Versatility and Differential Roles of Cysteine Residues in Human Prostacyclin Receptor Structure and Function*

Prostacyclin plays important roles in vascular homeostasis, promoting vasodilatation and inhibiting platelet thrombus formation. Previous studies have shown that three of six cytoplasmic cysteines, particularly those within the C-terminal tail, serve as important lipidation sites and are differentially conjugated to palmitoyl and isoprenyl groups (Miggin, S. M., Lawler, O. A., and Kinsella, B. T. (2003) J. Biol. Chem. 278, 6947-6958). Here we report distinctive roles for extracellular- and transmembrane-located cysteine residues in human prostacyclin receptor structure-function. Within the extracellular domain, all cysteines (4 of 4) appear to be involved in disulfide bonding interactions (i.e. a highly conserved Cys-92-Cys-170 bond and a putative non-conserved Cys-5-Cys-165 bond), and within the transmembrane (TM) region there are several cysteines (3 of 8) that maintain critical hydrogen bonding interactions (Cys-118 (TMIII), Cys-251 (TMVI), and Cys-202 (TMV)). This study highlights the necessity of sulfhydryl (SH) groups in maintaining the structural integrity of the human prostacyclin receptor, as 7 of 12 extracellular and transmembrane cysteines studied were found to be differentially indispensable for receptor binding, activation, and/or trafficking. Moreover, these results also demonstrate the versatility and reactivity of these cysteine residues within different receptor environments, that is, extracellular (disulfide bonds), transmembrane (H-bonds), and cytoplasmic (lipid conjugation).

Prostacyclin and its receptor play key roles in vascular smooth muscle relaxation and inhibition of platelet aggregation. A host of studies, including IP 2 receptor knock-out (IP Ϫ/Ϫ ) mice, has implicated dysfunctional IP activity in numerous cardiovascular abnormalities, including thrombosis, myocardial infarction, stroke, hypertension, and atherosclerosis (2)(3)(4)(5). Despite significant progress in our understanding of G-protein coupled receptors (GPCRs) in general, the details of human prostacyclin receptor structure-function remains largely unknown.
Cysteine chemistry is both fascinating and intriguing. The sulfhydryl or thiol (SH) reactive groups of this amino acid are very susceptible to oxidation and can readily form stable dimers (i.e. disulfide S-S bridges), which play important roles in the organization and maintenance of protein tertiary structure. Somewhat analogous to hydroxyl groups (OH) found on serines, sulfhydryl (SH) side chains are also polar and can participate in hydrogen bonding interactions and can additionally coordinate trace metals (e.g. zinc) (6). Sulfur atoms are also quite nucleophilic and react readily with electrophilic molecules to form a variety of thiol-linked derivatives (e.g. thioethers, thioesters, and thioacetals). Thus, cysteine side chains are common sites for various biological coupling and conjugation reactions, including palmitoylation, isoprenylation, disulfide cross-linking, and thiol-disulfide exchange (7)(8)(9). In this study our goal was to assess the structural and functional contributions of cysteines within the human prostacyclin receptor (hIP) receptor. Early mutagenesis studies in bovine rhodopsin (using Cys3 Ser mutations) revealed the importance of amino acids Cys-110 and Cys-187, which are essential in the formation of normal functional rhodopsin protein (10). It was later shown that these two highly conserved cysteine residues form a disulfide bond in the extracellular domain (11). Further mutagenesis and mass spectrometry confirmed the presence of this disulfide bond and its importance in stabilizing rhodopsin (12,13). Because of the initial discovery of this imperative disulfide bond within rhodopsin, many other investigations have gone on to demonstrate the presence and importance of similar S-S bonds in other GPCRs (14 -18). Mutagenesis of the thromboxane receptor (TP) at the conserved extracellular S-S cysteine positions (Cys-105 and Cys-183) revealed decreased binding affinity and low amplitude calcium signaling (19,20). Interestingly, in addition to this conserved disulfide bridge, further studies on the ␤2-adrenergic receptor as well as the gonadotropin-releasing hormone receptor have suggested the presence of a second, non-conserved disulfide bond (8,21).
The C-terminal tail of the human prostacyclin receptor contains six resident cysteines, which have been extensively studied. Five of these cysteine residues are located within CAAX consensus motifs thought to comprise sites for lipid anchoring (e.g. palmitoylation or isoprenylation). Given that a good deal of evidence has already been presented for both palmitoylation as well as isoprenylation at these C-terminal cysteine sites on the human prostacyclin receptor (1, 7), we have opted to exclude * This work was supported by grants from the NHLBI, National Institutes of Health (to J. H.) and the American Heart Association (to J. H. and J. S.). 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. 1 To whom correspondence should be addressed: Dept this region from our current investigation. The remaining 12 cysteine residues within the extracellular and transmembrane domains of the hIP receptor were individually converted to alanine using site-directed PCR mutagenesis. We report that in addition to the highly conserved disulfide interaction between Cys-92 (top of transmembrane (TM) III) and Cys-170 (exoloop  2), an additional, putative non-conserved disulfide bridge may exist between Cys-5 (N terminus) and Cys-165 (exoloop 2), within the extracellular domain of the hIP, as well as other prostanoid receptors (hDP and hEP2). Moreover, this interaction seems to serve a critical, yet distinct purpose compared with its conserved counterpart. Furthermore, in the transmembrane domain Cys-118 (TMIII), Cys-202 (TMV), and Cys-251 (TMVI) were found to be necessary for preserving normal receptor binding affinity, activation capacity, and/or cell-surface expression through probable hydrogen bonding networks. Such observations provide further insights into the molecular functioning of the hIP receptor and contribute to the mechanistic understanding of how the hIP protein is stabilized during the continuum of conformational changes that occur upon agonist-induced activation.
Cysteine-to-Alanine Mutations-Twelve (within the extracellular and transmembrane domains) of a total of 18 cysteine residues were individually mutated to alanine (Fig. 1A). The choice of amino acid substitution was critical as cysteine-toserine (Cys3 Ser) mutations in rhodopsin have been shown to have greater adverse effects on receptor conformation than cysteine-to-alanine (Cys3 Ala) substitutions (10,12). Additionally, the C-terminal residues were not included in this study because these have been extensively characterized (1). Human IP cDNA tagged at the C terminus with the 1D4 epitope tag (last 14 amino acids of rhodopsin) was cloned into the plasmid vector pMT4, and point mutations were generated using conventional methods of PCR mutagenesis as previously described (23). Ten microliters of the PCR product was used to transform competent DH5␣ Escherichia coli cells (ϳ2 ϫ 10 9 cells) followed by DNA extraction from selected clones. Large plasmid preparations were performed using Wizard Plus Maxiprep kits (Promega, Madison, WI), and all mutant constructs were confirmed via PCR DNA dideoxynucleotide chain-termination sequencing (Molecular Biology Core Facility, Dartmouth Medical School, Hanover, NH).
Transfection of COS-1 Cells and Membrane Preparations-Transient transfections were performed on COS-1 cells as previously described (24). In brief, cells were incubated in DNA (10 or 20 g/plate) in diethylaminoethyldextran (DEAE-dextran; Sigma) (0.2 mg/ml Dulbecco's modified Eagle's medium) and harvested 72 h post-transfection. Transfected COS-1 cells were washed in phosphate-buffered saline and harvested. Vortexing (providing shear forces) for 3 min in sucrose (0.25 M) was fol-lowed by low speed spin (ϳ1260 ϫ g) for 5 min, and the supernatant was collected. After a high speed centrifugation (ϳ30,000 ϫ g for 15 min) the pellet was then washed twice in 1ϫ HEM (20 mM Hepes, pH 7.4, 1.5 mM EGTA, and 12.5 mM MgCl 2 ) followed by re-suspension in 1ϫ HEM containing 10% glycerol and stored at Ϫ70°C (25). Either Bradford or BCA protein assays were performed to quantitate membrane proteins.
Ligand Binding Affinity-Ligand binding characteristics for the expressed receptors were initially determined through a series of competition binding assays using the radiolabeled ligand [ 3 H]iloprost. Analysis involved construction of reaction mixtures (in duplicate wells) containing 50 g of membrane, HEM buffer, and 15 nM [ 3 H]iloprost along with 1 of 11 different concentrations of cold (nonradiolabeled) iloprost ranging from 10 M to 0.1 nM. After 1.5 h of incubation at 4°C, reactions were stopped by the addition of ice-cold 10 mM Tris/HCl buffer, pH 7.4, and filtered onto Whatman GF/C glass-fiber filters using a Brandel cell harvester. The filters were washed 5 times with ice-cold Tris/HCl buffer, and radioactivity was measured in the presence of 5 ml of Ecoscint TM H scintillation fluid (National Diagnostics, Atlanta, GA). Nonspecific binding was determined by the addition of a 500-fold excess of nonradiolabeled iloprost. Data were analyzed using GraphPad Prism software (GraphPad software, Inc., San Diego, CA). IC 50 values were converted to K i using the Cheng-Prusoff equation, and K i values were expressed as a mean Ϯ S.E. For saturation binding experiments to determine B max and K D , the concentration of [ 3 H]iloprost was varied from 1 to 100 nM. Nonspecific binding was determined by the addition of a 500-fold excess of nonradiolabeled iloprost. Data were analyzed using GraphPad Prism software (GraphPad Software). Analysis of variance and Student's t tests were used to determine significant differences (p Ͻ 0.05).
Receptor Activation cAMP Determination-The wild-type and cysteine mutant constructs were analyzed for signal transduction capabilities. COS-1 cells were transiently transfected with 2 g of receptor DNA in 25-mm plates as described above. [ 3 H]cAMP was used in competition for a cAMP-binding protein against known concentrations of nonradiolabeled cAMP followed by determination of the unknowns. The reaction was allowed to proceed for 2 h at 4°C. Charcoal was used to remove excess unbound cAMP. Samples were counted in 5 ml of Ecoscint TM H (National Diagnostics). Results were analyzed with GraphPad Prism software. For the dose response, a non-linear, curve-fitting program (GraphPad Prism) was used, and the EC 50 (mean Ϯ S.E.) was determined for wildtype hIP1D4 and mutant constructs. Analysis of variance and Student's t tests were used to determine statistically significant differences (p Ͻ 0.05).
Molecular Modeling of hIP Receptor-A theoretical, threedimensional homology model of the seven transmembrane ␣-helices of the hIP was constructed using the internet-based protein-modeling server, SWISS-MODEL (GlaxoSmithKline) (26). The homology model was generated using the 2.8-Å-resolution x-ray crystallographic structure of bovine rhodopsin as the template (PDB code 1HZX). The transmembrane domains were energy-minimized, utilizing the Gromos96 force field to improve the stereochemistry of the model and remove unfavorable clashes (SWISS-MODEL). Polypeptide chains corresponding to the extracellular and cytoplasmic loop regions were manually constructed. The large C-terminal tail region of the hIP was excluded from this particular model. Initial torsion angles were derived from full polypeptide secondary structure prediction using the JPred multiple sequence alignment consensus server (27). The subsequent loop peptides were energyminimized using the NAMD molecular dynamics simulator (28). Resulting structures were sequentially attached to the transmembrane homology model. Compiled structures were energy-minimized using NAMD, including constraints for either 1) no extracellular disulfide bonds, 2) a single, conserved extracellular disulfide bond (Cys-92-Cys-170), or 3) dual extracellular disulfide bonds (Cys-92-Cys-170 and Cys-5-Cys-165). This latter structure, with dual extracellular (disulfide) constraints, yielded the lowest energy conformation and was utilized as the preferred hIP receptor model for subsequent comparisons. Although our model was a useful first pass tool in providing preliminary insights and predictions of sulfhydryl hIP structure function, biochemical and molecular pharmacological techniques were necessary to confirm our hypothesis.

RESULTS
Our current knowledge of the hIP receptor is limited. Of particular interest has been the role of cysteine residues within the extracellular and TM domains of the protein. Cysteines contain a highly nucleophilic sulfhydryl or thiol (SH) side chain that is capable of acting as a nucleophilic catalyst. Moreover, with a pK a of ϳ8, the chemical reactivity of cysteine sulfhydryls is easily modified by environmental conditions. Within a reducing environment, cysteine residues may be involved in bioconjugation reactions, whereas conversely, within an oxidizing environment, side chains may dimerize to form disulfide (S-S) linkages. These latter interactions play an invaluable role in intramolecular cross-linking of proteins that can increase molecular stability within harsh environments as well as confer resistance to proteolytic cleavage. We wished to elucidate the differential contributions of conserved and non-conserved cysteines within the transmembrane region of the hIP receptor as well as explore the potential existence of a second, putative non-conserved disulfide bond within the extracellular domain.
Extracellular Cysteines Are Required for Receptor Trafficking, Binding, and Activation-Site-directed mutagenesis was performed targeting all cysteine residues present within the hIP receptor, excluding those located in the C-terminal tail. Thus, 12 of the total 18 cysteine residues (Table 1, Fig. 1A) were individually mutated to alanine (Cys3 Ala) and assessed for functional effects on agonist binding (affinity) and receptor activation (potency), as measured via cAMP production. Binding affinity (K i ) could not be detected using competition binding ( 2B) be determined. These binding affinities are estimates due an inability to saturate the expressed receptors (Fig. 2B). They nevertheless indicate severe defects in binding affinity. Related to the severe binding defects, potency (EC 50 ) values could not be determined due to negligible efficacy (Table 1). These results demonstrate the severity of functional perturbation with mutation of any one of the four extracellular cysteine residues within the hIP and lend preliminary support to the notion that 1) the highly conserved Cys-92 and Cys-170 may form a disulfide bond (Fig. 1B), and 2) an alternative interaction may exist between the remaining extracellular cysteines Cys-5 and Cys-165 (perhaps a second disulfide bond), which appears to be equally important for hIP receptor binding and activation.
Estimates of cell surface trafficking using saturation binding on plasma membrane preparations confirmed decreased plasma membrane expression for all four extracellular

H]iloprost and 1 g of DNA/ml transfection solution) and dose-response (EC 50 ), characteristics for wild-type hIP and cysteine-to-alanine mutations
Bold indicates critical cysteine residues for hIP structure-function. ND, not detectable. DECEMBER 1, 2006 • VOLUME 281 • NUMBER 48

JOURNAL OF BIOLOGICAL CHEMISTRY 37229
Cys3 Ala substitutions compared with wild-type hIP ( Table 1). The additional use of confocal microscopy corroborated reduced cell surface expression and demonstrated defective trafficking with endoplasmic reticulum co-localization (calnexin) and retention (Fig. 2C). Wild-type hIP predominantly trafficked to the cell surface as observed with the red fluorescence surrounding the cell in the phase contrast micrograph overlay (Fig. 2C). Because of the nature of overexpression systems, some wild-type hIP was also detectable in the endoplasmic reticulum as observed with co-localization with the green fluorescence. Both C5A and C165A showed a marked reticular pattern co-localized predominantly to the endoplasmic reticulum, with reduced cell surface. C92A and C170A were also found predominantly in an endoplasmic reticulum location (perinuclear), again with reduced cell surface expression. Thus, in conjunction with the findings from previous binding and activation studies, these results also show defects for the four extracellular cysteines, suggesting that there may be dual functional connections between the highly conserved Cys-92-Cys-170 and the less conserved Cys-5-Cys-165. Sequence Alignments and Treatment with Reducing Agents Suggest Formation of Dual Extracellular Disulfide Bonds (Cys-5 to Cys-165 and Cys-92 to Cys-170)-In addition to our initial experimental findings, further supportive evidence that a second disulfide bond exists between the non-conserved extracellular cysteine residues Cys-5 and Cys-165 can be found in human prostanoid receptor sequence alignments (Fig. 1B), which shows that both residues are either exclusively present or exclusively absent within the various prostanoid receptors. One of the cysteines (Cys-5 or Cys-165) is always found in the presence of the other. Thus, at equivalent positions within the human IP, EP2, and DP receptors, both the Cys-5 and Cys-165 residues are found, whereas conversely, at equivalent positions within the human EP1, EP3, EP4, TP, and FP receptors, both cysteines are absent (Fig. 1B). In contrast, Cys-92 and Cys-170 are conserved at equivalent positions among all the prostanoid receptors, suggesting that Cys-92 interacts with Cys-170, and (when present) Cys-5 interacts with Cys-165.
If critical disulfide (S-S) bonds are present within the extracellular domain of the hIP receptor, then chemical reduction of the wild-type receptor should be associated with defects in structure (folding-conformation-stability) and function (binding and activation). Competition binding was performed in the presence of increasing concentrations of 1.4 -140 mM ␤-mercaptoethanol (␤-ME; Fig. 3A) and 1-100 mM 1,4-dithiothreitol (Fig. 3B), which break disulfide (S-S) bonds and maintain sulfhydryl (SH) groups in the reduced state. As can be viewed in Fig. 3, a sequential decrease in agonist binding was observed for wild-type hIP receptors treated with increasing concentrations of either 1,4-dithiothreitol or ␤-ME. The significant difference and decline in raw counts (specific binding) in response to increasingly higher concentrations of reducing agent is indicative of an increasing pool of defective IP (i.e. reduced S-S bonds to SH). Clear sigmoidal competition binding curves suggest a population of receptors in which the disulfide bond remained intact. As shown, the IC 50 values, measured in log-fold molar concentrations of iloprost agonist, log   Table 2 are the corresponding results for K D . Panel C, corresponding confocal microscopy (63ϫ resolution) images for each mutation showing predominant membrane trafficking only for wild-type protein. The hIP receptors, both wild type and mutants, are in red (1D4 monoclonal antibody). The endoplasmic reticulum is shown in green (anti-calnexin antibody), and the overlay additionally has blue nuclear staining (4Ј,6-diamidino-2-phenylindole) and a phase contrast microscopic image of the cell to localize the cells perimeter. Red arrows are used to localize areas of cell surface membrane. DECEMBER 1, 2006 • VOLUME 281 • NUMBER 48

JOURNAL OF BIOLOGICAL CHEMISTRY 37231
Additional evidence, albeit indirect, for the presence of dual disulfide bonds can be gleaned from the observation that the low levels of specific binding for both the C92A and C170A mutations quickly dissipates upon the addition of ␤-mercaptoethanol (Fig. 5), implying further reduction of a secondary disulfide bridge. The culmination of these results in combination with saturation binding and confocal microscopy suggests that both putative disulfide bonds are important.

Rhodopsin-based Computer Modeling Supports the Formation of Dual Disulfide Bonds with Differential Function-In
addition to experimental-functional analyses, computer-assisted modeling was used to assess whether such dual bonding was structurally feasible within an energy-minimized model of the hIP (see Fig. 7). Despite evident limitations in the manual construction and the addition of loop regions to our existing three-dimensional homology model, the composite, energyminimized structure (i.e. TM domain plus N terminus, 3 exoloops, and 3 cytoloops) suggested that the presence of a secondary disulfide bond was structurally feasible. Energy minimization placed the freely rotating N terminus between exoloops 2 and 3, preserving the conserved disulfide bridge TABLE 2 Binding affinity (K D ) characteristics for wild-type hIP and the four extracellular cysteine-to-alanine mutations in which binding affinity was not able to be detected using standard competition binding (Table 1) Receptor numbers were increased with increased DNA transfected (2 mg of DNA/ml of transfection solution) and increasing concentrations of [ 3 H]iloprost (saturation binding was performed in duplicate using 6 increasing concentrations) used. Saturation binding curves are shown in Fig. 2B. Bold indicates critical cysteine residues for hIP structure-function.    between Cys-92 (TMIII) and Cys-170 (exoloop 2) and facilitating the formation of a second stable disulfide linkage between Cys-5 (N terminus) and Cys-165 (exoloop 2) (see Fig. 7A). Additional molecular dynamics and secondary structural analyses are required for further refinement of the composite structure. Nevertheless, this homology model provides a basis for additional structural insights and predictive capabilities. Validation of the composite model can be achieved through subsequent experimental testing (mutagenesis studies) of model-based predictions.
Compensatory Mutation Partially Restores Binding Affinity via Non-conserved Disulfide Bond Reformation-Based upon initial assessments of the putative Cys-5-Cys-165 disulfide bridge within our homology model, in the context of the C5A mutation, conversion of an adjacent residue to cysteine might also result in covalent formation of a disulfide bond with Cys-165 (Fig. 6A). Furthermore, restoration of receptor function would most likely only occur in the presence of a covalent S-S interaction, because hydrogen bonding (of SH) would be unlikely to support such drastic changes at both amino acid positions. Thus, the adjacent residue Arg-6 was chosen as the candidate amino acid to be converted to cysteine (C5A-R6C) in an attempt to reinstate the uncoupled S-S linkage resulting from the initial C5A mutation. Competition binding analysis revealed a return of binding affinity with the compensatory mutation (C5A-R6C, K i ϭ 8.2 Ϯ 4.5 nM, n ϭ 3; wild-type hIP K i ϭ 9.6 Ϯ 2.8 nM n ϭ 4) in comparison to the undetectable K i value from the original C5A mutant receptor (Fig. 6B). However, the maximal binding was reduced considerably, suggesting only a small population had regained a disulfide bond. Such results are suggestive of a stabilizing disulfide interaction between Cys-6 and Cys-165, which mimics (incompletely) the putative non-conserved Cys-5-Cys-165 disulfide bond found in the native hIP receptor. To further support the restoration of a disulfide bond, reduction of C5A-R6C with 140 mM ␤-ME was performed with resulting reduction and a shift to the right of the C5A-R6C graph. The highest concentration of cold iloprost addition was now at background vector control levels 31.9 Ϯ 2.2 raw counts (cpm) (n ϭ 10).
Cys-118 in the Highly Conserved ERX Motif Affects Binding and Activation-In examining cysteines within the transmembrane domain of the hIP, Cys-118 is adjacent to the highly conserved endoplasmic reticulum motif located on the cytoplasmic side of TMIII. Sequence alignments for all 961 human GPCRs identified in the GPCRDB (GPCR database) showed that Tyr (56%), Phe1%), and Cys (4%) were the most common amino acids at this position. The C118A mutant was severely defective with undetectable binding with competition binding and poorly detectable saturation binding in addition to undetectable activation values (Table 1). Receptor expression was also significantly diminished (estimated B max C118A ϭ 0.5 Ϯ 0.2 pmol/mg of membrane protein, p Ͻ 0.05). Molecular modeling analysis indicated that Cys-118 is located at the beginning of exoloop 2 within a solvent-accessible cavity between TMIII and TMV. In certain conformations the sulfhydryl side chain may participate in a critical intra-␣-helical stabilizing H-bonding or thioester interaction with this residue (Fig. 7B).
Cys-251 May Provide a Stabilizing Interaction with TMVII-Cys-251 had adverse effects on hIP receptor binding and activation. It is completely conserved throughout all human prostanoid receptors (Fig. 1B). Although competition binding was unable to determine binding affinity (reduced affinity and expression), saturation binding detected a K D of 44.5 Ϯ 6.5 nM (n ϭ 3, p Ͻ 0.001), consistent with the reduced potency (Table  1). Upon examination of the Cys-251 residue within our previously developed three-dimensional molecular homology model of the hIP binding domain (25), it appears that the Cys-251 residue may be tethered to key residues within the immediate ligand binding pocket of the hIP receptor. As shown in Fig. 7C, the reactive (SH) side chain of Cys-251 (TMVI) forms a hydrogen-bond network in close proximity to Phe-278 and the ligand binding pocket. In previous studies Phe-278 was shown to directly interact with receptor-bound ligand and inhibit ligand dissociation (29).
C202A Exclusively Affects Expression-Cys-202 is nearly conserved across all the human prostanoid receptors, except for the hTP (Fig. 1B). It is located at the bottom of TMV near the beginning of the third cytoplasmic loop (cytoloop 3) (Fig. 7D). Although the C202A appeared to be wild-type-like in binding and activation, there was a significant decrease in plasma membrane receptor expression (B max Cys-202 ϭ 0.4 Ϯ 0.1, p Ͻ 0.05) ( Table 1). Within our homology model, the sulfhydryl (SH) group of Cys-202 appears to form a hydrogen-bonding network with Asn-203 and Ala-198 (Fig. 7D).

DISCUSSION
Prostacyclin (PGI 2 ) analogs are now widely used for the treatment of pulmonary hypertension (30 -33). More recently it has also been suggested that prostacyclin may also be useful as a therapeutic agent in treating lung (34) and colon cancers (35,36). Recent clinical studies showing increased cardiovascular events arising from selective cyclooxygenase-2 inhibition have also highlighted the important role of PGI 2 in cardiovascular protection (37,38). Furthermore, in vivo studies of prostacyclin receptor knock-out (IP Ϫ/Ϫ ) mice have shown that PGI 2 is a key modulator of platelet-vascular interactions and may play a role in protection against atherosclerosis (2,3). Thus, the culmination of these studies has lent a great deal of insight into the important roles of PGI 2 activity as 1) an effective pharmacological agent and 2) a critical physiological-homeostatic mediator. However, we are now only beginning to decipher the molecular structure and function of the G-protein-coupled receptor (i.e. hIP) that mediates these imperative functions of PGI 2 .
A Putative Second Disulfide Bond within the hIP Receptor-The goal of this study was to determine the functional roles played by extracellular-and transmembrane-located cysteine residues within the hIP to improve our molecular understanding of this important cardiovascular protein. Strong evidence has supported the existence and necessity (for proper ligand binding) of the conserved disulfide bond in other prostanoid receptors, such as the thromboxane receptor (19,20). Interestingly, mutations of the Cys-170 equivalent cysteine in the EP3 suggested that it is not critical for agonist binding and activation (40). In our present study we look at the differential effects on binding, activation, and expression of these putative conserved and non-conserved disulfide bonds within the extracellular domain of the hIP receptor. We demonstrate that the conserved disulfide bond (Cys-92-Cys-170) is insufficient for maintaining proper ligand binding characteristics, and an additional extracellular bond (Cys-5-Cys-165) may be required for hIP stability. Support for this notion of a second disulfide bond within the extracellular domain of the hIP could be observed upon sequential treatment with ␤-ME. Furthermore, the lack of significant change in either IC 50 or EC 50 values (only a change in specific counts) upon treatment with reducing agents strongly supports a secondary disulfide requirement for stability. Potential explanations for such added structural stringency could be the unusual orientation of receptor-bound PGI 2 (25), which maintains dual pentagonal rings facing TMI and TMII, as opposed to other similarly structured class A ligands (e.g. 11-cis-retinal and epinephrine), which face away from this region toward TMV and TMVI. Perhaps maintenance of this unique binding pocket requires an additional stabilizing bond extending from the N terminus (C5) to the second extracellular loop (Cys-165). The presence of putative dual disulfide bonds in the hIP has been suggested only recently (as a means of receptor oligomerization) in published work using differential tagging and Western analysis (41). Recent work has also demonstrated potential important roles for GPCR and hIP dimerization (42,43). Rather than direct disulfide bonding, the non-conserved extracellular free sulfhydryl groups may be involved in critical homo-or heterodimerization. Alternatively, an added disulfide bridge may be necessary to allow specific conformational changes required for receptor coupling and signaling via the G s -cAMP pathway. Although the underlying mechanism remains unclear, it is interesting to note that the three prostanoid receptors containing the putative, dual-bonding extracellular cysteines, namely the IP, DP, and EP2 receptors, all couple predominately to G s and modulate cAMP, whereas four of the remaining five prostanoid receptors that do not harbor the second S-S bonding cysteines signal through G q/11 (EP1, FP, TP) or through G i/o (EP3). The last member of this prostanoid receptor group (EP4) does, however, signal through G s .

Cysteines in Critical Hydrogen Bonding Networks and Lipid-anchoring Bioconjugations (Palmitoylation-Isoprenylation)-
The three critical transmembrane cysteines Cys-118, Cys-202, and Cys-251 are all highly conserved amino acids. Both Cys-202 and Cys-251 are 88 and 100% conserved across the prostanoid receptors, respectively. Residue Cys-118, although only mildly (38%) conserved across the prostanoid family of receptors, is located within the common ER(Y/C) motif, which is highly preserved in class A GPCRs. By virtue of the tightly packed ␣-helical environment, the requirement for membrane flexibility and conformational movement for receptor activation, disulfide (S-S) bonding, or cysteine bioconjugations are not commonly found within the transmembrane domain. As a result, hydrogen bonding networks are the most likely molecular interactions to form between cysteines and other amino acids within the transmembrane domain of the hIP. Our simple first pass homology model provides some support for such potential transmembrane H-bonding interactions. However, further detailed biophysical analysis would be required to definitively confirm such interactions.
Palmitoylation and isoprenylation are forms of post-translational modification in which hydrophobic carbon moieties (i.e. C-16 palmitic acid, or C-15, C-20 isoprenoids) are covalently attached to cysteine residues of membrane proteins. Five of the remaining six cysteines located along the C-terminal tail of the hIP are positioned within lipid-anchoring (palmitoylation-isoprenylation) CAAX consensus motifs. The first of these motifs begins at amino acid position 308 and contains three cysteines, CCLC (Cys-Cys-Leu-Cys) (Fig. 1A). This motif is highly conserved across class A GPCRs and provides a site for lipid anchoring, which tethers part of the C-terminal tail to the cytoplasmic membrane, forming a fourth intracellular loop. In the hIP, this site is thought to be predominately conjugated with palmitic acid (palmitoylation) at both Cys-308 and Cys-311. There also exists a second CAAX motif at the very end of the C terminus, starting at residue 383, CSLC (Cys-Ser-Leu-Cys). This particular motif is thought to be predominately isoprenylated (at Cys-383) and has been purported to result in a fifth intracellular loop, which may modify G-protein-coupling capacity (1,44). Although excluded from our particular investigation, the importance of these remaining six C-terminal cysteines on hIP structure and function should not be dismissed.
The culmination of this study highlights the importance and versatility of extracellular and transmembrane resident cysteines in hIP stability and function. As demonstrated, a number of highly conserved cysteines within the transmembrane domain (e.g. Cys-118, Cys-202, and Cys-251) of the hIP recep-tor were found to be critical for proper receptor function (i.e. binding, activation, and expression). Furthermore, all of the cysteine residues located within the extracellular domain (e.g. Cys-5, Cys-92, Cys-165, and Cys-170) were found to carry out significant and differential roles in the maintenance of hIP structural integrity and activity. Evidence from site-directed mutagenesis, functional binding activation assays, treatment with thiol-reducing agents, compensatory mutations, sequence homology comparisons, and computer-assisted modeling all support the notion of dual disulfide bridges within the extracellular domain of the hIP; that is, a conserved linkage between Cys-92 and Cys-170 and a putative non-conserved linkage between Cys-5 and Cys-165.
Considering the importance of the human prostacyclin receptor in cardiovascular disease along with the susceptibility to reductive-oxidative stressors of extracellular cysteines within the hIP receptor protein, one could speculate that the redox environment found within blood vessels may play a critical role in both the normal and pathophysiological structurefunction of the human prostacyclin receptor. Oxidative stress has been implicated in numerous cardiovascular abnormalities, including hypertension, atherosclerosis, and restenosis after angioplasty (39,45), and oxidation of hIP proteins within the vasculature could alter structure (e.g. inducing aberrant disulfide cross-linking) and inhibit function. Furthermore, naturally occurring single nucleotide polymorphisms to and from cysteine (Xaa3 Cys and Cys3 Xaa), especially within the extracellular domain of the receptor, could also have drastic effects given the importance of these amino acids in upholding conformational integrity. Currently none have been found in the hIP; however, mutations at equivalent sites in rhodopsin (Cys-110 and Cys-187) have been found to lead to the retinal degenerative disease retinitis pigmentosa (22).
This study identifies critical cysteine residues within the extracellular and transmembrane domains of the hIP that, by nature of their versatility, are necessary for proper receptor function. In addition to lipid bioconjugation, strong evidence supports the existence of dual disulfide bonds within the extracellular region of the hIP as well as critical transmembrane hydrogen bonding. All serve differential roles in maintaining proper receptor configuration for binding affinity, receptor activation, cell surface expression, and trafficking.