002G-8DnXID9/0 1023,"-08$3 00/0
`COJ1yright © The Amc1ican SodcLy for Pharmacology a nd Ex pct•imcntal Thcrapm1tics
`AH rights of reproduction ln any form reserved,
`MOLECULAR PHARMACOLOGY, 56:2,35- 242 (1999),
`
`Multiple Amylin Receptors Arise from Receptor Activity(cid:173)
`Modifying Protein Interaction with the Calcitonin Receptor
`Gene Product
`
`GEORGE CHRISTOPOULOS, KATIE J . PERRY, MARIA MORFIS, NANDA TILAKARATNE, YONGYI GAO, NEIL J. FRASER,
`MARTIN J. MAIN, STEVEN M. FOORD, and PATRICK M. SEXTON
`Molecular Pharmacology Laboratory, Department of Pharmacology, The University of Melbourne, Victoria, Australia (G.C., K.J.P. , M.M. , N. T.,
`Y.G., P.M.S.); and Receptor Systems Unit, Glaxo Wei/come Medicines Research Centre, Stevenage, Hertfordshire United Kingdom (N.J.F.,
`M.J.M., S.M.F.)
`Received April 14, 1999; accepted May 18, 1999
`
`This paper is available online at http://www.molpharm.org
`
`ABSTRACT
`Receptor activity-modifying proteins (RAMPs) are single-trans(cid:173)
`membrane proteins that transport the calcitonin receptor-like
`receptor (CRLR) to the cell surface. RAMP 1-transported CRLR
`is a calcitonin gene-related peptide (CGRP) receptor. RAMP 2-
`or RAMP 3-transported CRLR is an adrenomedullin receptor.
`The role of RAMPs beyond their interaction with CRLR, a class
`II G protein-coupled receptor, is unclear. In this study, we have
`examined the role of RAMPs in generating amylin receptor
`phenotypes from the calcitonin (CT) receptor gene product.
`Cotransfection of RAMP 1 or RAMP 3 with the human CT
`receptor lacking the 16-amino acid insert in intracellular domain
`1 (hCTR 11 _ ) into COS-7 cells induced specific 1251-labeled rat
`amylin binding. RAMP 2 or vector cotransfection did not cause
`significant increases in specific amylin binding. Competition(cid:173)
`binding characterization of the RAMP-induced amylin recep(cid:173)
`tors revealed two distinct phenotypes. The RAMP 1-derived
`amyl in receptor demonstrated the highest affinity for salmon CT
`(IC 50 , 3.01 ± 1.44 x 1 o- 10 M), a high to moderate affinity for rat
`amyl in (IC5 0 , 7.86 ± 4.49 X 1 o- 9 M) and human CGRPcx (IC5 0 ,
`
`2.09 ± 1.63 x 1 o- s M), and a low affinity for human CT (IC 50 ,
`4.4 7 ± 0. 78 X 10 -- 7 M). In contrast, whereas affinities for amyl in
`and the CTs were similar for the RAMP 3-derived receptor, the
`efficacy of human CGRPa was markedly reduced (IC50, 1 .12 ±
`0.45 X 1o- 7 M; P < .05 versus RAMP 1). Functional cyclic AMP
`responses in COS-7 cells cotransfected with individual RAMPs
`and hCTR 11 _ were reflective of the phenotypes seen in com(cid:173)
`petition for amyl in binding. Confocal microscopic localization of
`c-myc-tagged RAMP 1 indicated that, when transfected alone,
`RAMP 1 almost exclusively was located intracellularly. Cotrans(cid:173)
`fection with calcitonin receptor (CTR) 11 _
`induced cell surface
`expression of RAMP 1. The results of experiments cross-linking
`1251-labeled amylin to RAMP 1/hCTR-transfected cells with bis
`succidimidyl suberate were suggestive of a cell-surface asso(cid:173)
`ciation of RAMP 1 and the receptors. Our data suggest that in
`the CT family of receptors, and potentially in other class II G
`protein-coupled receptors, the cellular phenotype is likely to be
`dynamic in regard to the level and combination of both the
`receptor and the RAMP proteins.
`
`A.mylin is a 37-amino acid pancreatic hormone that shares
`amino acid homology with the calcitonin gene-related peptide
`(CGRP), calcitonin (CT), and the adrenomedullin family of
`peptides. It has the highest identity ( - 45%) with the CGRPs,
`an - 22% identity with the 38 C-terminal amino acids of
`adrenomedullin, and an 18 and 33% identity (with a gapped
`alignment) with rat/human and avian/teleost CTs, respec(cid:173)
`tively. The physiology of these peptides has been reviewed in
`detail (Muff et al., 1995; Wimalawansa, 1997). Circulating
`
`This work was fund ed in part by the National Health and Medical &;search
`Council of Australia and by GlaxoWellcome, Australia. P.M.S. is n Research
`Fellow uf lhe Nuliuual HeaHh umllvledical Research Council of Australia .
`
`levels of amy lin are raised in response to meal ingestion, and
`the peptide acts to potently inhibit ga stric emptying, post(cid:173)
`prandial glucagon secretion, and food intake. A.mylin also
`opposes the metabolic actions of insulin in skelet al muscle
`(Sexton and Perry, 1996; Young, 1997). Transgenic mice lack(cid:173)
`ing the amylin gene show abnormal weight gain, an obser(cid:173)
`vation that also suggests an important metabolic role for
`amylin (Devine and Young, 1998; Gebre-Medhin et al. , 1998).
`An independent gene encoding the amylin receptor has not
`been identified.
`McLatchie et al. (1998) recently identified and cloned a
`family uf at:t:essury pruLeim; termed receptor activity-modify-
`
`ABBREVIATIONS: CGRP, calcitonin gene-related peptide; CT, calcitonin; RAMP. receptor activity-modifying protein; CRLR, calcitonin receptor(cid:173)
`like receptor; a-TSH cel ls, a-thyroid-stimulating hormone thyrotroph cells; 883
`, bissuccidimidyl suberate; cAMP, cyclic AMP; CHO, Chinese
`hamster ovary; CTR, calcitonin receptor; DMEM, Dulbecco's modified Eagle's medium; FBS, fetal bovine serum; hCTR11 _ , human CT receptor
`lacking the 16 amino acid insert in intracellular domain 1; HEK, human embryonic kidney; rCTR 11 ._ , rat insert negative CT receptor isoform
`equivalent to rat C1 a CT receptor.
`------~--------------------------------
`
`235
`
`1
`
`EX2140
`Eli Lilly & Co. v. Teva Pharms. Int'l GMBH
`IPR2018-01427
`
`

`

`236
`
`Christopoulos et al.
`
`ing proteins CRAMPs), which was comprised of three mem(cid:173)
`bers designated RAMP 1, RAMP 2, and RAMP 3. These
`single-transmembrane domain proteins induced trafficking
`of the calcitonin receptor-like receptor (CRLR) to the cell
`surface, where it exhibited either a CGRP receptor pheno(cid:173)
`type (RAMP 1) or adrenomedullin receptor phenotypes
`(RAMP 2 or RAMP 3). RAMPs therefore provided a novel
`mechanism for engendering novel receptor phenotypes. Amy(cid:173)
`lin shows more sequence homology to CGRP than ad(cid:173)
`renomedullin, but it does not activate or bind to combinations
`of CRLR and RAMPs (McLatchie et al. , 1998). Amylin has
`even less sequence identity with the CTs, but there is evi(cid:173)
`dence that links amylin receptors with those for CT. The
`receptors tend to be colocalized (Sexton and Perry, 1996), and
`both receptors are recognized by antibodies raised against
`the hypervariable C terminus of the CT receptor (Perry et al.,
`1997). Furthermore, transfection of human CT receptors into
`human embryonic kidney (HEK)-293 cells, but not into Ti ni
`insect cells, induces low levels of amy lin receptors in addition
`to high levels of CT receptors (Chen et al., 1997). Moreover,
`transfection of the porcine CT receptor into different cellular
`backgrounds gives rise to different receptor phenotypes with
`transfection into Chinese hamster ovary (CHO)-K1 cells,
`yielding a receptor similar to the rat C1a CT receptor (Christ(cid:173)
`manson et al., 1994), whereas transfection into COS-7 or
`HEK-293 cells yields a receptor with moderate to high affin(cid:173)
`ity for amylin and poor responsiveness to human CT (Christ(cid:173)
`manson et al., 1994; Sexton et al., 1994a).
`These observations prompted an investigation of whether
`RAMP coexpression may also underlie the expression of amy(cid:173)
`lin receptors from the CT receptor gene product. Our data
`indicate that at least two independent amylin receptor phe(cid:173)
`notypes may be engendered by specific RAMP interaction.
`
`Experimental Procedures
`Materials. Salmon CT, human CT, human adrenomedullin, hu(cid:173)
`man CGRP<>, and rat amy lin were obtained from Bachem (Torrance,
`CAl. Dulbecco's modified Eagle's medium CDMEM), fetal bovine se(cid:173)
`rum (FBS), HEPES, G418, and Lipofectamine were obtained from
`GIBCO-BRL Life Technologies (Grand Island, NY). BSA was ob(cid:173)
`tained from Commonwealth Serum Laboratories (Parkville, Austra(cid:173)
`lia), anti-c-myc antibody was obtained from Invitrogen (Carlsbad,
`CA), and Alexa 488-conjugated goat anti-mouse sera and TOT0-3
`were obtained from Molecular Probes (Eugene, OR). Isobutylmeth(cid:173)
`ylxanthine was obtained from Sigma Chemical Co. (St. Louis, MOl,
`tissue culture plates and flasks were obtained from Nunc (Roskilde,
`Denmark), and anti-cyclic AMP (cAMP) antibody was a g·ift from Dr.
`Philip Marley (Department of Pharmacology, University of Mel(cid:173)
`bourne, Melbourne, Australia ). Endo F was obtained from Boehr(cid:173)
`inger Mannheim (Mannheim, Germany), and bissuccidimidyl suber(cid:173)
`ate (BS3 ) was obtained from Pierce Chemical Co. (Rockford, IL).
`Na 125I and 1251-labeled rat amylin (specific activity, 2000 Cilmmoi)
`were obtained from Amersham (Buckinghamshire, UK). 1251-labeled
`salmon CT (specific activity, ~ 700 Ci/mmol) was prepared as de(cid:173)
`scribed previously (Nicholson et al. , 1986). All other chemicals were
`of reagent grade or better.
`Cell Culture and eDNA Transfection. Green monkey kidney
`COS-7 cells were maintained in 175-cm2 flasks at 37•c in a humid(cid:173)
`ified atmosphere with 95% 0 2:5% C02, in complete DMEM supple(cid:173)
`mented with 10% FBS, 80 mg/1 gentamycin, 1 mg/1 minocycline, and
`15 mM HEPES. Human embryonic kidney (HEK)-293 cells stably
`expressing the rat C1a CT receptor [clones F12 ( ~60,000 receptors/
`cell) and Dll ( ~600,000 receptors/cell); Houssami et al., 1994] were
`
`maintained in antibiotic- and HEPES-supplemented DMEM con(cid:173)
`taining 5% FBS and 200 1'-g/ml G418. CHO-Kl cells (a gift from Dr.
`Steve Rees, GlaxoWellcome Medicines Research Center, Stevenage,
`UK) were maintained in DMEM/Ham's F12 media (50:50) supple(cid:173)
`mented with 10% FBS, 2 mM glutamine, and 0.5 mg/ml hygro(cid:173)
`mycin B.
`Coexpression of CT Receptor and RAMP eDNA. Cells in 24-
`or 6-well plates were grown to 70 to 80% confluency and transfected
`with 0.1/Lg (unless otherwise specified) of plasmid DNA encoding the
`most commonly expressed CT receptor isoforms from human [hC(cid:173)
`TR11 _; a gift from Dr. Emma Moore, Zymogenetics Inc., Seattle, WA
`(Kuestner et al., 1994)] or rat (C1a; Albrandt et al., 1993; Sexton et
`al., 1993) and RAMP eDNA (McLatchie et al., 1998) per 2 cm2 by
`using Lipofectamine, according to the manufacturer's instruction s.
`The CT receptor isoforms from rat and human are equivalent in their
`exon splicing and have been denoted hCTRn- and rCTRn _ for this
`Rtudy. In some experiments, increasing concentrations of RAMP
`DNA were transfected. Radioligand binding and cAMP assays were
`performed 48 h after transfection.
`Radioligand Binding. Receptor-expressing cells in 24-well
`plates, at 90 to 100% confluency, were incubated in binding buffer
`[DMEM containing 0.1% (w/v) BSA] with ~so pM 1251-labeled
`salmon CT or ~70 pM 1 2 51-labeled rat amylin (Sexton et al., 1993), in
`the absence (total binding) or presence of increasing concentrations
`of unlabeled ligands. Nonspecific binding was defined as binding in
`the presence of 1/LM homologous unlabeled peptide. After incubation
`for 60 min at 37•c, cells were washed with ice-cold PBS (140 mM
`NaCl, 2 mM KCI, 1 mM KH2P04 , and 8 mM Na 2HPO", pH 7.3) and
`solubilized with 0.5 M NaOH. Competition binding curves were
`analyzed with the Equilibrium Binding Data Analysis/Ligand soft(cid:173)
`ware package (Biosoft, Cambridge, UK).
`cAMP Assay. Transfected cells in 24-well plates, at 90 to 100%
`confluency, were preincubated in cyclase buffer [DMEM containing
`0.1% (w/v) BSA and 1 mM isobutylmethylxanthine] for 20 min at
`37•c . Cells subsequently were incubated for 25 min in the absence
`(basal) or presence of increasing concentrations of ligand. After in(cid:173)
`cubation, cells were washed with ice-cold PBS, and cAMP was ex(cid:173)
`tracted with 0.5 ml of absolute ethanol. cAMP levels were assayed by
`radioimmunoassay as described previously (Sexton et al., 1994a).
`Covalent Cross-Linking Analysis. Transfected cells in six-well
`plates were incubated for 60 min in binding buffer with an ~4 nM
`concentration of the specified radioligand in the absence (total bind(cid:173)
`ing), or presence of 1/LM homologous unlabeled peptide (nonspecific
`binding). After incubation, cells were washed with PBS and cross(cid:173)
`linked on ice for 35 min with 1 mM BS3
`. Cells were collected and
`solubilized in sample buffer [50 mM Tris HCl (pH 6.8) containing 2%
`(w/v) SDS, 0.1% bromophenol blue, 10% (v/v) glycerol, and 100 mM
`dithiothreitol] and centrifuged at 12,000g for 30 min at 4•c , and the
`supernatants were analyzed by 10% (w/v) SDS-polyacrylamide gel
`electrophoresis (Quiza et al., 1997). Gels were stained with Coomas(cid:173)
`sie blue R-250, destained, dried, and exposed to phosphor screens
`(Molecular Dynamics, Sunnyvale, CAJ. Deglycosylation was per(cid:173)
`formed as described previously (Quiza et al., 1997). TheM,. oflabeled
`bands was determined from a standard curve generated from the
`electrophoretic mobility of molecular weight markers that were co(cid:173)
`electrophoresed with the samples.
`Confocal Microscopic Localization of c-myc-Tagged RAMP
`1. RAMP 1 epitope tagged with the c-myc epitope at theN terminus
`(McLatchie et al., 1998) was transfected transiently into COS-7 cells
`seeded onto 22-mm glass coverslips in six-well plates, either alone or
`together with then- isoform of the rat or human CT receptor. Then,
`48 h after transfection, cells were fixed with 3.2% paraformaldehyde
`for 30 min at 22•c, the reaction was stopped with 150 mM glycine in
`PBS, and the cells were washed three times for 5 min in either PBS
`or PBS containing 0.3% TritonX-100. All subsequent treatments and
`washes were performed in either PBS for cell surface labeling or in
`PBS-Triton to allow intracellular identification of epitope-tagged
`protein. Cells were preblocked with 10% lamb serum for 30 min at
`
`2
`
`

`

`22•c, washed once with PBS or PBS-Triton, and then incubated with
`anti-c-myc antisera at a dilution of 1:500 for 1 h at 22"C. After
`incubation with primary antisera, cells were washed once and then
`incubated in the dark with Alexa 488-conjugated goat anti-mouse
`sera (1:100) for 1 hat 22•c. Cells were further incubated in the dark
`with PBS-Triton containing TOT0-3 (1:1000) and RNase (1:250) for
`1 h , washed twice with PBS-Triton for 5 min, dipped in distilled
`water, and mounted onto glass slides with fluorescent mounting
`media. Cells were visualized with a Bio-Rad confocal microscope
`(Bio-Rad Laboratories, Hercules, CA).
`
`Results
`Cotransfection of hCTR11 _ with increasing concentrations
`of RAMP 1 or RAMP 3 into COS-7 cells induced specific and
`high-affinity 1251-labeled amylin binding (Fig. 1, a and e).
`RAMP 2 had no significant effect (Fig. 1, a and e). Similar
`results were seen with the rCTRu - receptor isoform (Fig.
`1b). In contrast, in HEK-293 cells stably expressing the
`rCTRn - • only RAMP 1 induced alteration in the level of
`specific 1 2 51-labeled amylin binding (Fig. 1c). No significant
`change was seen in the level of 1 251-labeled salmon CT bind(cid:173)
`ing for any of the CT receptors studied (Fig. 1, d and f) .
`
`a
`
`2000
`
`I
`
`1600
`
`1!
`::::1
`0 1200
`.CI
`c
`t 800
`.1
`~
`u
`
`l
`
`Ill
`
`400
`
`0
`
`100
`150
`0
`50
`RAMP DNA transfected (ng)
`
`200
`
`d
`
`12000
`
`I
`
`.2-
`'tJ
`c:
`::::1
`.8
`t3
`~
`
`eooo
`
`6000
`
`4000
`
`2000
`
`0
`
`0
`
`50
`
`100 150 200 250
`
`RAMP DNA tranafected (ng)
`
`b
`e 1600
`a. 1400
`.2-
`~
`1200
`::::1
`.8 1000
`c
`~ 800
`II
`600
`..!.
`~
`u
`~
`
`400
`
`l 200
`
`Ill
`
`0
`
`e
`
`I 1800
`
`1800
`'tJ 1400
`c
`j 1200
`
`~
`u
`!E
`
`800
`600
`400
`
`0
`
`*
`10000 ~ .2-
`f 1000
`i
`! 200
`
`RAMPs Generate Amylin Receptors from Calcitonin Receptors
`
`237
`
`Transfection of increasing levels of RAMP into CHO-K1 cells,
`which endogenously express CT receptors, demonstrated in(cid:173)
`duction of 1251-labeled amylin binding with RAMPs 1 and 3
`but not with RAMP 2 or a vector control (Fig. 2). The expres(cid:173)
`sion of RAMP alone into COS-7 cells did not enable binding of
`either 1251-labeled amylin or 1 251-labeled salmon CT (not
`shown).
`The amylin receptors generated by the coexpression of the
`hCTRu -
`isoform with RAMP 1 and with RAMP 3 were
`analyzed in competition binding studies. Kd values for amy(cid:173)
`lin and salmon CT were 4.48 X 10- 9 M and 9.43 X 10- 10 M,
`respectively, for CTRJRAMP 1 and 5.38 X 10-9 M (amylin)
`and 5.8 x 10-9 M (salmon CT) for CTR/RAMP 3. The hC(cid:173)
`TRn _/RAMP 1 combination generated an amylin receptor
`equivalent to that identified previously in a-thyroid-stimu(cid:173)
`lating hormone thyrotroph cells (a-TSH cells; Perry et al.,
`1997). It had the highest affinity for salmon CT, a high to
`moderate affinity for rat amylin and human CGRPa, and a
`low affinity for human CT (Fig. 3a; Table 1). Human ad(cid:173)
`renomedullin had little interaction with this receptor, being
`at least 10-fold less potent than human CT (not shown). In
`c
`
`2000
`
`::::1
`
`I 1800
`
`'tJ c 1600
`.8
`c
`E .. ..I.
`j,
`~ 1200
`
`0
`
`50
`
`100
`
`150
`
`200
`
`RAMP DNA transfected (ng)
`
`0
`
`50 100 150 200 250
`
`RAMP DNA transfected (ng)
`
`1000
`
`f
`
`*
`
`II
`
`"',p.o~~~'~<fl-#~~'!>
`
`• !ll
`
`()
`
`I
`§
`.8
`t>
`15
`E
`1 4000
`=
`i
`
`10000
`
`8000
`
`6000
`
`2000
`
`0
`
`~'?'?;
`,p.o~ ~'?' ~<fl-
`"'
`#
`#
`#
`
`Fig. 1. Effect of cotransfection of RAI\IIP 1, RAlvlP 2, RAMP 3, or vector DNA on the expression of 1251-labeled amylin (a , b, c, e) or 12"1-labeled salmon
`,1' (d, n binding in
`S-7 cells tra nsien tly tnmsfcctcd wi th 100 ng of
`'1.' receptor (a, b, e, f) or HEK-293 cells stably expressi ng Lhe rC1'R11 _ (c: dl.
`• b. 12r.l-lt•beled amy lin binding in cell · transfected with 100 ng of rC'l'H 11
`a , "'"J-la beled amy lin binding in ce.lls transfected with 100 ng of hC'f R11
`•
`c, l!tiiJ-IabeJ d nmyli.n bin(ling in cells stably expressing the rnt C'fR 11 _. d, 1251-lnbohid salmon CT binding in cells sLnbly expressing U1c rCTR11 _.
`Rnpi'CHenWtive experiments (11 ~ 3) with Lriplic(ltc repeats. and f, pooled d11ta for specific 1251-labeled amylin binding te) or '""T-lal)clcd sul mon C1'
`binding (f) after cotransfection with 100 ng of RAMP and 100 ng ofhCTR11 ••. p, P < .05, ANOVA, multiple comparisons versus control (Tukey's Test;
`n = 7). P, RAMP 1. E, RAI\IlP 2. 9, RAMP 3. E, vector.
`
`3
`
`

`

`238
`
`Christopoulos et al.
`
`contrast, the binding of 1251-labeled amylin to the hCTR11 _/
`RAMP 3 combination was competed for by salmon CT, amy(cid:173)
`lin, and human CT in a manner similar to that seen with
`hCTRu_IRAMP 1, but human CGRPa was ~30-fold less
`effective (Fig. 3b; Table 1). As for the RAMP 1-induced phe(cid:173)
`notype, human adrenomedullin had the lowest affinity for
`this receptor and essentially did not compete for binding
`except at micromolar concentrations (not shown). Little
`change was seen in the level and specificity of 1251-labeled
`salmon CT binding to COS-7 cells after cotransfection with
`any of the RAMPs (Table 2). Similar results were seen with
`the rCTRn - isoform (not shown).
`The functional cAMP responses in COS-7 cells cotrans(cid:173)
`fected with individual RAMPs and hCTRn- were consistent
`with the pharmacology of the amy lin binding they induced.
`Cells cotransfected with receptor and vector control showed
`responses typical of a CT receptor (Kuestner et al., 1994;
`Albrandt et al., 1995; Gorn et al., 1995), with salmon and
`human CT displaying similar efficacy and amylin and CGRP
`only weakly stimulating cAMP accumulation. RAMP 1 and
`RAMP 3 increased amylin potency (Fig. 4a), whereas only
`RAMP 1 increased CGRP potency (Fig. 4b). RAMP cotrans(cid:173)
`fection caused a decrease in the efficacy of human CT (Fig.
`4c), whereas the efficacy of salmon CT essentially was unal(cid:173)
`tered by RAMP treatment (Fig. 4d). Consistent with its lim(cid:173)
`ited effect on specific amylin binding, RAMP 2 had little
`effect on peptide specificity and potency (Fig. 4).
`BS3 cross-linking of 1251-labeled amylin to RAMP VhC(cid:173)
`TRn __ -transfected cells revealed a broad receptor-binding
`protein with a Mr of ~80,000, whereas cells transfected with
`receptor plus vector control exhibited essentially no specific
`amylin binding (Fig. 5). 1251-labeled salmon CT labeled a
`band with a Mr of ~80,000 in both RAMP 1- and vector
`
`2000 .
`
`e
`-B- 1500
`
`1000
`
`500
`
`control-transfected cells. Endo F deglycosylation reduced the
`size of the 1251-labeled salmon CT-binding· protein to aM" of
`~54,000, consistent with the predicted size of the core recep-
`
`RAMP1
`
`·12
`
`·11
`
`-10
`
`·9
`
`-8
`
`-7
`
`-6
`
`RAMP3
`
`... ~f.~J~
`
`a
`
`5000
`
`e
`! 4000
`l
`j
`.E 3000
`'So ;
`1 2000
`
`1000
`
`b
`
`5000
`
`I 4000
`l
`j 3000
`Ji l
`1 2000
`
`~000
`
`" ' AMY
`l "
`aCT ll.
`1\
`\
`
`\
`
`··.\
`'.\ hCT
`CGRP·.\
`'.\
`• •••
`
`'
`
`-6
`
`-12
`
`·11
`-10
`·9
`·8
`-7
`Log10 (peptide] (M)
`Fig. 3. RAMP 1 and RAMP 3 generate two distinct amylin receptors. a,
`competition of 1261-labeled rat amylin binding to COS-7 cells cotrans(cid:173)
`fected with 100 ng ofhCTR11 _ and 100 ng of RAMP 1 DNA. b, competition
`of 1251-labeled amylin binding to cells cotransfected with hCTR11 _ and
`RAMP 3 DNA. Salmon CT (sCT, F), rat amy lin (AMY,/ ), human CGRPa
`(CGRP, a), human CT (hCT, f). Data are from a single representative
`experiment with triplicate repeats (n = 3). 1C50 values for pooled data are
`shown in Table 1.
`
`Q-1--.---....---r---._,..---r---..,.--
`150
`100
`200
`50
`250
`0
`RAMP DNA tranefected (ng)
`Fig. 2. Effect of transfection of RAMP 1 (F), RAMP 2 (E), RAMP 3 (9 ), or
`vector (E) DNA on the expression of 1251-labeled amylin binding in
`CHO-K1 cells that endogenously express a native CT receptor. RAMP 1
`and RAMP 3, but not RAMP 2 or vector DNA, increased specific 1251-
`labeled amylin binding. Data are from a representative experiment with
`triplicate repeats (n = three separate experiments).
`
`TABLE 1
`1C60 values (nM) for peptides in competition for 125l-labeled rat amylin
`binding to COS-7 cells cotransfected with 100 ng of hCTR11 _ and 100
`ng of RAMP (n = 3)
`
`RAMP 1
`
`RAMP3
`
`<>-TSH Cells*
`
`Competing peptide
`Rat amylin
`Salmon CT
`Human CT
`Human CGRPa
`
`7.86 :t 4.49
`0.309 ± 0.144
`447 :t 78
`4.68 :t 3.04
`
`6.35
`0.206
`862
`151
`
`1.70
`0.004
`619
`48**
`
`10.5 :t 0.32
`0.37 :t 0.10
`N.D.
`5.95 :t 1.70
`
`• From Perry et al., 1997.
`** P < ,05 RAMP 1 versus RAMP 3 (n = 3).
`
`4
`
`

`

`RAMPs Generate Amylin Receptors from Calcitonin Receptors
`
`239
`
`tor protein (Quiza et al., 1997). In contrast, deglycosylation of
`the 1251-labeled amylin-binding protein with Endo F gener(cid:173)
`ated two distinct bands, a lower band with a Mr of ~ 54,000
`and an upper band with a Mr of ~ 68,000 (Fig. 5). Similar
`results were seen in HEK-293 cells stably transfected with
`rCTRu - receptor homolog and transfected with RAMP 1 (not
`shown).
`Confocal microscopic localization of RAMP 1 incorporating
`a c-myc epitope tag in the N terminus revealed that, when
`transfected alone, little of the protein was expressed on the
`
`cell surface (Fig. 6, a and b). However, when cotransfected
`with rCTR11 _, significant cell-surface expression of the pro(cid:173)
`tein was observed (Fig. 6c).
`
`Discussion
`The discovery of RAMPs and the elucidation of their role in
`the trafficking of CRLR and its expressed cell-surface pheno(cid:173)
`type provided a novel potential mechanism for the diversifi(cid:173)
`cation and the regulation of receptor function. However, the
`
`TABLE 2
`IC_,0 values (M) for peptides in competition for 1251-labeled salmon CT binding to COS-7 cells cotransfected with 100 ng of hCTR11 _ and 100 ng of
`RAMP (n = 3)
`
`Receptor Alone
`
`RAMP!
`
`RAMP2
`
`RAlv1P3
`
`Competing peptide
`Rat amylin
`Salmon CT
`Human CT
`Human CGRPa
`
`5.37:!: 2.84 X 10- 7
`9.33 :!: 6.08 X 10-10
`1.55 :!: 0.23 X 10- 8
`> 10 6
`
`1.72 :!: 0.73 X 10·· 7
`2.24 :!: 0.30 X 10- 10
`2.95 :!: 0.52 X 10- 8
`> 10 6
`
`8.78
`1.34
`8.27
`
`1.96 X 10 ·· 8
`0.46 X 10- 10
`1.57 X 10- n
`> 10 6
`
`1.96 :!: 0.95 X 10 7
`3.17:!: 0.35 X 10- 10
`1.97 :!: 1.06 X 10- 8
`10 6
`
`CGRP responses
`
`b
`120
`
`100
`
`80
`
`60
`
`40
`
`20
`
`a 120
`
`Amylln responses
`
`E
`:::s e
`·= e
`ti
`~
`fl. I
`
`'5
`
`~
`0. I ()
`
`c
`
`100
`
`80
`
`60
`
`40
`
`20
`
`0
`
`120
`
`100
`
`so
`
`60
`
`40
`
`20
`
`0
`
`-13
`
`·12
`
`-11
`
`-10
`
`·9
`
`·8
`
`human CT responses
`
`R1
`R2
`
`-13
`
`-10
`-9
`·12
`·11
`Log10 [peptide) (M)
`
`.a
`
`Fig. 4. Effect of RAMPs on ligand-induced cAMP responses in COS-7 cells cotransfected with 100 ng ofhCTR11 _ and 100 ng of vector control((:), RAMP
`1 (F), RAMP 2 (E), or RAMP 3 (a ) DNA. a, amylin responses; b, CGRP responses; c, human CT responses; d, salmon CT responses. Data are from a
`single representative experiment (n = 3). Arrows indicate the direction of potency shift for each of the peptides in response to RAMP cotransfection.
`Basal cAMP levels were -350, 600, 720, and 400 pmol/ml/106 cells for vector-, RAMP 1-, RAMP 2-, and RAMP 3-cotransfected cells, respectively.
`Maximal responses were -2150, 2700, 3800, and 2200 pmol/ml/10 6 cells for vector-, RAMP 1-, RAMP 2-, and RAMP 3-cotransfected cells, respectively.
`
`5
`
`

`

`240
`
`Christopoulos et al.
`
`role of RAMPs beyond their interaction with CRLR is un(cid:173)
`clear. CRLR shares ~55% amino acid sequence identity with
`the CT receptor and is almost 80% identical in the transmem(cid:173)
`brane regions. This homology suggested that the CT receptor
`protein might also be a target for RAMP interaction. In this
`study, we demonstrate that RAMPs do indeed interact with
`the CT receptor gene product to induce novel receptor phe(cid:173)
`notypes. RAMP 1 cotransfection with CT receptors generated
`an amylin receptor equivalent to that identified in mouse
`a-TSH cells (Hanna et al., 1995; Perry et al., 1997). The
`profile of peptide interaction was also similar to amy lin re(cid:173)
`ceptors characterized in the nucleus accumbens (Beaumont
`et al., 1993), the kidney (Wookey et al., 1996), and skeletal
`muscle (Pittner et al., 1996). Although the affinity ofpeptides
`interacting with the nucleus accumbens appears higher
`(Beaumont et al., 1993), this is likely attributable to, at least
`in part, the difference in assay format, with live cells being
`used in the current study and membranes being used for the
`measurement of nucleus accumbens binding. Indeed, analy(cid:173)
`sis of binding competition in brain slices also yields lower
`affinity for competing peptides (Sexton et al., 1994b).
`Comparison of the RAMP 1- and RAMP 3-induced receptor
`phenotypes indicates that there are different forms of amy lin
`receptor with differential sensitivity to CGRP, and there is
`evidence for this in tissue preparations. Differential sensitiv(cid:173)
`ity of amylin binding to competition by CGRP within rat
`brain nuclei has been suggested by the results of autoradio(cid:173)
`graphic studies (van Russum et al., 1994). The disparity in
`affinity is modest, ~ 10-fold at most, but it is consistent with
`the difference between RAMP 1- and RAMP 3-induced recep(cid:173)
`tor profiles. Amylin binding to regions such as the dorsome(cid:173)
`dial and arcuate hypothalmic nuclei (low CGRP affinity) is
`consistent predominantly with the RAMP 3-induced pheno(cid:173)
`type. Amylin receptors in the nucleus accumbens core and
`the amygdala (high CGRP affinity) resemble the RAMP
`1-induced phenotype. Elsewhere, the affinity of CGRP is
`intermediate, which may imply varying levels of mixed-re-
`
`ceptor phenotypes. Both RAMP 1 and RAMP 3 are expressed
`significantly in brain (McLatchie et al., 1998). For cells in
`which an amylin receptor phenotype is induced, it is unclear
`why the relative potency of ligands in competition for 1 2"1·
`labeled salmon CT binding is not significantly altered. How(cid:173)
`ever, it is likely that cells cotransfected with CT receptor and
`
`a
`
`b
`
`"'1-aalmon CT
`
`"'1-i'at amylln
`
`vector
`
`RAMP1
`
`vector
`
`RAMP1
`
`c
`
`M,(x10 .. )
`
`-82
`
`-88
`
`100-
`90-
`
`70-
`eo-
`50-
`
`40-
`
`30-
`
`+
`
`+
`
`eCT 10• M
`Endo F
`
`- + amylln 10• M
`- + -
`-
`- + -
`- +
`- + - Endo F
`- +
`Fig. 5. Transfection of RAMP 1 with hCTR11 _ into COS-7 cells enables
`BS" to specifically cross-link 12' 1-labeled rat amylin to a broad glycosy(cid:173)
`lated protein with a M, of - ·80,000. Binding of 1 ~ 51-labeled sCT essen(cid:173)
`tially was unaltered by cotransfection of RAMP 1. However, after Endo F
`deglycosylation of cross-linked RAMP 1 transfected cells, two bands with
`aM, of - 68,000 and -54,000 were seen with 12'1-labeled amy lin. In cells
`transfected with hCTRn- alone, only the lower M, band was seen after
`cleglycosylation of 1251-labeled salmon CT cross-linked receptor. The data
`are from an individual representative experiment (n = 3).
`
`Fig. 6. Confocal microscopic localization of c-myc RAMP 1 transfected
`(c). a and c, cell surface labeling in
`alone (a, b) or with the rCTR11
`nonpermeabilized cells. b, labeling in cells permeabilizecl with 0.3% Tri(cid:173)
`ton X-100. The data are from an individual representative experiment
`(n = 6). The data indicate the strong induction of cell-surface expression
`of RAMP 1 with receptor coexpression.
`
`6
`
`

`

`RAMPs Generate Amylin Receptors from Calcitonin Receptors
`
`241
`
`RAMP 1 or RAMP 3 express mixed amylin-CT receptor phe(cid:173)
`notypes. Furthermore, we have speculated that specificity of
`peptides in competition 125I-labeled salmon CT binding is
`more reflective of affinity for inactive state receptor (Hous(cid:173)
`sami et al., 1995).
`These data show that in cells expressing CTRn_, RAMP
`expression determines the extent to which they respond to
`the CT family of peptides. For the majority of experiments,
`the expression of CTRn-• through cotransfection with
`RAMP, occurs together with RAMP. However, we also have
`demonstrated that RAMP expression gives rise to novel amy(cid:173)
`lin receptors in CHO-K1 cells endogenously expressing CT
`receptors and in cells stably expressing the rCTRn - . In the
`latter cell line, unlike COS-7 cells, only RAMP 1 was capable
`of inducing amylin-receptor binding, which suggests that
`cellular background, including native RAMP levels and, po(cid:173)
`tentially, other components such as G protein levels, plays a
`significant role in the derived receptor phenotype.
`As observed previously for CGRP and adrenomedullin
`(McLatchie et al., 1998), the expression of RAMP alone did
`not enable binding of either amy lin or salmon CT, indicating
`that RAMPs are not receptors by themselves. Confocal mi(cid:173)
`croscopic analysis of RAMP 1 distribution indicated that
`little cell-surface expression of the protein occurred when
`transfected alone, although significant intracellular protein
`expression was observed. Cotransfection of CT receptor with
`RAMP 1 induced the appearance of the RAMP at the cell
`surface, as has been observed previously with cotransfection
`of CRLR and RAMP 1 (McLatchie et al., 1998). However,
`unlike CRLR, which does not traffic to the cell surface in the
`absence of RAMP, CT receptor alone is strongly expressed at
`the cell surface, yielding a classical CT receptor phenotype.
`Thus, for the CT receptor gene product, RAMP appears to be
`acting principally as a phenotypic modulator and not as a
`trafficking protein. However, the possibility that RAMP may
`affect the processing and trafficking of newly formed CT
`receptor protein in the expression of novel receptor pheno(cid:173)
`types cannot be excluded.
`Although inconclusive, the results of deglycosylation stud(cid:173)
`ies with 1251-labeled amylin cross-linked to cell surface-ex(cid:173)
`pressed amylin receptor were suggestive of an association
`between the "CT receptor" protein and a protein the size of
`which was equivalent to RAMP 1, with the appearance of a
`band with a M r of - 64,000, in addition to the core protein
`band with aM, of - 54,000. Although it is possible that the
`higher-molecular-weight band reflects partial deglycosyla(cid:173)
`tion of the receptor, we believe that this is unlikely, because
`the vast majority of the salmon CT-binding protein runs as
`core protein under equivalent conditions. Thus, the data may
`indicate a close cell surface association of RAMP 1 (BS 3
`cross-links primary amino groups within -20 Aland the CT
`receptor gene product in the expression of the amy lin recep(cid:173)
`tor phenotypes. Although differences in the pattern of glyco(cid:173)
`sylation between the major amylin- and salmon CT-binding
`proteins occur in a-TSH cells (Perry et al., 1997), no apparent
`differences were seen in the current study, suggesting that
`alteration in the level of glycosylation is not required for the
`expression of amylin receptor phenotype. For the a-TSH
`amylin receptor, the additional carbohydrate occurred at a