- Email: [email protected]
Labelling of CRF1 and CRF2 receptors using the novel radioligand, [3H]-urocortin
Neurophannacology, Vol. 36, No. 10, pp. 1439-1446, 1997 0 1997 Elsevier Science Ltd. All rights reserved Printed in beat Britain 0028-3908/97 $17.00 + 0.00
Pergarnon PJk SOO28-3908(!V)OOO!M-1
Labelling of CRFl and CRF2 Receptors Using the Novel Radioligand, [3H]-urocortin J. GOTI’OWIK, V. GOETSCHY,
S. HENRIOT, E. KITAS, B. FLUHMAN, R. G. CLERC, J.-L. MOREAU, F. J. MONSMA and G. J. KILPATRICK”
Phanna Division, F-HofSmann-La Roche Ltd, 4070, Basel, Switzerland (Accepted 27 March 1997) Summary-The binding of the novel radioligand, [3H]-rat urocortin to homogenates of rat cerebellum and homogenates of cells stably transfected with the human CRFl, rat CRP2, and rat CR&p receptors was examined. In each case, specific reversible high affinity binding was observed (Kds between 0.18 and 0.31 nM). The density of sites was relatively low in the cerebellum (9 fmol/mg tissue) but high in the recombinant systems with expression levels of between 1.4 and 6.3 pmol/mg protein. Agents known to interact with CRF receptors potently competed for binding in each case. The pharmacological profile of binding to the recombinant receptors were consistent with data previously published using other radioligands. Thus, for the recombinant CRFl receptor, binding was inhibited with similar affinity by Urocortin, sauvagine, Urotensin 1 and CRF. The non-peptidic CRF antagonists (e.g. CP 154,526 and SC 241) also potently inhibited binding. The CRP2, and CRF2p receptor recombinant systems had a very similar pharmacological profile with a clear rank order of potency for the peptide ligands (Urocortin > Sauvagine > Urotensin 1 > CRF), whereas the non-peptide CRF receptor antagonists had no measurable affinity. The pharmacological profile of specific [3H]-urocortin binding to homogentates of rat cerebellum was consistent with specific labelling of a CRPr receptor. We conclude that [3H]-urocortin is a useful tool for the study of CRP receptors with the advantages that a filtration assay can be used, all CRF receptors can be labelled with the same ligand and the benefits associated with the low energy emittor, 3H. @ 1997 Elsevier Science Ltd. Keywords--j3H]i-urocortin,
CRF, CRH, CRFr, receptors, CRF2, receptors.
Corticotropin releasing factor (CRF) is thought to play an important role in integrating the stress response and has been implicated in thle pathophysiology of several diseases, including anxiety, depression, anorexia nervosa, Alzheimer’s disease and obesity (e.g. Chalmers er al., 1996; Owens and Nemeroff, 1993). Receptors for corticotropin releasing factor (CRF) have been divided into CRFr and CR& subtypes on the basis of molecular cloning (Chang et al., 1993; Chen et cd., 1993; Kishimoto et d., 1995; Lovenberg et al., 1995; Perrin et al., 1995; Stenzel et al., 1995; Vita et al., 1993;
Liaw et al., 1996). Two alternatively spliced variants of the CRFr and CRF2 receptors exist. The splice variants of the CRF2 receptor have been termed CRFza and CRF20. Additionally, a binding protein exists for CRF whose function is thought to be to terminate the effects of released CRF (Behan et al., 1995).
*To whom correspondence should be addressed. Telephone (41) 61 6888290 Fax (41) 61 6884484 E-mail: [email protected].
The study of CRF receptors has been hampered by the lack of convenient radioligands. [1251]-CRF (human or ovine) is the most frequently used radioligand (De Souza, 1987) but a centrifugation assay is required to measure binding to CRFr receptors and the relatively low affinity (17 nM for human CRF at the rat CRF2p receptor, Vaughan et al., 1995) is sub-optimal for it to be used as a ligand for CRF2 receptors. Recently, [‘251]-Tyr”-sauvagine has been employed to label CRFza receptors (Grigoriadis et al., 1996) but, here again, a centrifugation assay was employed. Urocortin, a new 40 amino acid peptide with 75% homology and 45% identity to CRF has recently been reported (Vaughan et al., 1995). Urocortin has a high affinity for both CRFi and CRF2 receptors but the colocalization with CRF2 receptors has led to the suggestion that urocortin is the natural ligand for CRF2 receptors. Very recently, human urocortin, which differs from the rat form in two amino acids at the N-terminus has been cloned (Donaldson et al., 1996). Because of its high affinity for both CRFr and CRF2 receptors, we considered that a radiolabelled derivative of this agent may be a useful tool to label CRF receptors.
1439
1440
J.
Gottowik ef al.
Thus we have obtained urcortin radiolabelled to a relatively high specific activity with tritium and now report on its binding characteristics to homogenates of rat cerebellum, recombinant human CRFi receptors and recombinant rat CRFh and CRFzp receptors. MATERIALS AND METHODS
Cloning of the human CRFl receptor A fragment representing the region between transmembrane domains 3 and 7 of the human CRFl receptor was generated by nested PCR using degenerate primers (upstream primer 1: S-(G/C)CAACT(A/T)(C/T)(A/G/ T)(A/G/T)CTGG(C/A)T(T/G)CT(G/C)-3’; upstream primer 2: S-(G/C)GA(C/T) GG(G/C)(G/C)TCTA(T/ C)CTTCACA-3’; downstream primer 1: 5’-ACCTC(A/ T/G)(C/T)(C/G)(GIA)TTG(A/C)(G/A)GAA(G’A)CAGTA3’; downstream primer 2; S-AC(C/G)A(A/C)(C/G/ A)A(G/A)(A/T)CCCTGGAA-3’), and a human cerebellum cDNA library in lambda GTll (Clontech, San Diego, CA, U.S.A.) as template DNA. This fragment, labeled with 32P by the random primers method, was used to screen the same library by hybridization. Two overlapping clones were identified which represented the 5’ and 3’ ends of the CRFr receptor coding sequence and included 64 and 276 bp of 5’ and 3’ non-coding sequence respectively. The cDNA inserts were amplified from the lambda clones using the Elongase PCR system (Life Technologies, Bethesda, MD, U.S.A.), and subcloned into the vector pAMP1 (Life Technologies) using the CloneAmp system (Life Technologies). Following verification of the sequence, the full coding region of the protein was reconstructed in pAMP1 using a unique, internal Stu 1 restriction site. The resulting insert was subcloned from pAMP1 into pCEP4 (InVitrogen, San Diego, CA, U.S.A.) using Not 1 and Kpn 1. The full sequence was again verified, with the final expression construct representing bases 199 to 1750 of the published hCI2Fr receptor sequence (Vita et al., 1993; GenBank accession number X72304). Cloning of the rat CRF2, and CRF~B receptors Using a 1.2 kb PCR DNA fragment of rat CRF2 receptor (generated from rat heart cDNA) as a hybridization probe, we screened a rat skeletal muscle Uni-ZAPTM XR cDNA library (Stratagene, La Jolla, CA, U.S.A.). From an initial screening of 0.5 x lo6 recombinants at high stringency we obtained three positive clones. The inserts were excised as a pBluescript phagemid from the Uni-ZAP vector. DNA sequence analysis revealed that two clones comprised the entire coding region of CRF2p including 20 and 30 bp of 5’ non-coding sequence, respectively. The third clone was truncated in 5’-coding sequences. The CRF2, receptor cDNA was constructed by exchanging the 5’-sequence of CRF20 receptor at the Sty I restriction site with a synthetic 132 bp DNA sequence (MedProbe A.S., Oslo, Norway) of CRF2,
according to the published sequence (Lovenberg et al., 1995; Gene bank accession number U 16253). The resulting CRF2, receptor fragment, which included 4 bp and 92 bp of 5’ and 3’ non-coding region, respectively, was subcloned from pBluescript (Stratagene, La Jolla, CA, U.S.A.) into the eukaryotic expression vector pCEP4 using Hind111 and Not1 restriction sites. The CRF2p receptor cDNA fragment, which included 20 bp and 92 bp of 5’ and 3’ non-coding region, respectively, was also subcloned from pBluescript into the pCEP4 (In Vitrogen) using Hind111 and Not1 restriction sites and the sequence was verified. Expression of CRF receptors The pCEP4-CRF receptor constructs were independently transfected into HEK-293 cells adapted for suspension culture (293s) using Lipofectamine (Life Technologies) according to the manufacturer’s recommendations. The cells were grown in HL medium containing 2% FCS and 100 U/ml Hygromycin (Calbiothem, San Diego, CA, U.S.A.) for selection. Hygromycin resistant clones were isolated and screened for expression of CRF receptor by [3H]-urocortin binding. For each receptor one clone was identified which expressed a high level of binding, and this clone was utilized for all subsequent studies. Preparation of membranes from HEK293 cells Cells were washed by two centrifugations (5 min, 500g) and the resulting pellet was resuspended in ice cold buffer 1 (Tris 50 mM; MgCl2 10 mM, EGTA 2 mM, pH 7.0) using a polytron homogenizer (15 set at maximal speed). This homogenate was centrifuged at 100 OOOgfor 60 min. The resulting pellet was resuspended in the same buffer to obtain a concentration corresponding to le7 cells/ml and the aliquots stored at -80°C. Preparation of rat cerebellar tissue Rats (male, RoRo, 110 g) were killed by decapitation. Cerebellar tissue was dissected, immediately frozen and stored at -80°C until use. On the day of the experiment, brain tissue was thawed and homogenized (Ultra turrax, T25 10000 rpm, 30 set) in 100 volumes of buffer 1 (4°C). Homogenates were centrifuged at 38000g for 15 min. The supernatant was discarded and the pellet resuspended by homogenisation in 100 vol of buffer 1, as above, and the centrifugation repeated. The final pellet was resuspended in 100 vol of buffer 2 (as for buffer 1 but also containing bovine serum albumin, 0.1%; bacitracin, 0.1 mM and aprotinin, 100 klU/ml). Binding assays Preliminary experiments revealed that urocortin bound avidly to plastics and glass. This was prevented by treating these materials with a siliconizing solution (2% ‘Surfasil’ in hexane; Pierce, IL, U.S.A.). The binding assays were conducted in a final volume of 300 ~1 (buffer 2) which contained cell ( - 20 ug protein
1441
Labelling of CW receptors using [3H]urocortin per assay for the hCRFt receptor and - 10 pg per assay
for the CRF2 receptors) or cerebellar tissue homogenates (final dilution 300 vol), [3H]-urocortin (synthesized by Amersham International, 104 or 145 Ci mmol-‘, routinely at 0.2 nM although for saturation analysis, 0.004-2 nM) and competing agent or buffer vehicle. For the CRF, receptor assays, non specific binding was defined by the inclusion of CP 154,526 (N-butyl-N-[2,5dimethyl-7-(2,4,6-trimel.hylphenyl)-7H-py~olo[2,3-d]pyrimidin-4-yl]-N-ethylamine, 1 PM). For the CRF2 receptor assays, non-specific binding was defined by the inclusion of cold human urocortin (300 nM). Assays were performed in duplicate or triplicate. Tubes were left for 2 hr at room temperature for the CRFi receptor assays and at 37°C for the CRF2 receptor assays. For the recombinant receptor systems the samples were then filtered through Whatman GF/B 96 well filter plates (pretreated with 0.3% polyethyleneimine) using a Packard filtermate 196 cell harvester and washed three times with 1 ml of cold buffer 2. 50 ~1 of Microscint 40 scintillation fluid was added to each filter well and the wells sealed. The filter plates were then agitated for at least 120 min after which time, the radioactivity retained on the filters was measured by liquid scintillation spectroscopy in a Packard Topcount scintillation counter. For the cerebellar tissue assays 3 ml of buffer 1 (plus BSA, O.l%, 4°C) was added to the tubes and this was filtered through Millipore Durapore hydrophilic filters, 0.45 pm over Millipore 3025 manifolds and washed once with 3 ml of buffer 1 (plus BSA, O.l%, 4°C). Filters were inserted into plastic scintillation vials and 10 ml of Packard Ultima Gold scintillation cocktail added. Radioactivity was then assessed by scintillation counting. Protein was determined using the Bio-Rad protein assay (Bradford, 1976). Data analysis Association and dissociation rate constants, equilibrium saturation parameters, Ki values and Hill numbers were calculated using the computer program package KINETIC/EBDA/LIGAND (McPherson, 1985; Munson and Rodbard, 1980).
Hoffmann-La Roche); Sauvagine (Sigma); Urotensin 1 (Sigma); human/rat CRF (Neo-system); ovine CRF (Sigma); alpha-helical CRF9_41 (Sigma); CRF&33 (Peninsula); CP 154,526 (N-butyl-N-[2,5-dimethyl-7-(2,4,6trimethylphenyl)-7H-pyrrolo[2,3-d]pyrirnidin-4-yl]-Nethylamine; F Hoffmann-La Roche); SC 241 ([3-(2bromo- 4 - isopropyl-phenyl) - 5 - methyl-3H-[ 1,2,3]triazolo[4,5-d]pyrimidin-7-yl]-bis-(2-methoxy-ethyl)-amine; F Hoffmann-La Roche); Elf Sanofi compound from patent EP 0 576 350 ((R,S)-(cyclopropyl-pyridin-4-yl-methyl)[4-(2,4-dichlorophenyl)-5-methyl-thiazol-2-yl]-propylamine; F Hoffmann-La Roche); Du Pont Merck compound from patent Wo 9510506 ((2-bromo-4-isopropyl-phenyl)(4,6-dimethyl-pyrimidin-2-yl)-methylamine; F. Hoffmann-La Roche).
RESULTS
Sequence analysis of CRFzo receptor Comparison of our sequence with the published rat CRFzb receptor sequence (Lovenberg et al., 1995) revealed three amino acid differences in the CRFzp specific part in both full length clones: Hiss to Thr, Argsc to Gln, and Prosi to Ala (Fig. 1). Interestingly, the amino acids we report here are conserved in the murine sequence of Perrin et al. (1995) (Fig. 1). If these differences in sequence are due to sequence polymorphism or errors in PCR (Lovenberg approach) remains subject to speculation. Specific binding For all of the assays, specific binding represented between 60% and 70% of total binding. Whilst filter binding was relatively high, there was no measurable “specific” binding to the GF/B or Millipore Durapore filters. No measurable specific binding could be detected to control or mock-transfected HEK 293 cells. Kinetics of [3H]-urocortin binding
Compounds The following compounds were used: [leu (9) - 3H]rat urocortin, 104 or 145 CYmmol (Amersham International); human Urocortin (Donaldson et al., 1996; F
A summary of the kinetic data is presented in Table 1. All assays were at equilibrium within two hours and ligand dissociation was also virtually complete for all assays 2 hr after adding excess of the agent used to define specific binding. Analysis of the kinetic data revealed a single site in the majority of experiments, therefore the data are analysed assuming a single site. The kinetic Kd, calculated
from the association
and dissociation
amino acid no. rCRF%
MGTPGSLPSR
QLLLCLYSLL
PLLUURRPGQ
fiLQDQPLUtTL
LEQYCHRm
RNFS
rCRF+ Lovenberg
MGHPGSLPSA
QLLLCLYSLL
PLLQIJROPGR
PLCIDOPLWTL
LEIIYCHRTTT
RNFS
mCRF2,
MGTPGSLPSR
OLLLCLFSLL
PULWRQPGQ
RPODOPLWTL
LERYCHRTTI
GNFS
MGQQISLPSR
QLLLCLFSLL
PULQURQPGQ
RPIIOPPLWTL
LEQYCHRTTI
GNFS
Perrin
mCRFZ, Kishimoto
54
I
Fig. 1. Partial ammo acid sequence comparison of rat (r) CRFzp receptor with mouse (m) CFU$ receptor (Perrin et al., 1995; Kishimoto et al., 1995). The underlined ammo acids are different in the sequence reported here compared to the rat CRF20 sequence published by Lovenberg et al. (1995) (accession no. T12244).
rate
J. Gottowik et al.
1442
Table 1. Kinetic analysis of [3H]-urocortin binding. Data are the mean ) SEM of three separate experiments. Association and dissociation rate constants were calculated using the computer program KINETIC.
.
rat CRFi (cerebellum)
human CRFi (HEK 293 cells)
rat CRFa, (HEK 293 cells)
rat CRFag (HEK 293 cells)
0.009+0.001 1.4kO.3 x 10’ 6.4* 1.7 x 10.‘%I
0.017 f 0.007 6.0*2.4x 10’ 2.8 + 1.6 x 10%4
0.015 + 0.002 1.5 kO.4 x lo8 9.5k2.7 x lo-“M
0.013 kO.002 1.7kO.8 x IO8 N”27 .6+3 - .2 x 10.“M
I
$MItn&> Kinetic Kd
Table 2. Equilibrium saturation analysis of [3H]-urocortin binding. Data are the mean f. SEM of three separate experiments. Data were analysed using the computer program LIGAND Tissue rat CRFi human CRFi rat CRF2, rat CRFafl
Cerebellum HEK293 cells HEK293 cells HEK293 cells
Bmax
Kd (M) 3.1* 0.4 x 1.8kO.l x 2.3 k 0.4 x 2.2 * 1.0 x
lo-lo 10-l’ 10-i’ lo-i0
9.0* l.lfmol/mg 1.4 & 0. lpmol/mg 6.3 k 0.2pmoUmg 3.4 t_ 0.4pmoVmg
tissue protein protein protein
Table 3. Inhibition constants for a range of agents to compete for [3H]-urocortin binding. Results are the mean of 3 separate experiments. SEM values for pKi values are co.2 and for Hill coefficients co.3 (excluding the Elf Sanofi compound in rat cerebellum, where the SEM was OS).indicates not tested. All values were calculated using the computer programs EBDA and LIGAND. The Elf Sanofi compound is from patent EP 0 576 350 ((R,S)-(cyclopropyl-pyridin-4-yl-methyl)-[4-(2,4-~chlorophenyl)-5-me~yl-~azol-2-yl]-propyl~ne). The Du Pont Merck compound is from patent Wo 9510506 ((2-bromo-4-isopropyl-phenyl)-(4,6-~e~yl-py~~~n-2-yl)-me~yl~ne). rat CRFi (cerebellum)
Compound
human Urocortin CRF Sauvagine Urotensin 1 Alpha-helical CRF9_41 ovine CRF CRF6-33 CP 154,526 SC 241 Du Pont Merck compound Elf Sanofi compound
human CRFi (HEK 293 cells)
rat CRFa, (HEK 293 cells)
rat CRFas (HEK 293 cells)
PKi
Hill number
PKi
Hill number
PKi
Hill number
PKi
Hill number
9.3 8.2 8.1 8.6 7.8 <6 9.0 7.5 7.1
0.69 0.76 0.87 0.80
8.3 7.9 7.6 8.3 7.5 7.9 <6 7.9
1.01 0.98 0.72 0.59 1.21 0.98
9.2 7.3 8.5 7.9 7.7 <6 <6 <6 <6 <6 <6
0.93 0.98 1.12 0.91 0.84
9.2 7.2 8.6 8.1 7.7 <6 <6 <6 46 16 <6
0.99 0.95 0.94 0.79 0.92
0.93 0.80 0.62 0.89
;:; 7.5
constants revealed high affinity [3H]-urocortin binding to
each of the tissues. These & values were close to those obtained in equilibrium saturation experiments (see next section). Equilibrium saturation analyses Data from these studies are presented in Table 2 and Fig. 2A, B, C and D. Saturable specific binding of [3H]urocortin was observed for each tissue. In each case, there was apparently only a single, high affinity site labelled. The Kd for [3H]-urocortin binding was similar in each preparation (between 0.18 and 0.31 r&I). The density of specific binding sites in the cerebellum was relatively low (9 fmol/mg tissue) whilst it was high in the recombinant receptor systems (1.4-6.3 pmol/mg protein). Competition studies Data from these studies are presented in Table 3 and Fig. 3A, B, C and D. Binding to rat cerebellum and homogenates of cells transfected with the hCRFr receptor showed a similar pharmacological profile with CRF,
0.93 0.97 0.81 0.93
sauvagine and urotensin 1 having similar high affinities. The non-peptide CRFi receptor antagonists, including CP 154,526 and SC 241 also inhibited binding potently in both cases although CP 154,526 appeared to have a lower affinity for the human CRFi receptor. Hill coefficients were close to unity and each competing agent inhibited binding to a similar level. Binding to recombinant rat CRF2, and CRF20 receptors showed a similar pharmacological profile but distinct from the CRFi receptor preparations, thus the
Fig. 2. Equilibrium saturation analysis of [3H]-urocortin binding to homogenates of (A) rat cerebellum, cells transfected with the human CRFt receptor,
(B) HEK 293 (C) HEK 293
cells transfected with the rat CRFz, receptor, and (D) HEK cells transfected with the rat CR&p receptor. Data are from single representative experiments performed in triplicate. ( n) Total binding; (A) non-specific binding, defined using CP 154526 (1 PM) for A and B and human urocortin (300 nM) for C and D; (a) specific binding.
Labelling of CRF receptors using [3H]urocortin
(4 ,.2OE-10
I .OOE-10
g
0.1
Ii p 0.06
&ODE-11 E E &OOE-II
z
I
m 0.04
a
4.00E-11
0.06
0.02 0
Z.OOE-11
O.OOE+O
I SOE-9
1 .OOE-9
500E-10
Fro‘3
Z.OOE-9
[Ml
(b) 5.OOE-10
0.8 0.7 go.6
4.00E-10
IO.5 h zo.4 z F 0.3
B 3.OOE-10 ;
10.2
2.00E-10
0.1 0
I .OOE-1 0
o oOE+O O.OOE+O
1 .OOE-9
5.00E-10
1.50E-9
Free Y]
(cl 9.oOE-10
0.7
&oOE-10
0.6
7.00E-10
so.5
6.0oE-10 E 5.00E-IO E 3 4.00E-10 3.00E-10 2.00E-10 1 .OOE-IO
O.OOE+O O.OOE+O
2.OOE-9
I .OOE-9
3.00~9
Free [MI
B.OOE-IO
(df ,
I
4.00E-10
g 3,OOE-10
i
2 OOE-10
LOOE-10
o.oOE+O O.OOE+O
S.OOE-10 Free [M]
1 OOE-9
0.4
)
1443
J. Gottowik et al.
1444
(b)
(4
I203
20
0 L l.OOE-11
l.OOE-10
l.OOE-9 Competing
l.OOE-8 agent
l.OOE-7
l.OOE-6
1 .OOE-11
1 .OOE-10
[M]
l.OOE-9 Competing
l.OOE-8 agent
1 .OOE-7
1 .OOE-6
[M]
(d)
(c) 120
120
80
20
0 1 .OOE-11
1 .OOE-10
1 .OOE-9 Competing
l.OOE-8 agent
l.OOE-7
l.OOE-6
[M]
l.OOE-11
l.OOE-10
l.OOE-9 Competing
l.OOE-8 agent
l.OOE-7
l.OOE-8
[M]
Fig. 3. Competition for [3H]-urocortin binding to homogenates of (A) rat cerebellum, (B) HEK 293 cells transfected with the human CRFr receptor (C) HEK 293 cells transfected with the rat CRF2, receptor, and (D) HEK cells transfected with the rat CRFzo receptor. Results are the mean + SEM of three separate experiments. ( n) Human urocortin (rat urocortin in cerebellar assays); (0) CRF, (0) urotensin 1; (+) sauvagine; (0) CP 154,526; (A) ovine CRF.
for the peptides was urocortinx sauvagine > urotensin 1 > CRF. All of the non-peptide CRFt receptor antagonists were inactive (at the concentrations tested) in competing for binding to the CRF2 receptor preparations. Hill coefficients were close to unity in almost all cases. Some low Hill coefficients were observed (e.g. for urotensin 1 at the hCRFi receptor). All compounds (which were active) inhibited binding to a similar level in each assay.
rank order of potency
DISCUSSION
We present data on the specific binding of [3H]urocortin to homogenates of rat cerebellum and HEK 293 cells which have been transfected with the human CRFt receptor, rat CRF2, or CRFzp receptors. In each case, [3H]-urocortin reversibly labelled a single high affinity site. The density of sites was relatively low in the cerebellum (9 fmol/mg tissue) but high in the recombi-
Labelling of CRF receptors using [3H]urocortin nant systems with expression levels of between 1.4 and 6.3 pmol/mg protein (equivalent to between 180 000 and 900 000 receptors per cell). For the recombinant systems, the affinity of various competing agents for [3H]urocortin binding was similar to those reported previously using other radioligands (Vaughan et al., 1995; Donaldson et al., 1996; Schulz et al., 1996; Grigoriadis et al., 1996). Thus, for the: cells transfected with the CRFl receptor, binding was inhibited with similar affinity by Urocortin, sauvagine, IJrotensin 1 and CRF. The small molecule CRF antagonists (e.g. CP 154526 and SC 241) also potently inhibited binding of [3H]-urocortin. For the CRF2, and CRFzb receptor recombinant systems, there was a clear rank order ad potency for the peptide ligands (Urocortin > Sauvagine > Urotensin 1 > CRF) and the small molecule CRF antagonists had no measurable affinity. It is notable that we found the pharmacology of the CRF2, and CRFz@ receptors to be identical. This is consistent with data from Donaldson et al. (1996) who examined the rat CRF2, receptor and mouse CRFzo receptor but differs from that of De Souza et al. (1995) who report that the pharmacological profile of the rat CRFzp receptor is more similar to the CRFl receptor than the CRF2, receptor. The reasons for this discrepancy are not clear but one mighlt predict that there would be no major difference in the pharmacology of the two alternatively spliced forms of the receptor given that the sequence difference is restricted to a small part of the N terminus. For the rat cerebellar homogenates, the pharmacological profile is consistent with this tissue mainly expressing the CRFl receptor. Most notably, the nonpeptide CRFl receptor antagonist CP 154,526 potently and fully competed for [3H]-urocortin binding. In fact, this compound appeared to be approximately 10 times more potent to inhibit I_3H]-urocortin binding to the rat cerebellar receptor than to the cloned human receptor. There are also some other small differences in affinity between the rat and human assays (e.g. Urocortin). Nevertheless, taking together, (1) the observed pharmacological profile, (2) the known presence of mRNA for the CRFl receptor in the rat cerebellum, and (3) the relative paucity of mRNA for the CRF2 receptor and CRF binding protein (Chalm’ers et al., 1996; Chalmers et al., 1995) leads us to conclude that we have specifically labelled the CRFl reoeptor in this tissue. The small differences in the affinity of some compounds may reflect species differences in the receptor. [3H]-Urocortin has clear advantages as a radioligand to label CRF receptors. Fiirstly, a filtration assay can be used. This is considerably easier than the centrifugation assays usually employed with other CRF receptor ligands such as [‘251]-Tyr”-CRF (e.g. De Souza, 1987) or [lz51]Tyr”-Sauvagine (Grigoriadis et al., 1996). Secondly, the same ligand can be used to label several receptors. In fact, this ligand may be useful to label the CRF binding protein for which it also has a high affinity (Vaughan et al., 1995). Lastly, the use of the low energy emittor, 3H,
1445
means reduces problems related to safety and containment, compared with agents labelled with 1251. There are some disadvantages to using [3H]-urocortin. Firstly this agent binds avidly to plastic and glass. However, this can be minimized by treating these materials with a siliconizing agent. A high level of filter binding is also observed when using GF/B (or GF/C filters). This problem is reduced by using recombinant systems expressing high levels of receptor, because the specific signal is high and by employing a 96-well filter which has a low surface area. Alternatively, as we found for cerebellar tissue where there is only a low density of binding, a hydrophilic filter should be employed. This markedly reduces filter binding. The lack of selectivity between the CRFl and CRF2 receptors may also reduce the usefulness of this agent in tissue which express both types of receptor, although the non-peptide CRFl receptor antagonists would be useful in distinguishing the two receptors. In conclusion, [3H]-urocortin appears to be a useful, new high affinity ligand for labelling CRF receptors, with the advantages that a filtration assay can be employed and all the known receptors can be labelled with this ligand. Therefore, [3H]-urocortin represents a useful addition to the tools available to study CRF receptors. authors would like to thank Drs M. Boes, Q. Branca and U. Widmer for synthesizing the nonpeptide CRF antagonists, Sabine Hankel and Stephane Grandjean for excellent technical assistance and Dr E. Schlaeger for growing the HEK 293 cells. They would also like to thank Dr Acknowledgements-The
Andrew Sleight for critical review of the manuscript.
REFERENCES Bradford M. (1976) A rapid and sensitive method for quantitation of microgram quantities of proteins using the principles of protein-dye binding. Analytical Biochemistry 72: 248-254. Behan D. P., De Souza E. B., Lowry P. J., Potter E., Sawchenko P. and Vale W. W. (1995) Corticotropin releasing factor (CRF) binding protein: A novel regulator of CRF and related peptides. Frontiers in Endocrinology 16: 362-382. Chalmers D. T., Lovenberg T. W., Grigoriadis D. E., Behan D. P. and De Souza E. B. (1996) Corticotropin-releasing factor receptors: from molecular biology to drug discovery. Trends in Pharmacological Sciences 12: 166-172. Chalmers D. T., Lovenberg T. W. and De Souza E. B. (1995) Localization of novel corticotropin releasing factor receptor (CRP2) mRNA expression to specific subcortical nuclei in rat brain: comparison with CRFl receptor mRNA expression. Journal of Neuroscience 15: 6340-6350. Chang C. -P., Pearse II R. V., O’Connell S. and Rosenfeld M. G. (1993) Identification of a seven transmembrane helix receptor for corticotropin-releasing factor and sauvagine in mammalian brain. Neuron 11: 1187-l 195. Chen R., Lewis K. A., Perrin M. H. and Vale W. W. (1993) Expression cloning of a human corticotropin-releasing factor receptor. Proceedings of the National Academy of Sciences U.S.A. 90: 8969-8971.
J. Gottowik et al.
1446
De Souza E. B. (1987) Corticotropin-releasing factor receptors in the rat central nervous system: characterisation and regional distribution. Journal of Neuroscience 7: 88-100. De Souza E. B., Lovenberg T. W., Chalmers D. T., Grigoriadis D. E., Liaw C. W., Behan D. P. and McCarthy J. R. (1995) Heterogeneity of corticotropin releasing factor receptors: multiple targets for the treatment of CNS and inflammatory disorders. Annual Reports in Medicinal Chemistry 30: 2130.
Donaldson C. J., Sutton S. W., Perrin M. H., Corrigan A. Z., Lewis K. A., Rivier J. E., Vaughan J. M. and Vale W. W. (1996) Cloning and characterization of human urocortin. Endocrinology
137: 2167-2170.
Grigoriadis D. E., Liu X.-J., Vaughn J, Palmer S. F., True C. D., Vale W. W. and Ling N, De Souza E. B. (1996) 1251-TyroSauvagine: a novel high affinity radioligand for the pharmacological and biochemical study of human corticotropin-releasing factorz, receptors. Molecular Pharmacology 50: 679-686.
Kishimoto T., Pearse II R. V., Lin C. R. and Rosenfield M. G. (1995) A sauvagine/corticotropin-releasing factor receptor expressed in heart and skeletal muscle. Proceedings of the National Academy of Science U.S.A. 92: 1108-l 112. Liaw C. W., Lovenberg T. W., Barry G., Oltersdorf T., Grigoriadis D. E. and De Souza E. B. (1996) Cloning and characterisation of the human corticotropin-releasing factor2 receptor complementary deoxyribonucleic acid. Endocrinology
137: 72-77.
Lovenberg T. W., Liaw C. W., Grigoriadis D. E., Clevenger W., Chalmers D. T., De Souza E. B. and Oltersdorf T. (1995) Cloning and characterisation of a functionally distinct corticotropin-releasing factor receptor subtype from rat brain. Proceedings of the National Academy of Science U.S.A. 92: 836-840.
McPherson, G.A. (1985) KINETIC, EBDA, LIGAND, LOWRY. A collection of radioligand binding analysis programs. Biosoft, Cambridge, U.K.
Munson P. J. and Rodbard D. (1980) Ligand: a versatile computerised approach for characterisation of ligand-binding systems. Analytical Biochemistry 107: 220-230. Owens M. J. and Nemeroff C. B. (1993) The role of corticotropin-releasing factor in the pathophysiology of affective and anxiety disorders: laboratory and clinical studies. In Corticotropin-releasing factor (Chadwick D. J., Marsh J. and A&ill K., Eds), pp. 296-316. Wiley, Chichester, U.K. Pet-tin M., Donaldson C., Chen R., Blount A., Berggren T., Bilezikjian L., Sawchenko P. and Vale W. (1995) Identification of a second corticotropin-releasing factor receptor gene and characterisation of a cDNA expressed in heart. Proceedings of the National 92: 2969-2973.
Academy
of Science
U.S.A.
Schulz D. W., Mansbach R. S., Sprouse J., Braselton J. P., Collins J., Corman M., Dunaiskis A., Faraci S., Schmidt A. W., Seeger T., Seymour P., Tingley III F. D., Winston E. N., Chen Y. L. and Heym J. (1996) CP-154,526: A potent and selective nonpeptide antagonist of corticotropin releasing factor receptors. Proceedings of the National Academy of Science U.S.A. 93: 10477-10482.
Stenzel P., Kesterson R., Yeung W., Cone R. D., Rittenber M. B. and Stenzel-Poore M. P. (1995) Identification of a novel murine receptor for corticotropin-releasing hormone expressed in the heart, Molecular Endocrinology. 9(5), 637645.
Vaughan J., Donaldson C., Bittencourt J., Perrin M. H., Lewis K., Sutton S., Chart R., Tumbull A. V., Lovejoy D., Rivier C., Rivier J., Sawchenko J. and Vale W. (1995) Urocortin, a mammalian neuropeptide related to fish urotensin 1 and to corticotropin releasing factor. Nature 378: 287-292. Vita N., Laurent P., Lefort S., Chalon P., Lelias J.-M., Kaghad M., Le Fur G., Caput D. and Ferrara P. (1993) Primary structure and functional expression of mouse pituitary and human brain corticotrophin releasing factor receptors. FEBS Letters 335: l-5.
Pergarnon PJk SOO28-3908(!V)OOO!M-1
Labelling of CRFl and CRF2 Receptors Using the Novel Radioligand, [3H]-urocortin J. GOTI’OWIK, V. GOETSCHY,
S. HENRIOT, E. KITAS, B. FLUHMAN, R. G. CLERC, J.-L. MOREAU, F. J. MONSMA and G. J. KILPATRICK”
Phanna Division, F-HofSmann-La Roche Ltd, 4070, Basel, Switzerland (Accepted 27 March 1997) Summary-The binding of the novel radioligand, [3H]-rat urocortin to homogenates of rat cerebellum and homogenates of cells stably transfected with the human CRFl, rat CRP2, and rat CR&p receptors was examined. In each case, specific reversible high affinity binding was observed (Kds between 0.18 and 0.31 nM). The density of sites was relatively low in the cerebellum (9 fmol/mg tissue) but high in the recombinant systems with expression levels of between 1.4 and 6.3 pmol/mg protein. Agents known to interact with CRF receptors potently competed for binding in each case. The pharmacological profile of binding to the recombinant receptors were consistent with data previously published using other radioligands. Thus, for the recombinant CRFl receptor, binding was inhibited with similar affinity by Urocortin, sauvagine, Urotensin 1 and CRF. The non-peptidic CRF antagonists (e.g. CP 154,526 and SC 241) also potently inhibited binding. The CRP2, and CRF2p receptor recombinant systems had a very similar pharmacological profile with a clear rank order of potency for the peptide ligands (Urocortin > Sauvagine > Urotensin 1 > CRF), whereas the non-peptide CRF receptor antagonists had no measurable affinity. The pharmacological profile of specific [3H]-urocortin binding to homogentates of rat cerebellum was consistent with specific labelling of a CRPr receptor. We conclude that [3H]-urocortin is a useful tool for the study of CRP receptors with the advantages that a filtration assay can be used, all CRF receptors can be labelled with the same ligand and the benefits associated with the low energy emittor, 3H. @ 1997 Elsevier Science Ltd. Keywords--j3H]i-urocortin,
CRF, CRH, CRFr, receptors, CRF2, receptors.
Corticotropin releasing factor (CRF) is thought to play an important role in integrating the stress response and has been implicated in thle pathophysiology of several diseases, including anxiety, depression, anorexia nervosa, Alzheimer’s disease and obesity (e.g. Chalmers er al., 1996; Owens and Nemeroff, 1993). Receptors for corticotropin releasing factor (CRF) have been divided into CRFr and CR& subtypes on the basis of molecular cloning (Chang et al., 1993; Chen et cd., 1993; Kishimoto et d., 1995; Lovenberg et al., 1995; Perrin et al., 1995; Stenzel et al., 1995; Vita et al., 1993;
Liaw et al., 1996). Two alternatively spliced variants of the CRFr and CRF2 receptors exist. The splice variants of the CRF2 receptor have been termed CRFza and CRF20. Additionally, a binding protein exists for CRF whose function is thought to be to terminate the effects of released CRF (Behan et al., 1995).
*To whom correspondence should be addressed. Telephone (41) 61 6888290 Fax (41) 61 6884484 E-mail: [email protected].
The study of CRF receptors has been hampered by the lack of convenient radioligands. [1251]-CRF (human or ovine) is the most frequently used radioligand (De Souza, 1987) but a centrifugation assay is required to measure binding to CRFr receptors and the relatively low affinity (17 nM for human CRF at the rat CRF2p receptor, Vaughan et al., 1995) is sub-optimal for it to be used as a ligand for CRF2 receptors. Recently, [‘251]-Tyr”-sauvagine has been employed to label CRFza receptors (Grigoriadis et al., 1996) but, here again, a centrifugation assay was employed. Urocortin, a new 40 amino acid peptide with 75% homology and 45% identity to CRF has recently been reported (Vaughan et al., 1995). Urocortin has a high affinity for both CRFi and CRF2 receptors but the colocalization with CRF2 receptors has led to the suggestion that urocortin is the natural ligand for CRF2 receptors. Very recently, human urocortin, which differs from the rat form in two amino acids at the N-terminus has been cloned (Donaldson et al., 1996). Because of its high affinity for both CRFr and CRF2 receptors, we considered that a radiolabelled derivative of this agent may be a useful tool to label CRF receptors.
1439
1440
J.
Gottowik ef al.
Thus we have obtained urcortin radiolabelled to a relatively high specific activity with tritium and now report on its binding characteristics to homogenates of rat cerebellum, recombinant human CRFi receptors and recombinant rat CRFh and CRFzp receptors. MATERIALS AND METHODS
Cloning of the human CRFl receptor A fragment representing the region between transmembrane domains 3 and 7 of the human CRFl receptor was generated by nested PCR using degenerate primers (upstream primer 1: S-(G/C)CAACT(A/T)(C/T)(A/G/ T)(A/G/T)CTGG(C/A)T(T/G)CT(G/C)-3’; upstream primer 2: S-(G/C)GA(C/T) GG(G/C)(G/C)TCTA(T/ C)CTTCACA-3’; downstream primer 1: 5’-ACCTC(A/ T/G)(C/T)(C/G)(GIA)TTG(A/C)(G/A)GAA(G’A)CAGTA3’; downstream primer 2; S-AC(C/G)A(A/C)(C/G/ A)A(G/A)(A/T)CCCTGGAA-3’), and a human cerebellum cDNA library in lambda GTll (Clontech, San Diego, CA, U.S.A.) as template DNA. This fragment, labeled with 32P by the random primers method, was used to screen the same library by hybridization. Two overlapping clones were identified which represented the 5’ and 3’ ends of the CRFr receptor coding sequence and included 64 and 276 bp of 5’ and 3’ non-coding sequence respectively. The cDNA inserts were amplified from the lambda clones using the Elongase PCR system (Life Technologies, Bethesda, MD, U.S.A.), and subcloned into the vector pAMP1 (Life Technologies) using the CloneAmp system (Life Technologies). Following verification of the sequence, the full coding region of the protein was reconstructed in pAMP1 using a unique, internal Stu 1 restriction site. The resulting insert was subcloned from pAMP1 into pCEP4 (InVitrogen, San Diego, CA, U.S.A.) using Not 1 and Kpn 1. The full sequence was again verified, with the final expression construct representing bases 199 to 1750 of the published hCI2Fr receptor sequence (Vita et al., 1993; GenBank accession number X72304). Cloning of the rat CRF2, and CRF~B receptors Using a 1.2 kb PCR DNA fragment of rat CRF2 receptor (generated from rat heart cDNA) as a hybridization probe, we screened a rat skeletal muscle Uni-ZAPTM XR cDNA library (Stratagene, La Jolla, CA, U.S.A.). From an initial screening of 0.5 x lo6 recombinants at high stringency we obtained three positive clones. The inserts were excised as a pBluescript phagemid from the Uni-ZAP vector. DNA sequence analysis revealed that two clones comprised the entire coding region of CRF2p including 20 and 30 bp of 5’ non-coding sequence, respectively. The third clone was truncated in 5’-coding sequences. The CRF2, receptor cDNA was constructed by exchanging the 5’-sequence of CRF20 receptor at the Sty I restriction site with a synthetic 132 bp DNA sequence (MedProbe A.S., Oslo, Norway) of CRF2,
according to the published sequence (Lovenberg et al., 1995; Gene bank accession number U 16253). The resulting CRF2, receptor fragment, which included 4 bp and 92 bp of 5’ and 3’ non-coding region, respectively, was subcloned from pBluescript (Stratagene, La Jolla, CA, U.S.A.) into the eukaryotic expression vector pCEP4 using Hind111 and Not1 restriction sites. The CRF2p receptor cDNA fragment, which included 20 bp and 92 bp of 5’ and 3’ non-coding region, respectively, was also subcloned from pBluescript into the pCEP4 (In Vitrogen) using Hind111 and Not1 restriction sites and the sequence was verified. Expression of CRF receptors The pCEP4-CRF receptor constructs were independently transfected into HEK-293 cells adapted for suspension culture (293s) using Lipofectamine (Life Technologies) according to the manufacturer’s recommendations. The cells were grown in HL medium containing 2% FCS and 100 U/ml Hygromycin (Calbiothem, San Diego, CA, U.S.A.) for selection. Hygromycin resistant clones were isolated and screened for expression of CRF receptor by [3H]-urocortin binding. For each receptor one clone was identified which expressed a high level of binding, and this clone was utilized for all subsequent studies. Preparation of membranes from HEK293 cells Cells were washed by two centrifugations (5 min, 500g) and the resulting pellet was resuspended in ice cold buffer 1 (Tris 50 mM; MgCl2 10 mM, EGTA 2 mM, pH 7.0) using a polytron homogenizer (15 set at maximal speed). This homogenate was centrifuged at 100 OOOgfor 60 min. The resulting pellet was resuspended in the same buffer to obtain a concentration corresponding to le7 cells/ml and the aliquots stored at -80°C. Preparation of rat cerebellar tissue Rats (male, RoRo, 110 g) were killed by decapitation. Cerebellar tissue was dissected, immediately frozen and stored at -80°C until use. On the day of the experiment, brain tissue was thawed and homogenized (Ultra turrax, T25 10000 rpm, 30 set) in 100 volumes of buffer 1 (4°C). Homogenates were centrifuged at 38000g for 15 min. The supernatant was discarded and the pellet resuspended by homogenisation in 100 vol of buffer 1, as above, and the centrifugation repeated. The final pellet was resuspended in 100 vol of buffer 2 (as for buffer 1 but also containing bovine serum albumin, 0.1%; bacitracin, 0.1 mM and aprotinin, 100 klU/ml). Binding assays Preliminary experiments revealed that urocortin bound avidly to plastics and glass. This was prevented by treating these materials with a siliconizing solution (2% ‘Surfasil’ in hexane; Pierce, IL, U.S.A.). The binding assays were conducted in a final volume of 300 ~1 (buffer 2) which contained cell ( - 20 ug protein
1441
Labelling of CW receptors using [3H]urocortin per assay for the hCRFt receptor and - 10 pg per assay
for the CRF2 receptors) or cerebellar tissue homogenates (final dilution 300 vol), [3H]-urocortin (synthesized by Amersham International, 104 or 145 Ci mmol-‘, routinely at 0.2 nM although for saturation analysis, 0.004-2 nM) and competing agent or buffer vehicle. For the CRF, receptor assays, non specific binding was defined by the inclusion of CP 154,526 (N-butyl-N-[2,5dimethyl-7-(2,4,6-trimel.hylphenyl)-7H-py~olo[2,3-d]pyrimidin-4-yl]-N-ethylamine, 1 PM). For the CRF2 receptor assays, non-specific binding was defined by the inclusion of cold human urocortin (300 nM). Assays were performed in duplicate or triplicate. Tubes were left for 2 hr at room temperature for the CRFi receptor assays and at 37°C for the CRF2 receptor assays. For the recombinant receptor systems the samples were then filtered through Whatman GF/B 96 well filter plates (pretreated with 0.3% polyethyleneimine) using a Packard filtermate 196 cell harvester and washed three times with 1 ml of cold buffer 2. 50 ~1 of Microscint 40 scintillation fluid was added to each filter well and the wells sealed. The filter plates were then agitated for at least 120 min after which time, the radioactivity retained on the filters was measured by liquid scintillation spectroscopy in a Packard Topcount scintillation counter. For the cerebellar tissue assays 3 ml of buffer 1 (plus BSA, O.l%, 4°C) was added to the tubes and this was filtered through Millipore Durapore hydrophilic filters, 0.45 pm over Millipore 3025 manifolds and washed once with 3 ml of buffer 1 (plus BSA, O.l%, 4°C). Filters were inserted into plastic scintillation vials and 10 ml of Packard Ultima Gold scintillation cocktail added. Radioactivity was then assessed by scintillation counting. Protein was determined using the Bio-Rad protein assay (Bradford, 1976). Data analysis Association and dissociation rate constants, equilibrium saturation parameters, Ki values and Hill numbers were calculated using the computer program package KINETIC/EBDA/LIGAND (McPherson, 1985; Munson and Rodbard, 1980).
Hoffmann-La Roche); Sauvagine (Sigma); Urotensin 1 (Sigma); human/rat CRF (Neo-system); ovine CRF (Sigma); alpha-helical CRF9_41 (Sigma); CRF&33 (Peninsula); CP 154,526 (N-butyl-N-[2,5-dimethyl-7-(2,4,6trimethylphenyl)-7H-pyrrolo[2,3-d]pyrirnidin-4-yl]-Nethylamine; F Hoffmann-La Roche); SC 241 ([3-(2bromo- 4 - isopropyl-phenyl) - 5 - methyl-3H-[ 1,2,3]triazolo[4,5-d]pyrimidin-7-yl]-bis-(2-methoxy-ethyl)-amine; F Hoffmann-La Roche); Elf Sanofi compound from patent EP 0 576 350 ((R,S)-(cyclopropyl-pyridin-4-yl-methyl)[4-(2,4-dichlorophenyl)-5-methyl-thiazol-2-yl]-propylamine; F Hoffmann-La Roche); Du Pont Merck compound from patent Wo 9510506 ((2-bromo-4-isopropyl-phenyl)(4,6-dimethyl-pyrimidin-2-yl)-methylamine; F. Hoffmann-La Roche).
RESULTS
Sequence analysis of CRFzo receptor Comparison of our sequence with the published rat CRFzb receptor sequence (Lovenberg et al., 1995) revealed three amino acid differences in the CRFzp specific part in both full length clones: Hiss to Thr, Argsc to Gln, and Prosi to Ala (Fig. 1). Interestingly, the amino acids we report here are conserved in the murine sequence of Perrin et al. (1995) (Fig. 1). If these differences in sequence are due to sequence polymorphism or errors in PCR (Lovenberg approach) remains subject to speculation. Specific binding For all of the assays, specific binding represented between 60% and 70% of total binding. Whilst filter binding was relatively high, there was no measurable “specific” binding to the GF/B or Millipore Durapore filters. No measurable specific binding could be detected to control or mock-transfected HEK 293 cells. Kinetics of [3H]-urocortin binding
Compounds The following compounds were used: [leu (9) - 3H]rat urocortin, 104 or 145 CYmmol (Amersham International); human Urocortin (Donaldson et al., 1996; F
A summary of the kinetic data is presented in Table 1. All assays were at equilibrium within two hours and ligand dissociation was also virtually complete for all assays 2 hr after adding excess of the agent used to define specific binding. Analysis of the kinetic data revealed a single site in the majority of experiments, therefore the data are analysed assuming a single site. The kinetic Kd, calculated
from the association
and dissociation
amino acid no. rCRF%
MGTPGSLPSR
QLLLCLYSLL
PLLUURRPGQ
fiLQDQPLUtTL
LEQYCHRm
RNFS
rCRF+ Lovenberg
MGHPGSLPSA
QLLLCLYSLL
PLLQIJROPGR
PLCIDOPLWTL
LEIIYCHRTTT
RNFS
mCRF2,
MGTPGSLPSR
OLLLCLFSLL
PULWRQPGQ
RPODOPLWTL
LERYCHRTTI
GNFS
MGQQISLPSR
QLLLCLFSLL
PULQURQPGQ
RPIIOPPLWTL
LEQYCHRTTI
GNFS
Perrin
mCRFZ, Kishimoto
54
I
Fig. 1. Partial ammo acid sequence comparison of rat (r) CRFzp receptor with mouse (m) CFU$ receptor (Perrin et al., 1995; Kishimoto et al., 1995). The underlined ammo acids are different in the sequence reported here compared to the rat CRF20 sequence published by Lovenberg et al. (1995) (accession no. T12244).
rate
J. Gottowik et al.
1442
Table 1. Kinetic analysis of [3H]-urocortin binding. Data are the mean ) SEM of three separate experiments. Association and dissociation rate constants were calculated using the computer program KINETIC.
.
rat CRFi (cerebellum)
human CRFi (HEK 293 cells)
rat CRFa, (HEK 293 cells)
rat CRFag (HEK 293 cells)
0.009+0.001 1.4kO.3 x 10’ 6.4* 1.7 x 10.‘%I
0.017 f 0.007 6.0*2.4x 10’ 2.8 + 1.6 x 10%4
0.015 + 0.002 1.5 kO.4 x lo8 9.5k2.7 x lo-“M
0.013 kO.002 1.7kO.8 x IO8 N”27 .6+3 - .2 x 10.“M
I
$MItn&> Kinetic Kd
Table 2. Equilibrium saturation analysis of [3H]-urocortin binding. Data are the mean f. SEM of three separate experiments. Data were analysed using the computer program LIGAND Tissue rat CRFi human CRFi rat CRF2, rat CRFafl
Cerebellum HEK293 cells HEK293 cells HEK293 cells
Bmax
Kd (M) 3.1* 0.4 x 1.8kO.l x 2.3 k 0.4 x 2.2 * 1.0 x
lo-lo 10-l’ 10-i’ lo-i0
9.0* l.lfmol/mg 1.4 & 0. lpmol/mg 6.3 k 0.2pmoUmg 3.4 t_ 0.4pmoVmg
tissue protein protein protein
Table 3. Inhibition constants for a range of agents to compete for [3H]-urocortin binding. Results are the mean of 3 separate experiments. SEM values for pKi values are co.2 and for Hill coefficients co.3 (excluding the Elf Sanofi compound in rat cerebellum, where the SEM was OS).indicates not tested. All values were calculated using the computer programs EBDA and LIGAND. The Elf Sanofi compound is from patent EP 0 576 350 ((R,S)-(cyclopropyl-pyridin-4-yl-methyl)-[4-(2,4-~chlorophenyl)-5-me~yl-~azol-2-yl]-propyl~ne). The Du Pont Merck compound is from patent Wo 9510506 ((2-bromo-4-isopropyl-phenyl)-(4,6-~e~yl-py~~~n-2-yl)-me~yl~ne). rat CRFi (cerebellum)
Compound
human Urocortin CRF Sauvagine Urotensin 1 Alpha-helical CRF9_41 ovine CRF CRF6-33 CP 154,526 SC 241 Du Pont Merck compound Elf Sanofi compound
human CRFi (HEK 293 cells)
rat CRFa, (HEK 293 cells)
rat CRFas (HEK 293 cells)
PKi
Hill number
PKi
Hill number
PKi
Hill number
PKi
Hill number
9.3 8.2 8.1 8.6 7.8 <6 9.0 7.5 7.1
0.69 0.76 0.87 0.80
8.3 7.9 7.6 8.3 7.5 7.9 <6 7.9
1.01 0.98 0.72 0.59 1.21 0.98
9.2 7.3 8.5 7.9 7.7 <6 <6 <6 <6 <6 <6
0.93 0.98 1.12 0.91 0.84
9.2 7.2 8.6 8.1 7.7 <6 <6 <6 46 16 <6
0.99 0.95 0.94 0.79 0.92
0.93 0.80 0.62 0.89
;:; 7.5
constants revealed high affinity [3H]-urocortin binding to
each of the tissues. These & values were close to those obtained in equilibrium saturation experiments (see next section). Equilibrium saturation analyses Data from these studies are presented in Table 2 and Fig. 2A, B, C and D. Saturable specific binding of [3H]urocortin was observed for each tissue. In each case, there was apparently only a single, high affinity site labelled. The Kd for [3H]-urocortin binding was similar in each preparation (between 0.18 and 0.31 r&I). The density of specific binding sites in the cerebellum was relatively low (9 fmol/mg tissue) whilst it was high in the recombinant receptor systems (1.4-6.3 pmol/mg protein). Competition studies Data from these studies are presented in Table 3 and Fig. 3A, B, C and D. Binding to rat cerebellum and homogenates of cells transfected with the hCRFr receptor showed a similar pharmacological profile with CRF,
0.93 0.97 0.81 0.93
sauvagine and urotensin 1 having similar high affinities. The non-peptide CRFi receptor antagonists, including CP 154,526 and SC 241 also inhibited binding potently in both cases although CP 154,526 appeared to have a lower affinity for the human CRFi receptor. Hill coefficients were close to unity and each competing agent inhibited binding to a similar level. Binding to recombinant rat CRF2, and CRF20 receptors showed a similar pharmacological profile but distinct from the CRFi receptor preparations, thus the
Fig. 2. Equilibrium saturation analysis of [3H]-urocortin binding to homogenates of (A) rat cerebellum, cells transfected with the human CRFt receptor,
(B) HEK 293 (C) HEK 293
cells transfected with the rat CRFz, receptor, and (D) HEK cells transfected with the rat CR&p receptor. Data are from single representative experiments performed in triplicate. ( n) Total binding; (A) non-specific binding, defined using CP 154526 (1 PM) for A and B and human urocortin (300 nM) for C and D; (a) specific binding.
Labelling of CRF receptors using [3H]urocortin
(4 ,.2OE-10
I .OOE-10
g
0.1
Ii p 0.06
&ODE-11 E E &OOE-II
z
I
m 0.04
a
4.00E-11
0.06
0.02 0
Z.OOE-11
O.OOE+O
I SOE-9
1 .OOE-9
500E-10
Fro‘3
Z.OOE-9
[Ml
(b) 5.OOE-10
0.8 0.7 go.6
4.00E-10
IO.5 h zo.4 z F 0.3
B 3.OOE-10 ;
10.2
2.00E-10
0.1 0
I .OOE-1 0
o oOE+O O.OOE+O
1 .OOE-9
5.00E-10
1.50E-9
Free Y]
(cl 9.oOE-10
0.7
&oOE-10
0.6
7.00E-10
so.5
6.0oE-10 E 5.00E-IO E 3 4.00E-10 3.00E-10 2.00E-10 1 .OOE-IO
O.OOE+O O.OOE+O
2.OOE-9
I .OOE-9
3.00~9
Free [MI
B.OOE-IO
(df ,
I
4.00E-10
g 3,OOE-10
i
2 OOE-10
LOOE-10
o.oOE+O O.OOE+O
S.OOE-10 Free [M]
1 OOE-9
0.4
)
1443
J. Gottowik et al.
1444
(b)
(4
I203
20
0 L l.OOE-11
l.OOE-10
l.OOE-9 Competing
l.OOE-8 agent
l.OOE-7
l.OOE-6
1 .OOE-11
1 .OOE-10
[M]
l.OOE-9 Competing
l.OOE-8 agent
1 .OOE-7
1 .OOE-6
[M]
(d)
(c) 120
120
80
20
0 1 .OOE-11
1 .OOE-10
1 .OOE-9 Competing
l.OOE-8 agent
l.OOE-7
l.OOE-6
[M]
l.OOE-11
l.OOE-10
l.OOE-9 Competing
l.OOE-8 agent
l.OOE-7
l.OOE-8
[M]
Fig. 3. Competition for [3H]-urocortin binding to homogenates of (A) rat cerebellum, (B) HEK 293 cells transfected with the human CRFr receptor (C) HEK 293 cells transfected with the rat CRF2, receptor, and (D) HEK cells transfected with the rat CRFzo receptor. Results are the mean + SEM of three separate experiments. ( n) Human urocortin (rat urocortin in cerebellar assays); (0) CRF, (0) urotensin 1; (+) sauvagine; (0) CP 154,526; (A) ovine CRF.
for the peptides was urocortinx sauvagine > urotensin 1 > CRF. All of the non-peptide CRFt receptor antagonists were inactive (at the concentrations tested) in competing for binding to the CRF2 receptor preparations. Hill coefficients were close to unity in almost all cases. Some low Hill coefficients were observed (e.g. for urotensin 1 at the hCRFi receptor). All compounds (which were active) inhibited binding to a similar level in each assay.
rank order of potency
DISCUSSION
We present data on the specific binding of [3H]urocortin to homogenates of rat cerebellum and HEK 293 cells which have been transfected with the human CRFt receptor, rat CRF2, or CRFzp receptors. In each case, [3H]-urocortin reversibly labelled a single high affinity site. The density of sites was relatively low in the cerebellum (9 fmol/mg tissue) but high in the recombi-
Labelling of CRF receptors using [3H]urocortin nant systems with expression levels of between 1.4 and 6.3 pmol/mg protein (equivalent to between 180 000 and 900 000 receptors per cell). For the recombinant systems, the affinity of various competing agents for [3H]urocortin binding was similar to those reported previously using other radioligands (Vaughan et al., 1995; Donaldson et al., 1996; Schulz et al., 1996; Grigoriadis et al., 1996). Thus, for the: cells transfected with the CRFl receptor, binding was inhibited with similar affinity by Urocortin, sauvagine, IJrotensin 1 and CRF. The small molecule CRF antagonists (e.g. CP 154526 and SC 241) also potently inhibited binding of [3H]-urocortin. For the CRF2, and CRFzb receptor recombinant systems, there was a clear rank order ad potency for the peptide ligands (Urocortin > Sauvagine > Urotensin 1 > CRF) and the small molecule CRF antagonists had no measurable affinity. It is notable that we found the pharmacology of the CRF2, and CRFz@ receptors to be identical. This is consistent with data from Donaldson et al. (1996) who examined the rat CRF2, receptor and mouse CRFzo receptor but differs from that of De Souza et al. (1995) who report that the pharmacological profile of the rat CRFzp receptor is more similar to the CRFl receptor than the CRF2, receptor. The reasons for this discrepancy are not clear but one mighlt predict that there would be no major difference in the pharmacology of the two alternatively spliced forms of the receptor given that the sequence difference is restricted to a small part of the N terminus. For the rat cerebellar homogenates, the pharmacological profile is consistent with this tissue mainly expressing the CRFl receptor. Most notably, the nonpeptide CRFl receptor antagonist CP 154,526 potently and fully competed for [3H]-urocortin binding. In fact, this compound appeared to be approximately 10 times more potent to inhibit I_3H]-urocortin binding to the rat cerebellar receptor than to the cloned human receptor. There are also some other small differences in affinity between the rat and human assays (e.g. Urocortin). Nevertheless, taking together, (1) the observed pharmacological profile, (2) the known presence of mRNA for the CRFl receptor in the rat cerebellum, and (3) the relative paucity of mRNA for the CRF2 receptor and CRF binding protein (Chalm’ers et al., 1996; Chalmers et al., 1995) leads us to conclude that we have specifically labelled the CRFl reoeptor in this tissue. The small differences in the affinity of some compounds may reflect species differences in the receptor. [3H]-Urocortin has clear advantages as a radioligand to label CRF receptors. Fiirstly, a filtration assay can be used. This is considerably easier than the centrifugation assays usually employed with other CRF receptor ligands such as [‘251]-Tyr”-CRF (e.g. De Souza, 1987) or [lz51]Tyr”-Sauvagine (Grigoriadis et al., 1996). Secondly, the same ligand can be used to label several receptors. In fact, this ligand may be useful to label the CRF binding protein for which it also has a high affinity (Vaughan et al., 1995). Lastly, the use of the low energy emittor, 3H,
1445
means reduces problems related to safety and containment, compared with agents labelled with 1251. There are some disadvantages to using [3H]-urocortin. Firstly this agent binds avidly to plastic and glass. However, this can be minimized by treating these materials with a siliconizing agent. A high level of filter binding is also observed when using GF/B (or GF/C filters). This problem is reduced by using recombinant systems expressing high levels of receptor, because the specific signal is high and by employing a 96-well filter which has a low surface area. Alternatively, as we found for cerebellar tissue where there is only a low density of binding, a hydrophilic filter should be employed. This markedly reduces filter binding. The lack of selectivity between the CRFl and CRF2 receptors may also reduce the usefulness of this agent in tissue which express both types of receptor, although the non-peptide CRFl receptor antagonists would be useful in distinguishing the two receptors. In conclusion, [3H]-urocortin appears to be a useful, new high affinity ligand for labelling CRF receptors, with the advantages that a filtration assay can be employed and all the known receptors can be labelled with this ligand. Therefore, [3H]-urocortin represents a useful addition to the tools available to study CRF receptors. authors would like to thank Drs M. Boes, Q. Branca and U. Widmer for synthesizing the nonpeptide CRF antagonists, Sabine Hankel and Stephane Grandjean for excellent technical assistance and Dr E. Schlaeger for growing the HEK 293 cells. They would also like to thank Dr Acknowledgements-The
Andrew Sleight for critical review of the manuscript.
REFERENCES Bradford M. (1976) A rapid and sensitive method for quantitation of microgram quantities of proteins using the principles of protein-dye binding. Analytical Biochemistry 72: 248-254. Behan D. P., De Souza E. B., Lowry P. J., Potter E., Sawchenko P. and Vale W. W. (1995) Corticotropin releasing factor (CRF) binding protein: A novel regulator of CRF and related peptides. Frontiers in Endocrinology 16: 362-382. Chalmers D. T., Lovenberg T. W., Grigoriadis D. E., Behan D. P. and De Souza E. B. (1996) Corticotropin-releasing factor receptors: from molecular biology to drug discovery. Trends in Pharmacological Sciences 12: 166-172. Chalmers D. T., Lovenberg T. W. and De Souza E. B. (1995) Localization of novel corticotropin releasing factor receptor (CRP2) mRNA expression to specific subcortical nuclei in rat brain: comparison with CRFl receptor mRNA expression. Journal of Neuroscience 15: 6340-6350. Chang C. -P., Pearse II R. V., O’Connell S. and Rosenfeld M. G. (1993) Identification of a seven transmembrane helix receptor for corticotropin-releasing factor and sauvagine in mammalian brain. Neuron 11: 1187-l 195. Chen R., Lewis K. A., Perrin M. H. and Vale W. W. (1993) Expression cloning of a human corticotropin-releasing factor receptor. Proceedings of the National Academy of Sciences U.S.A. 90: 8969-8971.
J. Gottowik et al.
1446
De Souza E. B. (1987) Corticotropin-releasing factor receptors in the rat central nervous system: characterisation and regional distribution. Journal of Neuroscience 7: 88-100. De Souza E. B., Lovenberg T. W., Chalmers D. T., Grigoriadis D. E., Liaw C. W., Behan D. P. and McCarthy J. R. (1995) Heterogeneity of corticotropin releasing factor receptors: multiple targets for the treatment of CNS and inflammatory disorders. Annual Reports in Medicinal Chemistry 30: 2130.
Donaldson C. J., Sutton S. W., Perrin M. H., Corrigan A. Z., Lewis K. A., Rivier J. E., Vaughan J. M. and Vale W. W. (1996) Cloning and characterization of human urocortin. Endocrinology
137: 2167-2170.
Grigoriadis D. E., Liu X.-J., Vaughn J, Palmer S. F., True C. D., Vale W. W. and Ling N, De Souza E. B. (1996) 1251-TyroSauvagine: a novel high affinity radioligand for the pharmacological and biochemical study of human corticotropin-releasing factorz, receptors. Molecular Pharmacology 50: 679-686.
Kishimoto T., Pearse II R. V., Lin C. R. and Rosenfield M. G. (1995) A sauvagine/corticotropin-releasing factor receptor expressed in heart and skeletal muscle. Proceedings of the National Academy of Science U.S.A. 92: 1108-l 112. Liaw C. W., Lovenberg T. W., Barry G., Oltersdorf T., Grigoriadis D. E. and De Souza E. B. (1996) Cloning and characterisation of the human corticotropin-releasing factor2 receptor complementary deoxyribonucleic acid. Endocrinology
137: 72-77.
Lovenberg T. W., Liaw C. W., Grigoriadis D. E., Clevenger W., Chalmers D. T., De Souza E. B. and Oltersdorf T. (1995) Cloning and characterisation of a functionally distinct corticotropin-releasing factor receptor subtype from rat brain. Proceedings of the National Academy of Science U.S.A. 92: 836-840.
McPherson, G.A. (1985) KINETIC, EBDA, LIGAND, LOWRY. A collection of radioligand binding analysis programs. Biosoft, Cambridge, U.K.
Munson P. J. and Rodbard D. (1980) Ligand: a versatile computerised approach for characterisation of ligand-binding systems. Analytical Biochemistry 107: 220-230. Owens M. J. and Nemeroff C. B. (1993) The role of corticotropin-releasing factor in the pathophysiology of affective and anxiety disorders: laboratory and clinical studies. In Corticotropin-releasing factor (Chadwick D. J., Marsh J. and A&ill K., Eds), pp. 296-316. Wiley, Chichester, U.K. Pet-tin M., Donaldson C., Chen R., Blount A., Berggren T., Bilezikjian L., Sawchenko P. and Vale W. (1995) Identification of a second corticotropin-releasing factor receptor gene and characterisation of a cDNA expressed in heart. Proceedings of the National 92: 2969-2973.
Academy
of Science
U.S.A.
Schulz D. W., Mansbach R. S., Sprouse J., Braselton J. P., Collins J., Corman M., Dunaiskis A., Faraci S., Schmidt A. W., Seeger T., Seymour P., Tingley III F. D., Winston E. N., Chen Y. L. and Heym J. (1996) CP-154,526: A potent and selective nonpeptide antagonist of corticotropin releasing factor receptors. Proceedings of the National Academy of Science U.S.A. 93: 10477-10482.
Stenzel P., Kesterson R., Yeung W., Cone R. D., Rittenber M. B. and Stenzel-Poore M. P. (1995) Identification of a novel murine receptor for corticotropin-releasing hormone expressed in the heart, Molecular Endocrinology. 9(5), 637645.
Vaughan J., Donaldson C., Bittencourt J., Perrin M. H., Lewis K., Sutton S., Chart R., Tumbull A. V., Lovejoy D., Rivier C., Rivier J., Sawchenko J. and Vale W. (1995) Urocortin, a mammalian neuropeptide related to fish urotensin 1 and to corticotropin releasing factor. Nature 378: 287-292. Vita N., Laurent P., Lefort S., Chalon P., Lelias J.-M., Kaghad M., Le Fur G., Caput D. and Ferrara P. (1993) Primary structure and functional expression of mouse pituitary and human brain corticotrophin releasing factor receptors. FEBS Letters 335: l-5.