British Journal of Pharmacology (2003) 140, 477–486
`
`& 2003 Nature Publishing Group All rights reserved 0007–1188/03 $25.00
`www.nature.com/bjp
`
`CL/RAMP2 and CL/RAMP3 produce pharmacologically distinct
`adrenomedullin receptors: a comparison of effects of
`adrenomedullin22–52, CGRP8–37 and BIBN4096BS
`
`1,2,5D.L. Hay, 1S.G. Howitt, 1A.C. Conner, 3M. Schindler, 4D.M. Smith & *,1D.R. Poyner
`
`1Pharmaceutical Sciences Research Institute, Aston University, Birmingham B4 7ET; 2Department of Metabolic Medicine,
`Imperial College School of Medicine, Hammersmith Hospital, Du Cane Road, London W12 0NN; 3Cardiovascular Research,
`Boehringer Ingelheim Pharma KG, Biberach 88397, Germany and 4AstraZeneca, CVGI, Mereside, Alderley Park, Macclesfield,
`Cheshire SK10 4TG
`
`1 Adrenomedullin (AM) has two known receptors formed by the calcitonin receptor-like receptor
`(CL) and receptor activity-modifying protein (RAMP) 2 or 3: We report the effects of the antagonist
`fragments of human AM and CGRP (AM22 – 52 and CGRP8 – 37) in inhibiting AM at human (h), rat (r)
`and mixed species CL/RAMP2 and CL/RAMP3 receptors transiently expressed in Cos 7 cells or
`endogenously expressed as rCL/rRAMP2 complexes by Rat 2 and L6 cells.
`2 AM22 – 52 (10 mm) antagonised AM at all CL/RAMP2 complexes (apparent pA2 values: 7.3470.14
`(hCL/hRAMP2), 7.2870.06 (Rat 2), 7.0070.05 (L6), 6.2570.17 (rCL/hRAMP2)). CGRP8 – 37 (10 mm)
`resembled AM22 – 52 except on the rCL/hRAMP2 complex, where it did not antagonise AM (apparent
`pA2 values: 7.0470.13 (hCL/hRAMP2), 6.7270.06 (Rat2), 7.0370.12 (L6)).
`3 On CL/RAMP3 receptors, 10 mm CGRP8 – 37 was an effective antagonist at all combinations
`(apparent pA2 values: 6.9670.08 (hCL/hRAMP3), 6.1870.18 (rCL/rRAMP3), 6.4870.20 (rCL/
`hRAMP3)). However, 10 mm AM22 – 52 only antagonised AM at the hCL/hRAMP3 receptor (apparent
`pA2 6.7370.14).
`4 BIBN4096BS (10 mm) did not antagonise AM at any of the receptors.
`5 Where investigated (all-rat and rat/human combinations), the agonist potency order on the CL/
`RAMP3 receptor was AMBbCGRP4aCGRP.
`rRAMP3 showed three apparent polymorphisms, none of which altered its coding sequence.
`6
`7 This study shows that on CL/RAMP complexes, AM22 – 52 has significant selectivity for the CL/
`RAMP2 combination over the CL/RAMP3 combination. On the mixed species receptor, CGRP8 – 37
`showed the opposite selectivity. Thus, depending on the species,
`it is possible to discriminate
`pharmacologically between CL/RAMP2 and CL/RAMP3 AM receptors.
`British Journal of Pharmacology (2003) 140, 477 – 486. doi:10.1038/sj.bjp.0705472
`Keywords: CGRP; CGRP8 – 37; adrenomedullin; adrenomedullin22 – 52; calcitonin receptor-like receptor; CL; RAMP2;
`RAMP3
`
`Abbreviations: AM, adrenomedullin; BIBN4096BS, 1-piperidinecarboxamide, N-[2-[[5amino-1-[[4-(4-pyridinyl)-1-piperazinyl]-
`carbonyl]pentyl]amino]-1-[(3,5-dibromo-4-hydroxyphenyl)methyl]-2-oxoethyl]-4-(1,4-dihydro-2-oxo-3(2H)-quina-
`zolinyl); CGRP, calcitonin gene-related peptide; CL, calcitonin receptor-like receptor; [cys(ACM)2,7]aCGRP,
`Cys-acetoamidomethyl2,7 human aCGRP; [cys(Et)2,7]aCGRP, Cys-ethylamide2,7 human aCGRP; r, rat; RAMP,
`receptor activity-modifying protein
`
`Introduction
`
`its presence in the
`AM is an essential vascular peptide;
`developing foetus governs the formation of intact vasculature
`and thus foetal survival (Caron & Smithies, 2001; Shindo et al.,
`2001). These studies support the well-recognised role of AM in
`cell growth, and further the concept that AM may be involved
`in angiogenesis (Miller et al., 1996; Withers et al., 1996;
`Nikitenko et al., 2002). The pharmacology of receptors
`responsive to AM has been examined in many tissues and cell
`lines (see Hinson et al., 2000; Hay & Smith, 2001 for reviews).
`
`*Author for correspondence; E-mail: D.R.Poyner@aston.ac.uk
`5Current address: The Liggins Institute, University of Auckland, 2-6
`Park Ave, Grafton, Auckland, New Zealand
`Advance online publication: 26 August 2003
`
`Specific AM receptors can be characterised by high affinity for
`AM and X100-fold lower affinity for the other members of the
`calcitonin family of peptides (Coppock et al., 1999). The
`effects of AM at such receptors can be inhibited by the AM
`(Eguchi et al., 1994). The
`antagonist
`fragment AM22 – 52
`aCGRP antagonist
`fragment CGRP8 – 37 and the amylin
`receptor antagonist AC187 can also antagonise specific AM
`receptors, but only at high concentrations (45 mm) (Coppock
`et al., 1999). AM also activates CGRP receptors, and these
`effects can be inhibited by CGRP8 – 37 (Nagoshi et al., 2002).
`However, this description is likely to be an oversimplification.
`AM22 – 52 is the only antagonist available which is specific for
`AM receptors, and this is a relatively low-affinity peptide.
`Without better antagonists, it is difficult to separate receptor
`
`1
`
`EX2057
`Eli Lilly & Co. v. Teva Pharms. Int'l GMBH
`IPR2018-01426
`
`

`

`478
`
`D.L. Hay et al
`
`Heterogeneity of adrenomedullin receptors
`
`subtypes which may exist in tissues that are likely to contain
`very complex mixtures of receptors, for example, vas deferens
`(Poyner et al., 1999; Wu et al., 2000).
`Generally, it seems that AM22 – 52 antagonises the effects of
`AM but not CGRP. However, high concentrations of AM22 – 52
`are often required due to the low affinity of this antagonist,
`and observations of unusual pharmacology, potentially
`attributable to the existence of subtypes of AM receptors,
`have been noted. In the rat vas deferens, the effects of AM or
`[Cys(Et)2,7]aCGRP (a putative ‘CGRP2’-receptor
`selective
`agonist; Dumont et al., 1997) were more potently antagonised
`by BIBN4096BS than those of either a or bCGRP (Wu et al.,
`2000). In the hind limb vascular bed of the cat, AM22 – 52 could
`not antagonise the effects of AM, but could inhibit the effects
`of CGRP (Champion et al., 1997). In this system, CGRP8 – 37
`could inhibit responses to CGRP but not AM.
`Two AM receptor subtypes have now been defined in
`molecular terms: AM1, composed of CL with RAMP2, and
`AM2, composed of CL and RAMP3 (McLatchie et al., 1998;
`Poyner et al., 2002). RAMP2 and RAMP3 can be differentially
`regulated in in vivo models of the disease (Ono et al., 2000).
`For example, in a rat model of obstructive neuropathy, CL,
`RAMP1 and RAMP2 mRNA levels were upregulated, but
`RAMP3 levels were unchanged (Nagae et al., 2000). At
`present, there is no pharmacological separation of AM1 and
`AM2 receptors although it has been reported that a mouse
`RAMP3/rat CL (rCL) complex is more sensitive to the effects
`of CGRP than its RAMP2 counterpart (Husmann et al.,
`2000). The mouse RAMP3/rat CL receptor is considered a
`mixed AM/CGRP receptor, but in terms of the effects of
`antagonists, specific AM1 and AM2 receptors have never been
`thoroughly characterised. Recent studies have examined the
`effects of CGRP8 – 37 and AM22 – 52 at human, bovine and
`porcine CL complexes with RAMPs1 – 3 (Aiyar et al., 2001;
`2002). However, these studies were performed in HEK293 cells
`which are known to express endogenous RAMPs (particularly
`RAMP2) and/or CL (Aiyar et al., 1996; Kuwasako et al.,
`2001). Most other analyses have been based on binding
`studies. Therefore, the functional effects of these antagonists at
`exclusive CL/RAMP3 complexes have never been examined.
`Furthermore, although CL has been cloned from several
`species (Elshourbagy et al., 1998; Aiyar et al., 2001; 2002),
`these have usually been coexpressed with human RAMPs
`(hRAMPs). There has been no study of a non-human CL
`expressed with a RAMP from the same species. It is not known
`how well these mixed species receptors reflect the pharmacol-
`ogy of the homologous receptors. In turn, this means that
`there is no reliable information on species variation from
`exogenously expressed, recombinant receptors.
`RAMP2 and RAMP3 are divergent in sequence (Sexton
`et al., 2001), and the regions of RAMP2 and RAMP3 with
`which AM interacts are not conserved between the two
`proteins (Kuwasako et al., 2001; 2002). This suggests that
`there could be pharmacological differences between these
`receptors. Therefore, this study was designed to make a
`detailed comparison of
`the highest affinity antagonists
`available for studying AM/CGRP receptors. At the same
`time,
`the effect of species composition on the observed
`pharmacology was also investigated. AM22 – 52 (Eguchi et al.,
`1994), CGRP8 – 37 (Chiba et al., 1989; Dennis et al., 1990) and
`the novel CGRP receptor antagonist BIBN4096BS (Doods
`et al., 2000) were used to inhibit AM responses at AM
`
`British Journal of Pharmacology vol 140 (3)
`
`receptors composed of various combinations of rat or human
`CL, RAMP2 or RAMP3. A more limited investigation into
`agonist potency ratios on CL/RAMP3 receptors was also
`carried out. This work was done using Cos 7 cells which have a
`null CL/RAMP background, making it an ideal cell line for
`studying the pharmacology of single populations of AM1 or
`AM2 receptors.
`
`Methods
`
`Cell culture
`
`Cos 7, Rat-2 and L6 cells were cultured in Dulbecco’s modified
`Eagles medium supplemented with 10% foetal bovine serum
`and 5% penicillin/streptomycin in a humidified 95% air/5%
`CO2 atmosphere. The cells were subcultured by removing the
`growth medium and washing the cells with cell culture-grade
`phosphate-buffered saline for 1 – 2 min. The cells were removed
`from the flasks with a small volume of trypsin/EDTA solution.
`Fresh growth medium was added to the cell suspension to
`neutralise the trypsin, and the cells were centrifuged at 500 g
`for 5 min. The supernatant was removed, and the cell pellet
`was resuspended in fresh growth medium. The cells were
`transferred to fresh flasks, or plated onto 48-well plates.
`
`Cloning of rat RAMP3 (rRAMP3)
`
`rRAMP3 was cloned from a rat lung cDNA library (Invitro-
`gen) using primers based on the published sequence (Oliver
`et al., 2001). These were 50-CTCGAGATGGCGACCCCGG
`CACAGCGGCTGCACC-30 and 50-GAATTCTCACAGAA
`GCCGGTCAGTGTGCTTGCTACG-30. After 30 rounds of
`polymerase chain reaction (921C, 60 s; 601C, 60 s; 721C, 60 s),
`using Pfu polymerase (Promega), the amplified product was
`identified as a 0.48 kilobase band on a 1.4% agarose gel. Its
`identity was confirmed by sequencing (Alta Biosciences,
`Birmingham, U.K.). The product was subcloned into pcDNA3
`using restriction enzyme sites EcoRI and BamHI, which were
`included in the primer design.
`
`Transient transfection
`
`Cells were transfected with various combinations of hRAMP2,
`hRAMP3, rRAMP3, N-terminally HA epitope-tagged hCL
`(kindly donated by Dr S.M. Foord, GSK, Stevenage, U.K.) or
`rCL (Njuki et al., 1993), using the calcium phosphate
`(Clontech) method of transient transfection. Transfections
`were undertaken essentially according to the manufacturer’s
`instructions, but with minor modifications. Test DNA (1 mg
`total per well of a 48-well plate) was mixed with sterile water
`and 2 m calcium chloride solution. This was mixed and left at
`room temperature for 10 min. The DNA mix was then added
`dropwise to an equal volume of HEPES-buffered saline. The
`HEPES-buffered saline was continually agitated as the DNA
`mix was added to it. This transfection solution was left at room
`temperature for 30 min. Ten times the volume of the normal
`growth medium was then added to the transfection solution.
`The old growth medium was replaced with the transfection
`solution. After a 5 – 16 h incubation period, the transfection
`mix was removed from the cells and replaced with fresh growth
`
`2
`
`

`

`medium. The plates were used for cyclic AMP assay 48 – 72 h
`after the medium was replaced.
`
`Results
`
`D.L. Hay et al
`
`Heterogeneity of adrenomedullin receptors
`
`479
`
`Characterisation of baseline receptor expression in Cos 7
`cells
`
`Careful characterisation of AM1 and AM2 receptors required
`the use of a batch of cells for transfection studies, which do not
`express CL or RAMPs endogenously. Cos 7 cells have
`previously been reported to contain only low levels of RAMPs
`(Tilakaratne et al., 2000). In agreement with this, Figure 1
`shows that in cells transfected with CL alone, 100 nm AM
`(Figure 1a) or concentrations of CGRP up to 1 mm (Figure 1b)
`failed to cause any increase in cyclic AMP production. This
`demonstrates that the cells lack any endogenous RAMPs. The
`cells also failed to respond to these concentrations of AM and
`CGRP when transfected with RAMP1 or RAMP3 alone,
`showing the absence of any endogenous CL (Figure 1). By way
`of positive controls, the cells did respond to AM when
`transfected with hRAMP3 and rCL, and to CGRP when
`transfected with hCL and hRAMP1 (Figure 1). These
`cells were cultured for over 50 passages and tested in this
`way every few passages. On no occasion were endogenous
`receptor components evident, making this a suitable cell line
`for characterising transfected AM1 and AM2 receptors.
`Attempts were made to further characterise the putative AM
`receptor L1 (Kapas et al., 1995) in transfection experiments
`using COS 7 cells. However, on no occasion was elevation of
`
`SFM
`
`100nM AM
`
`hRAMP3
`
`rCL
`
`pcDNA3
`
`rCL/hRAMP3
`
`hCL/hRAMP1
`
`hRAMP1
`
`hCL
`
`100
`
`75
`
`50
`
`25
`
`0
`
`a
`
`(% of maximum)
`cAMP production
`
`100
`
`75
`
`50
`
`25
`
`0
`
`b
`
`(% ofmaximum)
`cAMP production
`
`-13
`
`-12
`
`-11
`
`-8
`-9
`-10
`Log[CGRP]
`
`-7
`
`-6
`
`-5
`
`Figure 1 Characterisation of Cos 7 cells, (a) Responses of cells
`transfected with rCL/hRAMP3, cloning vector (pcDNA3), rCL and
`hRAMP3, and subsequently challenged with either 100 nm AM or
`serum-free medium (SFM). (b) Concentration – response curves to
`haCGRP in cells
`transfected with hCL/hRAMP1, hCL and
`hRAMP1. Points are the mean7s.e.m. of triplicate determinations.
`These are representative data from experiments repeated 10 times.
`
`British Journal of Pharmacology vol 140 (3)
`
`Assay of cyclic AMP production
`
`The growth medium was removed from the cells and replaced
`with serum and antibiotic-free Dulbecco’s modified Eagle
`medium containing 500 mm isobutyl methyl xanthine for
`30 min. All drugs were diluted in the same medium.
`Antagonists were added for 15 min before the addition of
`agonists in the range 1 pm – 1 mm for a further 15 min. Cyclic
`AMP was extracted with ice-cold 95 – 100% ethanol. Cyclic
`AMP was measured by radio-receptor assay as previously
`described (Poyner et al., 1992).
`
`Analysis of data
`
`For cyclic AMP studies, the data from each concentration –
`response curve were fitted to a sigmoidal concentration –
`response curve to obtain the maximum response, Hill
`coefficient and EC50, using the fitting routine PRISM
`Graphpad. From the individual curves, dose ratios were
`calculated. Where three antagonist concentrations were used, a
`Schild plot was constructed; after confirming that the slope
`was not significantly different from unity, it was constrained to
`1 to obtain the pKb. Where only one or two antagonist
`concentrations were used, an apparent pA2 was calculated
`from the formula log[antagonist]log(dose ratio1), after first
`confirming that there were no significant differences in the Hill
`coefficient or maximum response between the concentration –
`response curves in the presence and absence of antagonist.
`Statistical analysis was carried out by Student’s t-test, or by
`one-way ANOVA followed by Tukey’s test (where every value
`was compared against each other), or Dunnett’s test (where
`several values were being compared against a single control).
`The significance was accepted at Po0.05; two-tailed tests were
`used throughout. All values are quoted as means7s.e.m.
`
`Materials
`
`Rat AM and human AM22 – 52 were obtained from Bachem (St
`Helens, Merseyside, U.K.). Human aCGRP (haCGRP) and
`human aCGRP8 – 37 (haCGRP8 – 37) were from Calbiochem
`(Beeston, Nottingham, U.K.) or Neosystems (Strasbourg,
`France). [cys(ACM)2,7]aCGRP, rat amylin and rat calcitonin
`were from Bachem (St Helens, U.K.). [cys(Et)2,7]aCGRP was
`from Phoenix Pharmaceuticals (Mountain View, CA, U.S.A.),
`and humanbCGRP (hbCGRP) was from Sigma (Gillingham,
`Dorset, U.K.). Salmon calcitonin was purchased from Cam-
`bridge Research Biochemicals (Northwich, Cheshire, U.K.).
`All peptides were dissolved in distilled water and stored as
`aliquots at 201C or 701C (AM and AM22 – 52) in nonstick
`microcentrifuge tubes (Thermo Life Sciences, Basingstoke,
`U.K.). BIBN4096BS was a gift
`from Dr M. Schindler
`(Boehringer-Ingelheim, Biberach, Germany), and was pre-
`pared as previously described (Hay et al., 2002). Unless
`otherwise specified, chemicals were from Sigma or Fisher
`(Loughborough, U.K.). Cell culture reagents were from Gibco
`BRL (Paisley, Renfrewshire, U.K.) or Sigma.
`
`3
`
`

`

`480
`
`D.L. Hay et al
`
`Heterogeneity of adrenomedullin receptors
`
`Numbersinparenthesesarenvalues.*,**,***SignificantlydifferentfromcontrolpEC50,Dunnett’stestatPo0.05,0.01and0.001levels,respectively.ND,notdetermined.
`
`ND
`
`7.970.20(4)
`7.7270.28(3)
`8.5570.13(3)
`9.0170.09(3)
`8.8370.13(3)
`7.6470.72(3)
`
`ND
`
`8.3870.21(4)
`8.4970.21(3)
`8.7070.14(3)
`8.7570.11(3)
`8.8870.21(3)
`8.6170.72(3)
`
`AMalone
`
`10mm
`
`+BIBN4096BS
`
`7.0570.39(5)*
`7.0670.14(3)**
`7.5270.19(6)***
`
`–
`
`–
`
`7.3670.22(4)
`7.2170.2(6)***
`
`7.4370.07(5)
`
`–
`
`8.1770.17(3)**
`7.3470.01(3)*
`7.7770.01(3)*
`
`–
`
`7.5670.1(3)**
`
`10mm
`
`+CGRP8–37
`
`1mm
`
`8.2670.72(5)
`8.5670.25(3)
`9.4370.22(6)
`8.3970.08(4)
`8.7570.11(3)
`8.2370.3(4)
`9.0670.2(7)
`
`AMalone
`
`pEC507s.e.m(n)
`
`7.9170.24(3)
`8.2770.36(3)
`7.3470.16(4)***
`
`–
`
`–
`
`6.770.07(3)**
`6.4370.28(4)***
`
`–
`
`–
`
`8.2470.15(3)*
`
`–
`
`–
`
`–
`
`6.8370.14(4)***
`
`8.0670.13(3)
`8.7970.47(3)
`8.4170.13(3)*
`7.6470.14(3)*
`7.4570.15(3)*
`8.3270.42(3)
`7.2870.33(3)**
`
`10mm
`
`3mm
`
`+AM22–52
`
`1mm
`
`8.3070.17(3)
`8.8970.25(6)
`9.2670.21(7)
`8.7070.14(3)
`8.7670.12(3)
`8.1670.14(6)
`8.7470.08(6)
`
`AMalone
`
`rCL/rRAMP3
`rCL/hRAMP3
`hCL/hRAMP3
`
`L6
`
`rCL/rRAMP2Rat-2
`rCL/hRAMP2
`hCL/hRAMP2
`
`Receptor
`
`Table1ActionsofantagonistsonCL/RAMP2andCL/RAMP3complexes
`
`cyclic AMP evident in response to 1 mm of either AM or CGRP
`(data not shown). The effects of transfection of L1 in the
`presence of RAMPs were not examined in this
`study.
`Concurrent transfection of RAMPs with L1 was reported to
`be without effect in a previous study (Chakravarty et al., 2000).
`
`Characterisation of rRAMP3
`
`from the previously
`rRAMP3 showed three differences
`reported sequence (Oliver et al., 2001). Codon 128, previously
`reported as CTG, was TTG; codon 134, previously reported as
`GGC, was GGG, and codon 137, previously reported as GTG,
`was GTA. None of these alter the amino acids (i.e. L128, G134
`and V137). As Pfu polymerase, used for the polymerase chain
`reaction, has stringent proofreading ability, the probability of
`obtaining three errors as a result of this process, none of which
`alter the coding sequence, is very remote. Accordingly, these
`are likely to be polymorphisms.
`
`Effect of antagonists on AM responses in hCL/hRAMP2-
`transfected Cos 7 cells
`
`The effects of AM on cyclic AMP responses in hCL/hRAMP2
`cotransfected cells in the presence or absence of AM22 – 52,
`CGRP8 – 37 and BIBN4096BS are shown in Table 1. In the
`presence of AM22 – 52, the concentration – effect curve to AM
`was shifted to the right in a parallel fashion (Figure 2a). These
`data were used to generate a Schild plot (Figure 2e). As the
`slope of the line was not significantly different from unity, the
`slope was constrained to 1, and a pKB of 7.3470.14 (n¼ 11)
`estimated. CGRP8 – 37 also produced a significant change in the
`pEC50 to AM, with no significant change in Hill coefficient or
`maximum response (Table 1, Figure 2b). This antagonist was
`slightly less potent than AM22 – 52 (apparent pA2 7.0470.15,
`n¼ 9, Figure 2b), although the difference was not significant.
`BIBN4096BS at 10 mm had no significant effect on the response
`to AM (Table 1).
`
`Effect of antagonists on AM responses in rCL/hRAMP2
`Cos 7 cells
`
`The effects of AM on cyclic AMP responses in rCL/hRAMP2
`cotransfected cells are shown in Table 1. pEC50 values in the
`presence of AM22 – 52, CGRP8 – 37 and BIBN4096BS are also
`shown. In the presence of AM22 – 52, the concentration – effect
`curve to AM was shifted to the right in a parallel fashion (Hill
`slopes; control 0.7770.5, 1 mm AM22 – 52 0.570.1, 10 mm AM22 – 52
`1.2570.5). An apparent pA2 of 6.2570.17 (n¼ 3) (Figure 2c)
`was estimated from the shift. There was no significant
`difference in the pEC50 value to AM obtained in the presence
`of BIBN4096BS or CGRP8 – 37 (Table 1, Figure 2d).
`
`Effect of antagonists on rat AM1 receptors endogenously
`expressed in Rat-2 and L6 cell lines
`
`Rat-2 and L6 cell lines have previously been demonstrated to
`express CL and RAMP2 (Choksi et al., 2002), and are
`therefore good models of rat AM1 receptors. In L6 cells,
`AM22 – 52 (1 mm) produced a significant rightward shift in the
`concentration – effect curve to AM (Table 1, Figure 3a). From
`this shift, an apparent pA2 of 7.0070.05 (n¼ 3) was generated.
`CGRP8 – 37 (1 mm) was also effective at inhibiting the effects of
`
`British Journal of Pharmacology vol 140 (3)
`
`4
`
`

`

`D.L. Hay et al
`
`Heterogeneity of adrenomedullin receptors
`
`481
`
`L6 cells
`
`AM
`+ 1µM CGRP8-37
`
`100
`
`75
`
`50
`
`25
`
`0
`
`-13
`
`-12
`
`-11
`
`-8
`-9
`-10
`Log [AM]
`
`-7
`
`-6
`
`-5
`
`Rat 2 cells
`
`b
`
`cAMP production (% ofmaximum)
`
`d
`
`AM
`+ 1µM AM22-52
`
`100
`
`75
`
`L6 cells
`
`AM
`+ 1µM AM22-52
`+ 10µM BIBN
`
`100
`
`75
`
`50
`
`25
`
`0
`
`-13
`
`-12
`
`-11
`
`-8
`-9
`-10
`Log [AM]
`
`-7
`
`-6
`
`-5
`
`L6 cells
`
`100
`
`CGRP
`+1µM AM22-52
`
`a
`
`cAMP Production (% ofmaximum)
`
`c
`
`hCL/hRAMP2
`
`AM
`+ 1µM CGRP8-37
`+ 10µM CGRP8-37
`
`100
`
`75
`
`50
`
`25
`
`0
`
`-13
`
`-12
`
`-11
`
`-8
`-9
`-10
`Log [AM]
`
`-7
`
`-6
`
`-5
`
`rCL/hRAMP2
`
`100
`
`AM
`+ 10µM CGRP8-37
`
`b
`
`cAMP production (% of maximum)
`
`d
`
`hCL/hRAMP2
`
`AM
`+ 1µM AM22-52
`+ 3µM AM22-52
`+ 10µM AM22-52
`
`100
`
`75
`
`50
`
`25
`
`0
`
`-13
`
`-12
`
`-11
`
`-8
`-9
`-10
`Log [AM]
`
`-7
`
`-6
`
`-5
`
`rCL/hRAMP2
`
`a
`
`cAMP production (% of maximum)
`
`c
`
`100
`
`50
`
`25
`
`0
`
`-13
`
`-12
`
`-11
`
`-10
`
`-9
`
`-8
`
`-7
`
`-6
`
`-5
`
`cAMP production (% of maximum)
`
`75
`
`50
`
`25
`
`0
`
`-13
`
`-12
`
`-11
`
`-10
`
`-9
`
`-8
`
`-7
`
`-6
`
`-5
`
`cAMP production (% of maximum)
`
`Log [CGRP]
`
`Log [AM]
`
`the stimulation of cyclic AMP
`Figure 3 Characterisation of
`production by rat AM in L6 and Rat 2 cells (endogenous rCL/
`rRAMP2). Points are the mean7s.e.m. of triplicate determinations.
`Concentration – response curves are representative of three or four
`experiments. Data are expressed as the percentage of maximum
`cyclic AMP production, estimated by fitting each line to a logistic
`Hill equation, as described in Methods. Maximum cyclic AMP
`values were 290720 pmol per 106 cells for Rat 2 cells, and
`20007300 pmol per 106 cells for L6 cells; basal values were all
`below 10 pmol per 106 cells, (a) L6 cells, AM22 – 52 and BIBN4096BS
`against rAM; (b) L6 cells CGRP8 – 37 against rAM; (c) L6 cells,
`AM22 – 52 against aCGRP, (d) Rat 2 cells, AM22 – 52 against rAM.
`
`CGRP was inactive on Rat-2 cells at concentrations of up to
`1 mm, in accordance with published data (Coppock et al., 1999)
`(n¼ 3, data not shown), but AM caused a concentration-
`dependent stimulation of cyclic AMP production, as shown in
`Table 1. AM22 – 52 (1 mm) caused a rightward shift in the
`concentration – effect curve to AM, with an apparent pA2 of
`7.2870.06 (n¼ 3, Table 1, Figure 3d). We have previously
`demonstrated that the AM response in these cells can be
`1 mm CGRP8 – 37, but not by
`10 mm
`antagonised by
`BIBN4096BS (Hay et al., 2002). These data are included in
`Table 1 for comparison with the data presented here.
`
`Effect of antagonists on AM responses in hCL/hRAMP3
`Cos 7 cells
`
`In hCL/hRAMP3 cotransfected cells, the concentration – effect
`curve to AM was shifted to the right in the presence of AM22 – 52
`or CGRP8 – 37 (Table 1, Figure 4a, b). Figure 2e shows the
`Schild plot generated from the antagonist shifts with AM22 – 52,
`from which a pKB of 6.7370.14 (n¼ 10) was estimated. It was
`significantly less potent at this receptor (Po0.01) than at the
`hCL/hRAMP2 complex. An apparent pA2 of 6.9670.08
`(n¼ 9, Figure 4b) was generated for CGRP8 – 37. This was not
`significantly different from its effects at the hCL/hRAMP2
`complex. BIBN4096BS was inactive at up to 10 mm.
`
`Effect of antagonists on AM responses in rCL/hRAMP3
`Cos 7 cells
`
`In the presence of CGRP8 – 37, the concentration – effect curve
`to AM was shifted to the right in a parallel fashion (Hill slopes;
`control 1.370. 16, 1 mm CGRP8 – 37 1.6370.38, 10 mm CGRP8 – 37,
`
`British Journal of Pharmacology vol 140 (3)
`
`75
`
`50
`
`25
`
`0
`
`-13
`
`-12
`
`-11
`
`-10
`
`-9
`
`-8
`
`-7
`
`-6
`
`-5
`
`Log [AM]
`
`cAMP production (% of maximum)
`
`AM
`+ 1µM AM22-52
`+ 10µM AM22-52
`
`75
`
`50
`
`25
`
`0
`
`-13
`
`-12
`
`-11
`
`-10
`
`-9
`
`-8
`
`-7
`
`-6
`
`-5
`
`cAMP production (% ofmaximum)
`
`hCL/hRAMP2
`hCL/hRAMP3
`
`-6.0
`
`-5.5
`Log [AM22-52]
`
`-5.0
`
`-4.5
`
`log [AM]
`
`3.0
`
`2.5
`
`2.0
`
`1.5
`
`1.0
`
`e
`
`log (DR-1)
`
`0.5
`-6.5
`
`the stimulation of cyclic AMP
`Figure 2 Characterisation of
`production by rat AM in Cos 7 cells transfected with CL/RAMP2
`combinations. Points are the mean7s.e.m. of triplicate determina-
`tions. Concentration – response curves are representative of three to
`seven experiments. Data are expressed as
`the percentage of
`maximum cyclic AMP production, estimated by fitting each line
`to a logistic Hill equation, as described in Methods. Maximum cyclic
`AMP values were 250720 pmol per 106 cells for hCL/hRAMP2, and
`450760 pmol per 106 cells for hCL/hRAMP3; basal values were all
`below 10 pmol per 106 cells, (a) hCL/hRAMP2, AM22 – 52; (b) hCL/
`(c)
`(d)
`rCL/
`hRAMP2, CGRP8 – 37;
`rCL/hRAMP2, AM22 – 52;
`hRAMP2, CGRP8 – 37;
`(e) Schild plot, antagonism of AM by
`AM22 – 52 on hCL/hRAMP2 and hCL/hRAMP3 receptors.
`
`AM, eliciting a significant change in the pEC50 (Po0.05,
`Table 1, Figure 3b). An apparent pA2 value of 7.0370.12
`(n¼ 3) was calculated from these data. The effect of CGRP on
`cyclic AMP in these cells could not be inhibited by 1 mm AM22 – 52
`(Figure 3c). BIBN4096BS was unable to antagonise the effects
`of AM in L6 cells up to concentrations of 10 mm (Table 1,
`Figure 3a). Slow kinetics of BIBN4096BS have previously been
`reported (Schindler & Doods, 2002); hence, the incubation
`time for pretreatment with BIBN4096BS was increased from
`15 to 60 min. However, this antagonist was still unable to
`inhibit the effects of AM. The pEC50 values were 8.5670.32
`without BIBN4096BS, compared to 8.3070.25 (both n¼ 2) in
`the presence of the antagonist. Therefore, the lack of effect
`of this antagonist in the studies described above is unlikely
`to be due to the short (15 min) antagonist incubation time.
`BIBN4096BS inhibited the binding of 125I-iodohistidyl-CGRP
`to membranes made from COS 7 cells cotransfected with hCL
`and hRAMP1, with a pKi of 10.8570.21. This is in line with
`its pKi, on SK-N-MC cells which also express hCL and
`hRAMP1 (Schindler & Doods, 2002), confirming that the
`antagonist was active.
`
`5
`
`

`

`482
`
`D.L. Hay et al
`
`Heterogeneity of adrenomedullin receptors
`
`Effect of antagonists on AM responses in rCL/rRAMP3
`Cos 7 cells
`
`At concentrations up to 10 mm, AM22 – 52 failed to antagonise
`the actions of AM (Table 1, Figure 4e). However, 10 mm
`CGRP8 – 37 did cause a significant shift in the AM concentra-
`from which an apparent pA2 of
`tion – response curve,
`6.1870.18 (n¼ 5, Figure 4f) was calculated.
`
`Effect of AM, a and bCGRP and other peptides in rCL/
`hRAMP3 and rCL/rRAMP3 Cos 7 cells
`
`hCL/hRAMP3
`
`AM
`+ 1µM CGRP8-37
`+ 10µM CGRP8-37
`
`100
`
`75
`
`50
`
`25
`
`0
`
`-13
`
`-12
`
`-11
`
`-9
`-10
`Log [AM]
`
`-8
`
`-7
`
`-6
`
`-5
`
`b
`
`cAMP production (% ofmaximum)
`
` hCL/hRAMP3
`
`AM
`+1µM AM22-52
`+3µM AM22-52
`+ 10µM AM22-52
`
`100
`
`75
`
`50
`
`25
`
`0
`
`-13
`
`-12
`
`-11
`
`-8
`-9
`-10
`Log [AM]
`
`-7
`
`-6
`
`-5
`
`a
`
`cAMP production (% of maximum)
`
`rCL/hRAMP3
`
`AM
` + 1µM AM22-52
`+ 10µM AM22-52
`
`c
`
`100
`
`75
`
`50
`
`25
`
`0
`
`rCL/hRAMP3
`
`AM
`+ 10µM CGRP8-37
`
`d
`
`100
`
`75
`
`50
`
`25
`
`0
`
`-11
`
`-8
`
`-7
`
`-6
`
`-5
`
`AM, a and bCGRP, all elevated cyclic AMP with pEC50 values
`as indicated in Table 2 and Figure 5a and b. For both
`receptors, the rank potency order was AM4bCGRPzaCGRP.
`In particular, on the rCL/rRAMP3 complex, AM and bCGRP
`were not significantly different in potency from each other
`(Tukey’s test, P40.05). Other peptides were also tested on
`rCL/hRAMP3 receptor. Here, aCGRP,
`[Cys(Et)2,7]aCGRP
`and rat amylin were equipotent. [Cys(ACM)2,7] aCGRP, rat
`calcitonin, salmon calcitonin and the three antagonists used
`above were unable to elevate cyclic AMP when acting at the
`rCL/hRAMP3 receptor.
`
`Discussion
`
`The functional effects of antagonists at complexes of CL with
`RAMP2 and RAMP3 are little studied. The limited data
`previously available suggested that there were no differences
`between these receptors, despite the low homology between
`RAMP2 and RAMP3 (Aiyar et al., 2001; 2002; Sexton et al.,
`2001). Furthermore, the studies that have looked at the
`detailed pharmacology of these receptors have usually been
`performed in HEK293 cells which are known to express
`RAMP2 endogenously. CL/RAMP3 has often been over-
`looked in the literature. There is a considerable lack of
`information on this receptor complex although AM stimulates
`cyclic AMP production with similar potency to its actions at
`CL/RAMP2 (McLatchie et al., 1998; Aiyar et al., 2001; 2002;
`Kuwasako et al., 2002). While expression of RAMP3 mRNA
`is low in rats (Chakravarty et al., 2000), in humans it is at least
`as abundant as RAMP2 (McLatchie et al., 1998).
`The data in this study suggest that there are pharmacolo-
`gical differences between CL/RAMP2 and CL/RAMP3
`receptors (Figure 6). AM22 – 52 was a significantly more
`effective antagonist of AM at
`the CL/RAMP2 receptor
`compared to the CL/RAMP3 receptor, regardless of species
`
`-13
`
`-12
`
`-9
`-10
`log [AM]
`
`rCL/rRAMP3
`
`AM
`+ 1µM CGRP8-37
`+ 10µM CGRP8-37
`
`100
`
`75
`
`50
`
`25
`
`0
`
`-13
`
`-12
`
`-11
`
`-8
`-9
`-10
`Log [AM]
`
`-7
`
`-6
`
`-5
`
`cAMP production (% ofmaximum)
`
`f
`
`cAMP production (% ofmaximum)
`
`-13
`
`-12
`
`-11
`
`-8
`-9
`-10
`Log [AM]
`
`-7
`
`-6
`
`-5
`
`rCL/rRAMP3
`
`AM
`+ 1µM AM22-52
`+ 10µM AM22-52
`
`100
`
`75
`
`50
`
`25
`
`0
`
`-13
`
`-12
`
`-11
`
`-8
`-9
`-10
`Log [AM]
`
`-7
`
`-6
`
`-5
`
`cAMP production (% of maximum)
`
`e
`
`cAMP production (% of maximum)
`
`the stimulation of cyclic AMP
`Figure 4 Characterisation of
`production by rat AM in Cos 7 cells transfected with CL/RAMP3
`combinations. Points are the mean7s.e.m. of triplicate determina-
`tions. Concentration – response curves are representative of three to
`six experiments. Data are
`expressed as
`the percentage of
`maximum cyclic AMP production, estimated by fitting each line
`to a logistic Hill equation, as described in Methods. Maximum cyclic
`AMP values were 450760 pmol per 106 cells for hCL/hRAMP3,
`300725 pmol per 106 cells for hCL/rRAMP3 and 630770 pmol per
`106 cells for rCL/rRAMP3; basal values were all below 10 pmol per
`106 cells, (a) hCL/hRAMP3, AM22 – 52; (b) hCL/hRAMP3 CGRP8 – 37;
`(c) rCL/hRAMP3, AM22 – 52; (d) rCL/hRAMP3, CGRP8 – 37; (e) rCL/
`rRAMP3, AM22 – 52; (f) rCL/rRAMP3, CGRP8 – 37.
`
`1.4570.45), giving an apparent pA2 of 6.4870.20 (n¼ 3,
`Table 1, Figure 4d). AM22 – 52 up to 10 mm failed to cause any
`significant shift (Table 1, Figure 4c). In common with the rCL/
`hRAMP2-transfected cells, BIBN4096BS did not produce any
`significant effect on AM-stimulated cyclic AMP responses
`(Table 1).
`
`Table 2 Peptide potency values for the rCL/hRAMP3 and rCL/rRAMP3 complex
`
`rCL/hRAMP3
`pEC507s.e.m. (n)
`Relative potency
`
`rCL/rRAMP3
`pEC507s.e.m. (n)
`Relative potency
`
`rAM
`aCGRP
`bCGRP
`[Cys(Et)2,7]aCGRP
`rAmylin
`
`8.7870.03 (12)
`6.3870.06 (5)
`7.9470.1 (5)
`6.3770.1 (3)
`6.470.05 (3)
`
`1
`0.0039
`0.14
`0.0038
`0.0041
`
`8.5670.12 (4)
`7.3770.06 (3)
`8.1670.35 (5)
`
`1
`0.064
`0.40
`
`[Cys(ACM)2,7] aCGRP, rat calcitonin, salmon calcitonin, AM22 – 52. CGRP8 – 37 and BIBN4096BS, all had no effect on cyclic AMP
`accumulation at concentrations up to 1 mm on the rCL/hRAMP3 receptor.
`
`British Journal of Pharmacology vol 140 (3)
`
`6
`
`

`

`D.L. Hay et al
`
`Heterogeneity of adrenomedullin receptors
`
`483
`
`composition. CGRP8 – 37 was less discriminating, but in the
`mixed rat/human receptor it had the opposite selectivity to
`AM22 – 52. As a consequence of these effects, AM22 – 52 was
`numerically more potent than CGRP8 – 37 at all CL/RAMP2
`receptors
`(although the difference was only statistically
`significant for the mixed rat/human receptor), whereas the
`opposite was
`true for
`the CL/RAMP3 receptors
`(with
`statistically significant differences between the antagonists at
`the all-rat and rat/human receptors). BIBN4096BS showed no
`measurable antagonist activity at any of the CL/RAMP2 or
`CL/RAMP3 combinations, confirming its
`status as an
`extremely selective antagonist at CL/RAMP1 receptors (see
`Hay et al., 2002, for a full account of the pharmacology of this
`compound at CL/RAMP1 receptors expressed on L6 and SK-
`N-MC cells). We also determined agonist potency ratios at the
`mixed species and all-rat CL/RAMP3 receptor. This revealed
`that the potency of bCGRP approached that of AM (at the all-
`rat receptor they were statistically indistinguishable). This
`finding was similar to that reported by Fraser et al. (1999) for
`the all-human CL/RAMP3 receptor.
`The data revealed marked differences depending upon the
`species composition of the CL/RAMP receptors. For CL/
`RAMP2 receptors, the AM22 – 52 and CGRP8 – 37 had signifi-
`cantly lower affinities at the mixed species receptor compared
`to the human and rat receptors. There were