues within the docked peptides and the receptor residues that were covalently labeled were only slightly shorter for the model in series one that was based on the WDN approximation with its target (6) than in the models from series two and three based on the orientations proposed for the CRF2beta and VPAC1 receptors, respectively (4, 5). This is perhaps less surprising when it is appreciated that most of the covalently labeled receptor residues reside in flexible regions of the receptor amino terminus, outside of the conserved core regions that are structurally defined in the NMR data sets (4).

Despite having radically different orientations of the receptor amino terminus relative to the receptor transmembrane core in the models from these three series of studies, the aminoterminal end of secretin was found to reside above the third extracellular loop in each of the best molecular models. This is one of the few consistent experimental observations that have been made with several members of this receptor family, with the amino terminus of natural peptide ligands (secretin, parathyroid hormone and calcitonin) shown to be spatially approximated with the third extracellular loop of their respective receptors (22, 27, 28).

The FRET distance data collected in the current report separated the best models from each series more effectively than did the photoaffinity labeling data. Distance violations in the FRET data provided the clearest indication that the best model in series one was better than those from the other two series. The major difference in the best models from each of the three series of studies was the orientation of the proposed endogenous agonist sequence, WDN (6). This sequence was in contact with the region above transmembrane VI that has been shown to be photoaffinity-labeled by photolabile derivatives of WDN (12) in the series one model. In contrast, it was projected upwards in different angles in the best models coming from series two and three. The positions of this endogenous peptide sequence in those models were clearly incompatible with its proposed interaction with the body of the secretin receptor (6).

It is important to note that the observation of the endogenous WDN sequence within the secretin receptor amino terminus to represent a full agonist at that receptor (12) has been generalized to other members of this receptor family. The analogous regions between the second and third conserved cysteine residues of the amino terminus of the calcitonin receptor and of the VPAC1 receptor have also been demonstrated to possess full agonist activity at their respective receptors (12). There is even some cross-reactivity for this action among these receptors (12). The target of action of these endogenous peptides has been established by photoaffinity labeling as well as by biological activity of these sequences on receptor mutants that were not responsive to natural peptide ligands (12).

The model we are currently proposing took advantage of nine photoaffinity labeling spatial approximation constraints and 16 FRET constraints, all measured in the context of a fully functional ligand-receptor complex. This represents a refinement of the preliminary model we first proposed recently based exclusively on photoaffinity labeling data (6). This molecular model is also compatible with the proposed function and mechanism of action of the endogenous agonist WDN epitope within the receptor amino terminus (12). We therefore believe that this model is most relevant to the actual binding conformation of a peptide ligand to an intact receptor in this important family. However, given the relatively large size of the fluorescence donors and acceptors utilized in the current study and the relatively long distances for each of the measurements, this does not allow further meaningful higher resolution refinement of this structure at this stage.

It will now be important to refine the helical bundle and loop structures of this receptor. This will require additional independent high resolution constraints for spatial approximations and interactions that involve this region. Higher resolution insights will also be key to extending our understanding of the proposed endogenous agonist mechanism for activation of this receptor (12). This may lead to identification of potential drugable pockets for small molecule ligands that could regulate receptors in this important family.


   FOOTNOTES
 
* This work was supported by National Institutes of Health Grant DK46577 (to L. J. M.), National Health and Medical Research Council of Australia (NHMRC) Grant 436780 (to P. M. S.), and by the Fiterman Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Back

1 An NHMRC Principal Research Fellow. Back

2 To whom correspondence should be addressed: Mayo Clinic, 13400 E. Shea Blvd., Scottsdale, AZ 85259. Tel.: 480-301-6650; Fax: 480-301-6969; E-mail: miller{at}mayo.edu.

3 The abbreviations used are: FRET, fluorescence resonance energy transfer; CHO, Chinese hamster ovary; KRH, Krebs-Ringer-HEPES. Back


   ACKNOWLEDGMENTS
 
We thank Laura-Ann Bruins and Renee M. Happs for excellent technical assistance and Evelyn Posthumus for secretarial assistance.



   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 EXPERIMENTAL PROCEDURES
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Ishihara, T., Nakamura, S., Kaziro, Y., Takahashi, T., Takahashi, K., and Nagata, S. (1991) EMBO J. 10, 1635-1641[Medline] [Order article via Infotrieve]
  2. Asmann, Y. W., Dong, M., Ganguli, S., Hadac, E. M., and Miller, L. J. (2000) Mol. Pharmacol. 58, 911-919[Abstract/Free Full Text]
  3. Lisenbee, C. S., Dong, M., and Miller, L. J. (2005) J. Biol. Chem. 280, 12330-12338[Abstract/Free Full Text]
  4. Grace, C. R., Perrin, M. H., DiGruccio, M. R., Miller, C. L., Rivier, J. E., Vale, W. W., and Riek, R. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 12836-12841[Abstract/Free Full Text]
  5. Tan, Y. V., Couvineau, A., Murail, S., Ceraudo, E., Neumann, J. M., Lacapere, J. J., and Laburthe, M. (2006) J. Biol. Chem. 281, 12792-12798[Abstract/Free Full Text]
  6. Dong, M., Lam, P. C., Gao, F., Hosohata, K., Pinon, D. I., Sexton, P. M., Abagyan, R., and Miller, L. J. (2007) Mol. Pharmacol. 72, 280-290[Abstract/Free Full Text]
  7. Harikumar, K. G., Hosohata, K., Pinon, D. I., and Miller, L. J. (2006) J. Biol. Chem. 281, 2543-2550[Abstract/Free Full Text]
  8. Dong, M., Asmann, Y. W., Zang, M., Pinon, D. I., and Miller, L. J. (2000) J. Biol. Chem. 275, 26032-26039[Abstract/Free Full Text]
  9. Dong, M., Li, Z., Zang, M., Pinon, D. I., Lybrand, T. P., and Miller, L. J. (2003) J. Biol. Chem. 278, 48300-48312[Abstract/Free Full Text]
  10. Dong, M., Pinon, D. I., Cox, R. F., and Miller, L. J. (2004) J. Biol. Chem. 279, 1167-1175[Abstract/Free Full Text]
  11. Dong, M., Zang, M., Pinon, D. I., Li, Z., Lybrand, T. P., and Miller, L. J. (2002) Mol. Endocrinol. 16, 2490-2501[Abstract/Free Full Text]
  12. Dong, M., Pinon, D. I., Asmann, Y. W., and Miller, L. J. (2006) Mol. Pharmacol. 70, 206-213[Abstract/Free Full Text]
  13. Powers, S. P., Pinon, D. I., and Miller, L. J. (1988) Int. J. Pept. Protein Res. 31, 429-434[Medline] [Order article via Infotrieve]
  14. Munson, P. J., and Rodbard, D. (1980) Anal. Biochem. 107, 220-239[CrossRef][Medline] [Order article via Infotrieve]
  15. Harikumar, K. G., Pinon, D. I., Wessels, W. S., Dawson, E. S., Lybrand, T. P., Prendergast, F. G., and Miller, L. J. (2004) Mol. Pharmacol. 65, 28-35[Abstract/Free Full Text]
  16. Harikumar, K. G., and Miller, L. J. (2005) J. Biol. Chem. 280, 18631-18635[Abstract/Free Full Text]
  17. Abagyan, R. A., Totrov, M. M., and Kuznetsov, D. A. (1994) J. Comp. Chem. 15, 488-506[CrossRef]
  18. Abagyan, R., and Totrov, M. (1994) J. Mol. Biol. 235, 983-1002[CrossRef][Medline] [Order article via Infotrieve]
  19. Metropolis, N. R. A., Rosenbluth, M. N., Teller, A. H. and Teller, E. (1953) J. Chem. Phys. 21, 1087-1092[CrossRef]
  20. Fernandez-Recio, J., Totrov, M., and Abagyan, R. (2003) Proteins 52, 113-117[CrossRef][Medline] [Order article via Infotrieve]
  21. Clore, G. M., Nilges, M., Brunger, A., and Gronenborn, A. M. (1988) Eur. J. Biochem. 171, 479-484[Medline] [Order article via Infotrieve]
  22. Dong, M., Li, Z., Pinon, D. I., Lybrand, T. P., and Miller, L. J. (2004) J. Biol. Chem. 279, 2894-2903[Abstract/Free Full Text]
  23. Hooft, R. W., Sander, C., Scharf, M., and Vriend, G. (1996) Comput. Appl. Biosci. 12, 525-529[Abstract/Free Full Text]
  24. Laskowski, R. A., Moss, D. S., and Thornton, J. M. (1993) J. Mol. Biol. 231, 1049-1067[CrossRef][Medline] [Order article via Infotrieve]
  25. Grace, C. R., Perrin, M. H., Gulyas, J., Digruccio, M. R., Cantle, J. P., Rivier, J. E., Vale, W. W., and Riek, R. (2007) Proc. Natl. Acad. Sci. U. S. A. 104, 4858-4863[Abstract/Free Full Text]
  26. Sun, C., Song, D., Davis-Taber, R. A., Barrett, L. W., Scott, V. E., Richardson, P. L., Pereda-Lopez, A., Uchic, M. E., Solomon, L. R., Lake, M. R., Walter, K. A., Hajduk, P. J., and Olejniczak, E. T. (2007) Proc. Natl. Acad. Sci. U. S. A. 104, 7875-7880[Abstract/Free Full Text]
  27. Dong, M., Pinon, D. I., Cox, R. F., and Miller, L. J. (2004) J. Biol. Chem. 279, 31177-31182[Abstract/Free Full Text]
  28. Bisello, A., Adams, A. E., Mierke, D. F., Pellegrini, M., Rosenblatt, M., Suva, L. J., and Chorev, M. (1998) J. Biol. Chem. 273, 22498-22505[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 Mol. Pharmacol.Home page
M. Dong, P. C.-H. Lam, D. I. Pinon, P. M. Sexton, R. Abagyan, and L. J. Miller
Spatial Approximation between Secretin Residue Five and the Third Extracellular Loop of Its Receptor Provides New Insight into the Molecular Basis of Natural Agonist Binding
Mol. Pharmacol., August 1, 2008; 74(2): 413 - 422.
[Abstract] [Full Text] [PDF]

Home page
 Mol. Endocrinol.Home page
M. Dong, F. Gao, D. I. Pinon, and L. J. Miller
Insights into the Structural Basis of Endogenous Agonist Activation of Family B G Protein-Coupled Receptors
Mol. Endocrinol., June 1, 2008; 22(6): 1489 - 1499.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
282/45/32834    most recent
M704563200v1
Right arrow Alert me when this article is cited
Right arrow