The Function of Extracellular Cysteines in the Human Adrenomedullin Receptor

Home , Acetylcholine receptor, Thyrotropin receptor

S25

Original Article

Kenji KUWASAKO, Kazuo KITAMURA, Tomohiko UEMURA, Yasuko NAGOSHI, Johji KATO, and Tanenao ETO

When co-expressed with receptor activity-modifying protein (RAMP) 2, calcitonin receptor-like receptor (CRLR) functions as an adrenomedullin (AM) receptor (CRLR/RAMP2). In the present study, we examined the function of the cysteine (C) residues in the extracellular loops of human (h)CRLR (C212, C225 and C282) and in the extracellular domain of hRAMP2 (C68, C84, C99 and C131). Using site-directed mutagenesis, the cysteine residues were substituted, one at a time, with alanine (A). Co-expression in HEK293 cells of hRAMP2 with the hCRLR C212A or C282A mutant signiﬁcantly reduced the 50% of effective concentration

(EC50) for AM-evoked cyclic adenosine monophosphate (cAMP) production, despite full cell surface expression of the receptor heterodimer. Co-expression of the C225A mutant had no effect on [125I]AM binding or receptor signaling. These results suggest that the cysteine residues in the first (C212) and the second (C282) extracellular loops form a disulfide bond that is important for stabilizing the receptor in the correct conformation for ligand binding and activation. Cells expressing hCRLR with an hRAMP2 mutant (C68A, C84A, C99A or C131A) showed no specific AM binding or AM-stimulated cAMP accumulation. Though abundant in the intracellular compartment, these receptors were not detected at the cell surface, suggesting that all four cysteine residues are essential for efficient transport to the plasma membrane. Cysteine residues in the extracellular loops of hCRLR and in the extracellular domain of hRAMP2 thus appear to play distinct roles in the cell surface expression and function of the receptor heterodimer. (Hypertens Res 2003; 26 (Suppl): S25ÐS31)

Key Words: adrenomedullin, receptor activity-modifying protein, calcitonin receptor-like receptor

(6–9). Introduction Adrenomedullin (AM) receptors were recently shown to be heterodimers comprised of a novel accessory protein, the G protein-coupled receptors (GPCRs), which are character- receptor activity-modifying protein (RAMP), and an orphan ized in part by their seven transmembrane domains, consti- receptor, the calcitonin receptor-like receptor (CRLR) (10). tute a large family of cell surface regulatory molecules (1). When CRLR is co-transfected with RAMP2 or -3, the two Evidence derived from biochemical and site-directed muta- proteins are transported together to the plasma membrane genesis studies indicates that a disulfide bond between two where they function as an AM-specific receptor (CRLR/ cysteines in the first and second extracellular loops of these RAMP2 or CRLR/RAMP3). The functionality of AM recep- receptors is crucial for their normal cell surface expression tors, which mediate both intracellular cyclic adenosine (2–5) and for stabilization of the ligand-binding pocket monophosphate (cAMP) production and calcium mobiliza-

From the First Department of Internal Medicine, Miyazaki Medical College, Miyazaki, Japan. This study was supported in part by the Grants-in Aids for Scientific Research on Priority Areas and for the 21st Century COE Program (Life Science) from the Ministry of Education, Culture, Sports, Science and Technology, Japan. Address for Reprints: Kenji Kuwasako, M.D., Ph.D., First Department of Internal Medicine, Miyazaki Medical College, 5200 Kihara, Kiyotake, Miyazaki 889–1692, Japan. E-mail: [email protected] Received July 25, 2002; Accepted in revised form September 2, 2002. S26 Hypertens Res Vol. 26, Suppl (2003) tion (11), suggests CRLR is a member of the GPCR super- Cell Culture and DNA Transfection family. Consistent with that idea, the 461-amino acid human (h)CRLR contains seven transmembrane domains and three Human embryonic kidney (HEK) 293 cells were maintained cysteine (C) residues in the first (C212 and C225) and the in Dulbecco’s modified Eagle’s medium supplemented with second (C282) extracellular loops. In addition, the 160- 10% fetal bovine serum, 100 U/ml penicillin G, 100 µg/ml amino acid RAMPs possess a short cytoplasmic domain, a streptomycin and 0.25 µg/ml amphotericin B at 37¼C under single transmembrane domain and a large extracellular do- a humidified atmosphere of 95% air/5% CO2. For exper- main containing 4–7 cysteine residues (12). In the present imentation, cells were seeded onto 24-well plates and, upon study, we used site-directed mutagenesis to substitute, one at reaching 70% confluence, were transiently co-transfected a time, the cysteine residues in the extracellular loops of with hRAMP2 and myc-hCRLR or one of its mutant expres- hCRLR or the extracellular domain of hRAMP2 with the sion constructs, or with hCRLR and myc-hRAMP2 or one of aim of determining their respective roles in the structure and its mutants, using Lipofectamine transfection reagents (Invit- function of the AM receptor. rogen) according to the manufacturer’s instructions. Briefly, the cells were incubated for 3 h in 250 µl Optimem 1 medi- µ Methods um containing 200 ng/well plasmid DNA and 2 l/well Lipofectamine. As a control, some cells were transfected with an empty vector (pCAGGS/Neo). All experiments were Materials performed 48 h after transfection. [125I]hAM (specific activity 2 µCi/pmol) was produced in our laboratory (13). hAM was purchased from Peptide Fluorescence-Activated Cell-Sorting (FACS) Analysis Institute (Osaka, Japan). Mouse fluorescein isothionate (FITC)-conjugated monoclonal anti-myc antibody (anti-myc- Flow cytometry was performed to assess the levels of whole FITC antibody) was from Invitrogen (Carlsbad, USA). All cell and cell surface expression of myc-hCRLR, myc- other reagents were of analytical grade and obtained from hRAMP2 and their respective point mutants in HEK293 various commercial suppliers. cells. To evaluate cell surface expression, following transient transfection, cells were harvested, washed twice with phos- phate-buffered saline (PBS), resuspended in ice-cold FACS Expression Constructs buffer (11), and then incubated for 60 min at 4¼C in the dark hCRLR (14) and hRAMP2 (10) were modified to provide with anti-myc-FITC antibody (1:1,000 dilution). For evalua- a consensus Kozak sequence as previously described tion of whole cell expression, cells were permeabilized using (14). Initially, a myc epitope tag (EQKLISEEDL) was ligat- IntraPrepTM reagents (Beckman Coulter, Fullerton, USA) ac- ed, in-frame, to the 5′end of the hCRLR and hRAMP2 cording to the manufacturer’s instructions, after which they cDNAs, and the native signal sequences were removed and were incubated for 15 min at room temperature in the dark replaced with MKTILALSTYIFCLVFA (15). The myc- with anti-myc-FITC antibody (1:1,000 dilution). Following hCRLR and -hRAMP2 were then cloned into pCAGGS/Neo two successive washes with FACS buffer, both groups of expression vectors using the 5′XhoI and 3′NotI sites to gen- cells were subjected to flow cytometry in an EPICS XL flow erate pCAGGS-hCRLR and -hRAMP2 (11). cytometer (Beckman Coulter) and analyzed using EXPO 2 Single amino acid substitutions were carried out using a software (Beckman Coulter). Quick Change kit (Stratagene, La Jolla, USA) according to the manufacturer’s instructions. pIRES-myc-hCRLR and Radioligand Binding Assays -hRAMP2, which were constructed by subcloning the coding sequence of myc-hCRLR and -hRAMP2 into pIRES1/Neo To assess whole-cell radioligand binding, transfected HEK293 (CLONTECH, Palo Alto, USA), served as the template. For cells in 24-well plates were washed twice with warmed PBS each mutation, two complementary 30-mer oligonucleotides and incubated for 20 min at 37¼C with 0.1% bovine serum (sense and antisense) were designed to contain the desired albumin (BSA) in PBS to reduce endogenous AM binding, mutation in their middle. To allow rapid screening of mutat- after which the remaining adherent cells were washed with ed clones, the primers carried an additional silent mutation ice-cold PBS. The cells were then incubated for 6 h at 4¼C that introduced or removed a restriction site. The presence of with 20 pmol/l [125I]hAM in the presence (for nonspecific each mutation of interest and the absence of undesired ones binding) or absence (for total binding) of 1 µmol/l unlabeled was confirmed by DNA sequencing. myc-hCRLR and hAM in modified Krebs-Ringers-N-2-hydroxyethylpipe- -hRAMP2 mutants were then cloned into pCAGGS/Neo. razine-N′-2-ethane sulfonic acid (HEPES) medium (11). The products from polymerase chain reaction were se- Thereafter, the cells were washed twice with ice-cold PBS quenced using an Applied Biosystems 310 Genetic Analyzer and harvested with 0.5 mol/l NaOH. The associated cellular (Applied Biosystems, Foster City, USA). radioactivity was measured in a γ-counter. Specific binding was defined as the difference between total binding and non- Kuwasako et al: Extracellular Cysteines in AM Receptor Function S27

addition of lysis buffer (Amersham Biosciences, Bucking- hamshire, UK). The resultant lysates were centrifuged at 2,000 rpm for 10 min at 4¼C, and the cAMP content in samples of the supernatants was assayed using a commercial enzyme immunoassay kit according to the manufacturer’s (Amersham) instructions for a non-acetylation protocol.

Statistical Analysis Results are expressed as the means±SEM of at least three independent experiments. Differences between two groups were evaluated with Student’s t-tests; differences among multiple groups were evaluated with a one-way analysis of variance followed by Scheffe’s tests. Values of p＜0.05 were considered to indicate statistical signiﬁcance.

Results

Whole Cell and Cell Surface Expression of hCRLR Mutants In this study, we used HEK293 cells for gene transfection because they express only low levels of hRAMP2 and hCRLR mRNA (expression ratio＝25:1), and they lack functional AM receptors (16). We initially analyzed the whole cell expression of epitope-tagged hCRLR mutants in permeabilized cells using FACS (Fig. 1A). Surface and intracellular immunoreactivity was detected in only 1.7±0.18% of cells expressing the Fig. 1. FACS analysis of HEK293 cells expressing hRAMP2 empty vector (Mock), which was well within the 2% limit of and mutant hCRLR. A, Whole cell expression of the indicat- resolution characteristic of FACS analysis. When expressed ed myc-tagged proteins in permeabilized cells. Forty-eight alone or with hRAMP2, myc-hCRLR was detected in 23.7± hours after transfection, cells were permeabilized with 0.18% and 19.8±1.2% of cells, respectively. Likewise, co- TM IntraPrep reagents and then incubated for 15 min at room expression of hRAMP2 with an hCRLR C212A, C225A or temperature with monoclonal anti-myc-FITC antibody; C282A mutant led to their full expression in 19–24% of Mock incubation with the antibody served as the control. cells. Thus, the transfection efficacy for the mutant receptor Whole cell expression of each construct was estimated by flow cytometry. Bars represent the means±SEM of three was comparable to that for myc-hCRLR/hRAMP2. independent experiments; ＊ p＜0.001 vs. the control. B, Cell We next analyzed the cell surface expression of these surface expression of the indicated myc-tagged proteins in hCRLR mutants in non-permeabilized cells (Fig. 1B). The non-permeabilized cells. Following transient transfection, empty vector and myc-hCRLR appeared at the surface of cells were incubated for 1 h at 4¼C with monoclonal anti- 1.5±0.35% and 13.3±0.79% of cells, respectively. When myc-FITC antibody; Mock incubation with the antibody hRAMP2 was co-expressed with an hCRLR C212A, C225A served as the control. Cell surface expression of each con- or C282A mutant, cell surface immunoreactivity was detect- struct was estimated by flow cytometry. Bars represent the ed in 9–19% of cells, which again was similar to myc- ＊ means±SEM of three independent experiments; p＜0.001 hCRLR/RAMP2. vs. the control.

Radioligand Binding to hCRLR Mutant Receptors speciﬁc binding. Analysis of the binding of 20 pmol/l [125I]AM to receptors comprised of hRAMP2 complexed with the indicated mutant revealed no remarkable differences among cells expressing cAMP Measurements the empty vector or myc-hCRLR (Fig. 2). On the other hand, Cells were exposed to hAM in Hanks’ buffer containing in cells co-expressing hRAMP2 and myc-hCRLR, C212A, 20 mmol/l HEPES, 0.1% BSA and 0.5 mmol/l 3-isobutyl-1- C225A or C282A, the speciﬁc AM binding was 85–95-fold methylxanthine (Sigma Chemical, St. Louis, USA) for higher than in control cells (Mock). 15 min at 37¼C, after which the reactions were terminated by S28 Hypertens Res Vol. 26, Suppl (2003)

Fig. 2. [125I]AM binding to HEK293 cells expressing hRAMP2/hCRLR mutants. Total (open bars) and nonspecific (closed bars) binding are shown. Cells were transiently transfected with the indicated myc-tagged genes and then incubated for 6 h at 4¼C with 20 pmol/l [125I]hAM in the presence (for nonspecific binding) or absence (for total binding) of 1 µmol/l unlabeled hAM. Bars represent the means±SEM of three independent experiments; ＊ p＜0.008 vs. corresponding nonspecific binding.

Fig. 4. FACS analysis of HEK293 cells expressing hCRLR and mutant hRAMP2. A, Whole cell expression of the indicated myc-tagged proteins in permeabilized cells. Forty- eight hours after transfection, cells were permeabilized with IntraPrepTM reagents and then incubated for 15 min at room temperature with monoclonal anti-myc-FITC antibody; Mock incubation with the antibody served as the control. Whole cell expression of each construct was estimated by flow cytometry. Bars represent the means±SEM of three independent experiments; ＊ p＜0.001 vs. the control. B, Cell surface expression of the indicated myc-tagged proteins in non-permeabilized cells. Following transient transfection, cells were incubated for 1 h at 4¼C with monoclonal anti-myc-FITC antibody; Mock incubation with the antibody served as the Fig. 3. Agonist-evoked cAMP production in HEK293 cells control. Cell surface expression of each construct was esti- co-expressing hRAMP2 with C212A, C225A or C282A. Cells mated by flow cytometry. Bars represent the means±SEM of were transiently transfected with empty vector or myc- three independent experiments; ＊ p＜0.03 vs. the control. hCRLR (A) and with hRAMP2 plus myc-hCRLR, C212A, C225A or C282A (B), after which they were incubated for 15 min at 37¼C with the indicated concentrations of hAM and then lysed. The resultant lysates were analyzed for cAMP dicative of the absence of functional AM receptors in this content. Symbols represent the averages of two independent cell type. Virtually identical responses were obtained with experiments. cells expressing myc-hCRLR alone (Fig. 3A). This is in contrast to our earlier finding that AM-induced cAMP production was 2-fold higher in HEK293 cell transfectants stably expressing CRLR than in intact cells (11). Most likely this cAMP Production Mediated via hCRLR Mutant Receptors difference is largely explained by differences in transfection The EC50 for AM-induced cAMP production in intact efficacy between transient and stable transfectants. HEK293 cells was ＞2.0×10－7 mol/l (Fig. 3A), which is in- In cells co-expressing hRAMP2 and myc-hCRLR or Kuwasako et al: Extracellular Cysteines in AM Receptor Function S29

10000 * 8000 *

6000

4000

I-AM binding (cpm) * 2000 125

RLR Mock -RAMP2 C68A/CRLRC84A/CRLRC99A/CRLR myc C131A/CRLRC155A/CRLR Fig. 6. Agonist-evoked cAMP production in HEK293 cells -RAMP2/C co-expressing hCRLR with C68A, C84A, C99A, C131A or myc C155A. Cells were transiently transfected with empty vector, Fig. 5. [125I]AM binding to HEK293 cells expressing hCRLR/ myc-hRAMP2 or hCRLR plus C68A, C84A, C99A or C131A hRAMP2 mutants. Total (open bars) and nonspecific (closed (A) and with hCRLR plus myc-hRAMP2 or C155A (B), after bars) binding are shown. Cells were transiently transfected which they were incubated for 15 min at 37¼C with the indi- with the indicated myc-tagged genes and then incubated for cated concentrations of hAM and then lysed. The resultant 6 h at 4¼C with 20 pmol/l [125I]hAM in the presence (for non- lysates were analyzed for cAMP content. Symbols represent specific binding) or absence (for total binding) of 1 µmol/l the averages of two independent experiments. unlabeled hAM. Bars represent the means±SEM of three independent experiments; ＊ p＜0.04 vs. corresponding nonspecific binding. transmembrane domain (C155). When we analyzed the immunoreactivity of non-permeabilized and permeabilized cells co-expressing hCRLR and the hRAMP2 C155A mu- C225A, AM elicited concentration-dependent increases in tant, we found that the mutant receptor was fully expressed cAMP that reached levels nearly 80-fold greater than in cells at the cell surface and in the intracellular compartment at a expressing the empty vector, and there was no apparent dif- level comparable with myc-hRAMP2/hCRLR (Fig. 4A and －10 ference in potency: the EC50 for AM was 6.2×10 mol/l B). with both cell types (Fig. 3B). Cells expressing hRAMP2 and C212A or C282A responded comparatively weakly to Radioligand Binding to hRAMP2 Mutant Receptors AM in the concentration range tested: the EC50 values were ＞6.8×10－8 mol/l and ＞2.2×10－8 mol/l, respectively (Fig. Analysis of [125I]AM binding revealed no remarkable differ- 3B). ences among cells expressing the empty vector or myc- hRAMP2 alone (Fig. 5). The specific AM binding in cells expressing hCRLR and myc-hRAMP2 or C155A was 50– Whole Cell and Cell Surface Expression of hRAMP2 60-fold greater than that in Mock, while binding in cells ex- Mutants pressing the other mutants (C68A, C84A, C99A and C131A) Figure 4A shows the whole cell expression of epitope-tagged was ＜5-fold greater than in Mock. hRAMP2 mutants in permeabilized cells detected using FACS. The frequency of immunoreactivity in cells express- cAMP Production Mediated via hRAMP2 Mutant Receptors ing the empty vector, myc-hRAMP2 or myc-hRAMP2/ hCRLR was 1.5±0.32%, 27.7±1.4% and 40.6±1.9%, re- We found no remarkable differences in AM-evoked cAMP spectively. When co-expressed with hCRLR, the hRAMP2 production among cells expressing empty vector (Mock), C68A, C84A, C99A or C131A mutant was detected in myc-hRAMP2 alone, or one of the four hRAMP2 mutants 24–29% of cells. By contrast, cell surface immunoreactivity (C68A, C84A, C99A and C131A) together with hCRLR was detected in ＜10% of cells (Fig. 4B), suggesting that (Fig. 6A). By contrast, cells expressing C155A and hCRLR －10 these four cysteine residues in the extracellular domain of readily responded to AM, and the EC50 of 5.4×10 mol/l hRAMP2 are all involved in the cell surface expression of was comparable to that obtained with cells expressing myc- －9 the hRAMP2/hCRLR heterodimer. The empty vector, myc- hRAMP2/hCRLR (EC50＝1.6×10 mol/l) (Fig. 6B). Ap- hRAMP2 and myc-hRAMP2/hCRLR appeared at the surface parently, C155, situated in the transmembrane domain of of 1.9±0.08%, 13.1±0.64% and 30.2±1.4% of cells, re- hRAMP2, is not involved in the cell surface expression or spectively. function of the AM receptor. hRAMP2 also contains a cysteine residue in its one S30 Hypertens Res Vol. 26, Suppl (2003)

proper delivery of the receptor proteins to the cell surface Discussion membrane, which is consistent with previous findings that mutation of the participating cysteine residues drastically re- Previous pharmacological and biochemical analyses of mu- duced the maturation and transport of the proteins from the tant GPCRs in which the cysteine residues in the first and endoplasmic reticulum due to misfolding and aggregation second extracellular loops were replaced by serine or alanine (19). We recently reported that the seven-amino acid seg- indicated that the substituting amino acid had little effect on ment between the second and third cysteine residues in hu- the expression and function of platelet-activating factor re- man and rat RAMP2 most likely contributes to the structure ceptor (2) and vasopressin V2 receptor (5). On the other of the ligand-binding pocket, but not to the ligand-binding hand, with rhodopsin, another seven-transmembrane-domain sites (20, 21). This means that C84 and C99 are probably the GPCR, replacement of the two cysteine residues in the first residues are involved in forming a disulfide bond. Additional and the second extracellular loops with serine disrupted the experiments will be needed to clarify whether disulfide tertiary structure and reduced the stability of the receptor bonds are also formed within the RAMP2 molecule or be- protein (17). This was probably due to the hydrophilic nature tween RAMP2 and CRLR. of the serine side chain; indeed, when the two cysteine In summary, we have shown that C212 and C282 likely residues were replaced with alanines, the tertiary structure of form a disulfide bond critical to the structure and function of the resultant rhodopsin mutant appeared to be almost normal the hCRLR. We have also shown that substituting any one of (18). Therefore, in order to minimize the effects of the sub- four cysteine residues (C68, C84, C99 and C131) in stitution on proper protein folding, we decided to replace the hRAMP2 yields an apparently normal receptor protein that is extracellular cysteine residues of hCRLR and hRAMP2 with not transported to the cell surface. Cysteine residues in the alanine rather than with serine. extracellular loops of hCRLR and in the extracellular do- Most seven-transmembrane-domain GPCRs contain a pair main of hRAMP2 thus appear to play distinct roles in the of cysteine residues in the first and second extracellular cell surface expression and function of AM receptors. loops that are linked via a disulfide bond (1–9). With respect to platelet-activating factor receptor (2), µ-opioid receptor (3), M3 muscarinic acetylcholine receptor (4), and vaso- References pressin V2 receptor (5), the extracellular disulfide bonds ap- 1. Strader CD, Fong TM, Tota MR, Underwood D, Dixon pear necessary for proper transport of the receptor to the RAF: Structure and function of G protein-coupled recep- plasma membrane―i.e., the mutant receptors failed to ap- tors. Annu Rev Biochem 1994; 63: 101–132. pear at the cell surface. In the present study, by contrast, co- 2. Gouill CL, Parent JL, Pleszczynski MR, Stankova J: Role expression of hRAMP2 with an hCRLR mutant in which the of the Cys90, Cys95 and Cys173 residues in the structure and cysteine residues in the first (C212) and the second (C282) function of the human platelet-activating factor receptor. extracellular loops were substituted with alanine resulted in FEBS Lett 1997; 402: 203–208. full expression of the receptor heterodimer at the cell surface 3. Gaibelet G, Capeyrou R, Dietrich G, Emorine LJ: Identifi- (Fig. 1B) and high-affinity AM binding comparable to that cation of the µ-opioid receptor of cysteine residues respon- seen with unmutated CRLR/RAMP2 (Fig. 2); nevertheless, sible for inactivation of ligand binding by thiol alkylating – AM-induced cAMP production was significantly diminished and reducing agents. FEBS Lett 1997; 408: 135 140. 4. Zeng FY, Soldner A, Schoneberg T, Wess J: Conserved ex- (Fig. 3B). This finding of diminished receptor function is tracellular cysteine pair in the M3 muscarinic acetylcholine similar to that obtained with thyrotropin receptor (5), β2- receptor is essential for proper receptor cell surface local- adrenergic receptor (7), corticotropin-releasing factor recep- ization but not for G protrine coupling. J Neurochem 1999; tor (8) and thromboxane A2 receptor (9). The fact that muta- 72: 2404–2414. tion of C225 in the first extracellular loop of hCRLR had no 5. Schulein R, Zuhlke K, Oksche A, Hermosilla R, Furkert J, effect on [125I]AM binding or receptor function (Figs. 2 and Rosenthal W: The role of conserved extracellular cysteine 3B) suggests that C212 and C282 form a disulfide bond that residues in vasopressin V2 receptor function and properties is crucial for maintaining proper receptor functionality. Still, of two naturally occurring mutant receptors with additional we cannot rule out the possibility that C212 and C282 link to extracellular cysteine residues. FEBS Lett 2000; 466: – a cysteine in the large extracellular domain of hCRLR, 101 106. hRAMP2 or both. 6. Kosugi S, Ban T, Akamizu T, Kohn LD: Role of cysteine residues in the extracellular domain and exoplasmic loop of When co-transfected with hCRLR, none of the four the transmembrane domain of the TSH receptor: effect of hRAMP2 mutants tested (C68A, C84A, C99A and C131A) mutation to serine on TSH receptor activity and response to appeared at the cell surface (Fig. 4B). The abundantly pro- thyroid stimulating autoantibodies. Biochem Biophys Res duced mutant receptors remained in the intracellular com- Commun 1992; 189: 1754–1762. partment (Fig. 4A) and showed no specific ligand binding or 7. Noda K, Saad Y, Graham RM, Karnik SS: The high affinity signaling capability (Figs. 5 and 6A). Apparently, all four of state of the β2-adrenergic receptor requires unique interac- the hRAMP2 cysteine residues tested are essential for the tion between conserved and non-conserved extracellular Kuwasako et al: Extracellular Cysteines in AM Receptor Function S31

loop cysteines. J Biol Chem 1994; 269: 6743–6752. protein following introduction of a cleavable signal peptide. 8. Chiang N, Kan WM, Tai HH: Site-directed mutagenesis of J Biol Chem 1992; 267: 21995–21998. cysteinyl and serine residues of human thromboxane A2 re- 16. Nagoshi Y, Kuwasako K, Ito K, et al: The calcitonin recep- ceptor in insect cells. Arch Biochem Biophys 1996; 334: tor-like receptor/receptor activity-modifying protein 1 het- 9–17. erodimer can function as a calcitonin gene-related peptide- 9. Qi LJ, Leung AT, Xiong Y, Marx KA, Samra ABA: Extra- (8–37)-sensitive adrenomedullin receptor. Eur J Pharmacol cellular cysteines of the corticotropin-releasing factor re- 2002; 450: 237–243. ceptor are critical for ligand interaction. Biochemistry 1997; 17. Karnik SS, Sakmar TP, Chen HB, Khorana HG: Cysteine 36: 12442–12448. residues 110 and 187 are essential for the formation of cor- 10. McLatchie LM, Fraser NJ, Main MJ, et al: RAMPs regu- rect structure in bovine rhodopsin. Proc Natl Acad Sci late the transport and ligand specificity of the calcitonin-re- USA1988; 85: 8459–8463. ceptor-like receptor. Nature 1998; 393: 333–339. 18. Davidson FF, Loewen PC, Khorana HG: Structure and 11. Kuwasako K, Shimekake Y, Masuda M, et al: Visualiza- function in rhodopsin: replacement by alanine of cysteine tion of the calcitonin receptor-like receptor and its receptor residues 110 and 187, components of a conserved disulfide activity-modifying proteins during internalization and recy- bond in rhodopsin, affects the light-activated metarhodopsin cling. J Biol Chem 2000; 275: 29602–29609. II state. Proc Natl Acad Sci U S A 1994; 91: 4029–4033. 12. Sexton PM, Albiston A, Morfis M, Tilakaratne N: Receptor 19. Doms RW, Lamb RA, Rose JK, Helenius A: Folding and activity modifying proteins. Cell Signal 2001; 13: 73–83. assembly of viral membrane proteins. Virology 1993; 193: 13. Matsui E, Kitamura K, Yoshida M, et al: Biosynthesis and 545–562. secretion of adrenomedullin and proadrenomedullin N-ter- 20. Kuwasako K, Kitamura K, Ito K, et al: The seven amino minal 20 peptide in a rat model of endotoxin shock. Hyper- acids of human RAMP2 (86–92) and RAMP3 (59–65) are tens Res 2001; 24: 543–549. critical for agonist binding to human adrenomedullin recep- 14. Aiyar N, Rand K, Elshourbagy NA, et al: A cDNA encod- tors. J Biol Chem 2001; 276: 49459–49465. ing the calcitonin gene-related peptide type 1 receptor. J 21. Kuwasako K, Kitamura K, Onitsuka H, et al: Rat RAMP Biol Chem 1996; 271: 11325–11329. domains involved in adrenomedullin binding specificity. 15. Guan XM, Kobilka TS, Kobilka BK: Enhancement of FEBS Lett 2002; 519: 113–116. membrane insertion and function in a type IIIb membrane