Eur. J. Biochem. 189, 625-635 (1990) FEBS 1990

Characterisation and visualisation of [ 3H]dermorphin binding to p opioid receptors in the rat brain Combined high selectivity and affinity in a natural peptide agonist for the morphine (p) receptor

Mohamed AMICHE’, Sandrine SAGAN I, Amram MOR’, Didier PELAPRAT’, William ROSTENE’, Antoine DELFOUR ’ and Pierre NICOLAS

’ Laboratoire de Bioactivation des Peptides, Institut Jacques Monod, Universite Paris, Paris 7, France ’ Unite 55, lnstitut National de la Santi- et de la Recherchc Mkdicale, Hopital Saint Antoine, Paris, France

(Received May 3/0ctober 2, 1989) - EJB 89 0559

Dermorphin, Tyr-DAla-Phe-Gly-Tyr-Pro-Ser-NHz, a potent opioid peptide isolated from amphibian skin, is endowed with outstanding structural and biological features. It has no common structure with mammalian opioid peptides and is a unique example of a peptide, synthesized by an animal cell, which contains a D-amino acid in its native sequence. We have undertaken a complete evaluation of the receptor selectivity of dermorphin, together with the binding characteristics and receptor distribution of [3H]dermorphin in the rat brain.

1. Dermorphin was tested for its relative affinity to p-, 6- and x-opioid receptors by determining its potency in displacing the selective preceptor ligand [3H]Tyr-~Ala-Gly-MePhe-Gly-ol (where Gly-ol = glycinol), the pro- totypic &receptor ligand [3H]Tyr-~Pen-Gly-Phe-~Pen (where DPen = j,P-dimethylcysteine) and the x ligand [3H]ethylketocyclazocine from rat brain and/or guinea pig cerebellum membrane preparations. Inhibitory constant (Ki) values of dermorphin were 0.7 nM, 62 nM and > 5000 nM respectively for p, 6 and 31 sites, indicating a selectivity ratio Ki(6)/Ki(p) = 88. Under similar conditions, Tyr-DAla-Gly-MePhe-Gly-ol, which is regarded as one of the most selective high-affinity p-agonist available, exhibited a selectivity ratio of 84.

2. Specific binding properties of tritium-labeled dermorphin (52 Ci/mmol) were characterized in the rat brain. Equilibrium measurements performed over a large range of concentrations revealed a single homogeneous population of high-affinity binding sites (Kd = 0.46 nM; B,,, = 92 fmol/mg membrane protein).

3. Profound differences were observed in the potencies displayed by various selective opiates and opioids ligands in inhibiting the specific binding of [3H]dermorphin. The rank order of potency was in good agreement with that obtained with other p-selective radiolabeled ligands.

4. Receptor autoradiography in vitro was used to visualize the distribution of [3H]dermorphin binding sites in rat brain. The labeling pattern paralleled that observed using other p probes. Binding parameters and selectivity profile of [3H]dermorphin on slide-mounted sections were similar to those obtained with membrane homogenates.

5. Finally, intracerebroventricular administration of synthetic dermorphin into mice showed that this peptide is the most potent analgesic known to date, being up to 5 and 670 times more active than a-endorphin and morphine, respectively. Higher doses induced catalepsy.

The overall data collected demonstrate that dermorphin is the first among the naturally occurring peptides to be highly potent and ncarly specific super-agonist towards the morphine (p) receptor. The high binding specificity and affinity of dermorphin together with its very low non-specific binding, its high resistance to enzymatic degradation and its ability to cross the blood brain barrier make this natural peptide very attractive for dissecting the role(s) and for identifying molecular and conformational determinants of high-affinity binding to the morphine receptor.

Correspondence to P. Nicolas, Laboratoire de Bioactivation des Peptides, Institut Jacques Monod, Tour 43, Universite Paris 7,2 Place Jussieu, F-75252 Paris Cedex 05, France

Abbreviations. Dermorphin = Tyr-DAla-Phe-Gly-Tyr-Pro-Ser- NH’; other peptides are abbreviated as cnkephalin variants according to IUPAC-IUB Recommendations, see Eur. J . Biochem. /38,32 - 33: [Leu5]enkephalin = Tyr-Gly-Gly-Phe-Leu; [Met5]enkephalin = Tyr- Gly-Gly-Phe-Met; [~Ala~,MePhe~,Gly-ol~]enkephalin = Tyr-DAla- Gly-MePhe-Gly-ol where MePhe = N-methylphenylalanine and Gly- 01 = glycinol; [~Ser’]enkephalin-Thr = Tyr-DSer-Gly-Phe-Leu-Thr ; [~Pen’,~]enkephalin = Tyr-DPen-Gly-I’he-DPen where Pen = b,p- dimethylcysteine; [oPen2,Pen5]enkephalin = Tyr-open-Gly-Phe-Pen ; [~Thr~lenkephalin-Thr = Tyr-uThr-Gly-Phe-Leu-Thr ; [DAlaZ,DLeu5]- enkephalin = Tyr-oAla-Gly-Phe-DLeu; ADs0, median antinocicep- tive dose; ICs0, 50% inhibitory concentration.

The dermatous granular glands of amphibians synthesize and expel an extraordinarily rich variety of biologically active peptides [l, 21. This has allowed the characterization of more than 30 peptides including thyrotropin-releasing hormone, caeruleins, angiotensins, xenopsin, sauvagine, bombesin, tachykinin and the antimicrobial peptides magainins [3 - 51. It was soon established that most of these amphibian peptides have counterparts, either identical or closely similar, within the neurosecretory cells of the mammalian gastrointestinal tract and/or the brain [6-81. For instance, amphibians and mammals have identical thyrotropin-releasing hormone and bradykinin, while amphibian sauvagine, xenopsin and caerulein are closely related to mammalian corticotropin-re- leasing factor, neurotensin and cholecystokinin, respectively.

626

............... .. ... ..................... .................... .............. :. ......, :i::::i: .. ..j:i: ............ ..

Since frog skin may contain up to I00000 times higher doses than mammalian tissues, amphibian peptides represent a use- ful tool in the identification of homologues in mammals and are highly appropriate for the understanding of their biosyn- thesis and physiological functions [9].

Among secretory peptides isolated from amphibian skin, the potent opioid peptide dermorphin is endowed with out- standing structural and biological features [4, 10, 111. Der- morphin, Tyr-DAla-Phe-Gly-Tyr-Pro-Ser-NH2, has no com- mon structure with the sequence of mammalian opioid peptides and is unique among peptides synthesized by animal cells in having an alanine residue in position 2 in its D-COnfigUr- ation. Recent evidence [12] has been obtained showing that dermorphin is synthesized via larger biosynthetic precursors, establishing this peptide as the first example of a D-amino- acid-containing peptide being synthesized by the ribosomal route. The predicted amino acid sequence of the pro- dermorphins (Fig. 1) contained four or five similar repeats of 35 amino acids, each of these including a copy of the dermorphin progenitor sequence, Tyr-Ala-Phe-Gly-Tyr-Pro- Ser-Gly, bracketted by typical prohormone-processing signals. In the corresponding cloned cDNAs, a normal alanine codon occurred at the position where ~ A l a is present in mature dermorphin. This observation suggests the existence of a novel co- or post-translational reaction ensuring the conversion of an L-amino acid to its D-isomer. However, the recognition of a D-alanine by a tRNA (L-Ala) cannot be excluded.

Up to 20 different mammalian opioid peptides have been isolated, all of them containing the amino acid sequence of [Met5] or [Le~~lenkephalin. All these peptides have been shown to derive from the processing of only three separate biosynthetic precursors [I3 - 151. Comparison of the primary structure of dermorphin to that of the mammalian opioid peptides revealed such drastic differences that it would be unlikely that the enkephalins, dynorphins or endorphins rep- resent the generic counterparts of dermorphin in mammals. Thus, the occurrence of dermorphin, or peptides with closely similar sequences, in mammalian tissues remains to be unam- biguously established [16- 191.

With respect to its pharmacological activities, dermorphin exhibits a profile similar to that of morphine and has been reported to outclass all hitherto known natural and synthetic opiates and opioid peptides in producing long-lasting anal- gesia after intracerebroventricular administration [lo, 20 - 221. Structure,/activity relationship studies carried out in vivo with synthetic analogs have shown the presence of the ~ A l a in position 2 to be essential for analgesic activity, whereas the minimum structural requirement for morphine-like activities is the N-terminal tetrapeptide [23 -281.

Preliminary data from competition for binding with radiolabeled opiates and opioids have shown dermorphin to be more potent in inhibiting the binding of P-endorphin, Tyr- DAla-Gly-MePhe-Gly-ol or dehydromorphine than that of Tyr-DAla-Gly-Phe-DLeu or ethylketocyclazocine to rat brain membranes [29, 301. Although these data suggested that dermorphin is more selective for p binding sites, the rather low selectivity index of most of the primary ligands used has impeded any accurate determination of the exact profile and re- lative affinity of dermorphin. In addition, it has not been pos- sible to check the contribution to other opioid or non-opioid binding sites due to the lack of radiolabeled dermorphin. These points are of particular interest since the functional relationships between the mammalian opioid peptides and the three major types of opioid receptors, termed p, 6 and x, are not yet clear. In particular, none of these mammalian peptides

H I I I

1 TUr-D net-Phe-His-Leu-net-Asp-~z I (Dermenkephalin)

............... .... .................... , ..i..i....._ >.. ............. ..? . .............. ........... .................

Processing t

I I I I

Lys-Arg-Tyr-net-Phe-His-Leu-net-AslI -6lY-GlU -Ala-LyS-LyS

t

I Processing

(Derrnorphin) Tyr-D Ala-Phe-Gly-Tyr-Pro-Ser- m12

Fig. 1. Highly schemutic representution nf the structural orgunisation of the derrnorphin biosynthetic precursors Jrom the arboreal ,frog Phyllomedusa sauvagei [12/. The N-terminal dashed sequence is the predicted signal peptide sequence. Positions of dermorphin progenitor sequence, Tyr-Ala-Phe-Gly-Tyr-Pro-Ser, which is flanked by Lys-Arg at thc amino end and by Gly-Glu-Ah-Lys-Lys at the carboxyl end, are shaded. One of the precursors (upper representation) has a dermorphin copy replaced with a distinct progenitor sequence, Tyr- Met-Phe-His-Leu-Met-Asp, whose processing yields the novel 6- opioid peptide dermenkephalin [51, 521, i.e. Tyr-DMet-Phe-His-Leu- Met- Asp-NH,

has been shown to bind selectivity to the morphine (p) opioid receptor, the subtype which is believed to mediate analgesia.

The high potency of dermorphin in inducing analgesia and the striking structural differences when compared with the other known opioid peptides, prompted us to characterize the membrane receptors through which dermorphin elicits its pharmacological activities. In this study, we have undertaken a complete evaluation of the binding parameters, receptors selectivity profile and receptor distribution pattern of [3H]dermorphin in the rat brain. Results demonstrate that dermorphin is the first of the naturally occurring peptides to be a highly potent and fully selective super-agonist towards the morphine (p) receptor.

MATERIALS AND METHODS

Chemicals

[T~r’-3,5-~H] [DAla2, MePhe4, Gly-olS]Enkephalin (47 Ci/ mol) was purchased from the Commisariat a 1’Energie Atomique (Saclay, France). [ T~r’-3,5-~H] [~Pen~-’]Enkepha- lin (51.3 Ci/mmol) and [3H]ethylketocyclazocine (27 Ci/ mmol) were purchased from Du Pont/New England Nuclear. The radiochemical purity of the tritiated ligands was checked by HPLC before used and found to be >98%. Unlabelled [~Ala’, MePhe4, Gly-o15]enkephalin, [~Pen**~]enkephalin, [Met’lenkephalin, [~Thr~lenkephalin-Thr, [~Ser~lenkephalin- Thr, dynorphin A (porcine) and [DAla2, ~Leu’1enkephalin were from Peninsula. Naloxone was from Sigma (St. Louis). Human P-endorphin was a gift of Dr. C. H. Li (San Francisco). Ethylketocyclazocine was a gift from Dr. J. Costentin (Rouen). Morphine was a generous gift from Dr. J. M. Zajac (Paris). Purity of the peptides and opiates was routinely checked by HPLC, TLC, mass spectrometry and amino acid composition as described [31].

627

95

90

80 8 70

60 3 50

4 0 -

2 30 4 20

10

- -

Tritiated dermorphin

Tritiated dermorphin ([Tyr'.5-2,5-3H]dermorphin, 52 Ci/ mmol) was prcpared (Service de Biochimie, CEA, Saclay, France) by catalytic tritiation of [2,5-i0dotyrosyl'~~]dermor- phin and purified by HPLC. The purity of the labeled peptide was assessed by analytical HPLC on a CIS reverse-phase column (Lichrosorb RP 18.5 pm) using a O-6O% linear gradient of acetonitrile in 0.1 % trifluoroacetic acid. A single peak of radioactivity co-migrating with unlabeled dermorphin was observed. Single spots could be detected after ascending thin-layer chromatography on a silica gel plate (Merck) of [3H]dermorphin in the presence or the absence of 10 pg unlabeled dermorphin, in 1-butanol/acetic acid/water (4: 1 : l), with an Rf = 0.46 corresponding to authentic dermorphin. The ultraviolet spectrum of [3H] dermorphin was similar to that of synthetic dermorphin with maximum at 275 nm in water and 278 nm in methanol as expected for tyrosyl residues. Specific radioactivity was determined by amino acid compo- sition and ultraviolet spectroscopy. Tritiated dermorphin was kept frozen under liquid nitrogen for several months without loss of its chemical and biological properties.

- -

- - - - - - -

5)

Peptide synthesis

Dermorphin was prepared by stepwise manual solid-phase synthesis on 1 g benzhydrylamine polymer (0.5 mmol NH2/ g) using the preformed symmetric anhydride technique as previously described 1311. The crude synthetic peptide was purified by a combination of gel filtration, ion-exchange chromatography and preparative reverse-phase HPLC. The purity of synthetic dermorphin was assessed by analytical reverse-phase HPLC, thin-layer chromatography and amino acid analysis as described [31]. Positive-ion fast-atomic-bom- bardment spectrometry analysis of synthetic dermorphin showed a strong (M + H)' quasi-molecular ion at m/z = 804 with very little fragmentation. This molecular mass corre- sponds to that predicted theoretically.

The in vivo agonist potency of synthetic dermorphin was assessed by the tail-flick method [32] after intracere- broventricular administration to mice. A log/log probit dose/ response curve for the antinociceptive effect produced by in- tracerebroventricular injection of graded doses of dermorphin into mice are shown in Fig. 2. Also shown in this figure are dose/response curves of morphine and human /3-endorphin. Parallelism between dose/response curves was observerd, suggesting that morphine and dermorphin could produce their pharmacological effect by acting on a similar population of opioid receptors in vivo. The times for the peak effect and the duration of analgesia produced by morphine, dermorphin and P-endorphin were identical (not shown). On a molar basis, dermorphan is about 5 times and 670 times more active than p- endorphin and morphine, respectively (AD50 = 0.0058,0.030 and 4.20 nmol/mouse, respectively for dermorphin, P- endorphine and morphine). Higher doses of dermorphin or P- endorphin induce catalepsy.

Amino acid analysis

The amino acid composition of synthetic dermorphin was determined on an hydrolysate of the peptides (200 pmol) after phenylthiocarbamoylation of the released residues by phenyl- isothiocyanate followed by HPLC analysis as described [31].

I , I l l I I # , I I I I l l I

0.005 0.01 0.02 0.05 0.1 "0.5 1 2 5 10

Dose ( nmol 1

Fig. 2. Log-probit response curves fo r antinociceptive effect produced by intracerebroventricular injection of morphine ( D), P-endorphin (A) and synthetic dermorphin (0). After control latency was obtained, groups of 10 mice were injected ( 5 PI) and the tail-flick response was determined at 30 min

Opioid binding assays

Membrane fractions from rat brain (minus cerebellum) were prepared as described [33] with some modifications. Briefly, decerebellated whole brain of male 1 SO-g Sprague Dawley rats (Charles River, France) were homogenized by four cycles of polytron and centrifugation (20 min, 12000 g ) at 4°C. The buffer used was 50 mM Tris/Cl pH 7.4. The pellet of washed membranes equivalent to 10 brains was dispersed in 100 ml 50 mM Tris/Cl pH 7.4, 20% glycerol and stored at - 80°C. The final protein concentration of this extract was 7.0 mg/ml as determined by the method of Lowry [34] using bovine serum albumin as standard. Binding assays were perfomed at 24°C in 50 mM Tris/Cl pH 7.4, plus 0.1 YO bovine serum albumin and 0.01 YO bacitracin. Each assay contained, in a final volume of 2 ml, the membrane preparation (1.4 mg membrane protein) and the tritiated ligand at the desired concentration with or without unlabeled hgand. The non- specific binding was determined in the presence of 1 pM unlabeled ligand. The tubes were incubated for the desired time period (60 min for [3H] [DAla', MePhe4, Gly- ol'lenkephalin, [3H]demorphin and [3H]ethylketocyclazo- cine; 150 min for [3H] [~Pen'*~]enkephalin) while being stirred every 15 min. The binding reaction was terminated by rapid vacuum filtration through 0.1 YO poly(ethy1enimine)-coated Whatman glass fiber filters (GF/B). Filters were washed twice with 10 ml cold 50 mM Tris/Cl pH 7.4, 0.1% bovine serum albumin and transfered to vials containing 5 ml PCS scintil- lation fluid (Amersham). Specific binding was considered as the difference in radioactivity trapped on the filters either in the absence or in the presence of 3000 nM of the unlabeled primary ligand. All determinations were performed on dupli- cates. The 50% inhibitors concentration values (ICso) were obtained from nonlinear least-squares regression to a two- parameter logistic equation of the percentage specific bindings vs log(dose) curves. The inhibitory constant (K , ) of the various unlabeled ligands was calculated from the relation Ki = ICs0/ (1 + L/K,) [35], where L is the concentration of the labeled ligand and Kd its equilibrium dissociation constant determined by saturation binding analysis. All binding experiments were performed with the same preparation of brain membranes and purified radiolabels. Degradation of tritiated ligands was

628

.10

.08

.06 ?

0' .04 a m

LL

.02

0 0 50 100

[3H1Tyr-oAl~-Gly-MePhe-Gly-ol bound (fmol/mg)

.04

.03

0 .02 :

: .01

U

U . m

0 0 50

[~H1Ethylketocyclnzocine bound(f m o l l mg )

B

Fig. 3. Scatchard analysis of [ 3 H ] [oAla2, MePhe4, Cly-o15]enkephalin ( A ) , [ 3 H ] [nPen'*']enkephalin ( B ) , [3H]ethylketoc~clazocine ( C ) and [3H]dermorphin (0) specijl'c binding to rat brain membranes (1.4 rng membrane protein) at 24°C in TrisjClpH 7.4. Non-specific binding of the 3H-labeled peptides in A, B and D was measured in the presence of 1 pM of the corresponding non-labeled ligand. [3H]Ethylketo- cyclazocine total binding was measured in the presence of 200 nM [ ~ A l a * , MePhe4, Gly-o15]enkephalin @-receptor blocker) and 200 nM [~Pen*,~]enkephalin (&receptor blocker) and the non-specific binding determined in the presence of 1 pM ethylketocyclazocinc plus both blockers. Values of Kd and B,,, were calculated from the slope and intercept estimated either by linear or non-linear least-squares regression analysis

estimated at 24'C under the experimental conditions used for the binding assay, by chromatography on a Waters C18 Sep- Pak column following a procedure similar to that of Gay and Lahti [36]. The breakdown of the radiolabels ranged from 1 YO after a 1-h incubation to 6% after a 5-h incubation.

Quantitative azrtoradiography in vitro

Male Wistar rats (200 - 250 g) were sacrificed by decapi- tation, their brains rapidly removed and frozen, 20-pm sec- tions were cut on a cryostat at - 16"C, thaw-mounted onto gelatin-coated slides and stored at - 80 "C until use. The slide- mounted sections (four sections per slide) were incubated for 90 min at 22°C with 600 p11.5 nM [3H]dermorphin in 50 mM Tris/Cl pH 7.4 containing 1 mg/ml bovine serum albumin and 0.1 mg/ml bacitracin. Non-specific binding was determined in the presence of 1000 nM unlabeled dermorphin. After incu- bation, the sections were twice washed for 15 min in ice- cold 50 mM Tris/CI pH 7.4 containing 1 mg/ml bovine serum albumin. After washing, sections were quickly dipped twice ( 5 s each) in ice-cold water, dried with cold air, then tightly juxtaposed to tritium-sensitive Ultrofilm (LKB). The films were exposed for 6 weeks, developed for 3 min in Kodak D 19b developer at 2 0 T , then dipped in water and fixed with Kodak Unifix. Adjacent sections were incubated in the pres- ence of 10 nM [DAla', MePhe4, Gly-o15]enkephalin or 100 nM

[ ~ P e n ~ ~ ~ l e n k e p h a l i n in order to measure the displacement obtained after occupation of the p or 6 opioid receptors, respectively. Semi-quantitative data were obtained from autoradiograms using a Rag 200 apparatus (Biocom, France).

Pharmacological assay

The effect of intracerebroventricularly administered opi- ates or peptides on heat-escape latency was assessed by the tail-flick method [32] using groups of mice (male Swiss Webster, 20-25 g) per dose as described [37]. Median antino- ciceptive dose (AD,,) and 95% confidence limits were calcu- lated from the logarithm of the dose vs probit ('/n analgesia) curves.

RESULTS

Saturation binding analysis and receptor selectivity of (Tyr '-3J3 H ] [ D d a 2 , MePhe4, Gly-o15/enkephalin, (T~r ' -3 ,5-~H] (~Pen~~ ' ]enkepha l in and L3 H]ethylketocycluzocine binding to rat brain opioid receptors

Saturation studies and binding profile of the selective p- ligand [DAla2, MePhe4, Gly-o15]enkephalin [38], the proto- typical probe [~Pen~*']enkephalin [39, 401 and the non-selec- tive x ligand ethylketocyclazocine [41] were evaluated for the

629

Tablc 1. Equilibrium binding parameters for r 3 H ] [ ~ A l a ’ , MePhe4, Cly-01~ Jenkephulin, [ 3 H ] [~Pen~~’]enkephul in , r 3 Hldernzorphin und [3 H]etliylketocyclazocine binding to rat bruin membranes ( I .4 mg membrane protein) at 24“CpH 7.4 Values reported are the means of 2,3 or 4 independent experiments. Values in parentheses are SEM. Binding parameters for ethylketocyclazocine were obtained in the presence of 200 nM [~Ala’ , MePhe4, Gly-o15]enkephalin (p blocker) and 200 nM [~Pen~~~]enkephal in (6 blocker); nd = not determined

3H-labeled ligand Kd B,,, ~

Scatchard Eadie- H ofstee Scatchard Eadie-Hofstee

nM

[~Ala ’ , MePhe4, Gly-o15]Enkephalin 1.18 (0.10) [~Pen~,~]Enkephal in 1.16 (0.16) Dermorphin 0.46 (0.03) Ethylketocyclazocine X1 0.38 (0.05) x2 12.1 (1.8)

~~

fmol/mg protein

1.10 (0.09) 114.8 (6.0) 110.6 (5.4) 1.07 (0.14) 52.6 (3.8) 51.1 (3.4) 0.44 (0.03) 92.0 (4.5) 90.1 (4.2)

nd nd

13.5 (1.7) nd 86.8 (7.2) nd

Table 2. Potencies of opioid ligands in inhibiting the binding at the p sites ( r 3 H ] [ ~ A l a ’ , MePhe4, Gly-ol’lenkephalin, 0.3 n M ) , the 6 site ([’HJ [~Pen’,~/enkephalin, 0.9 n M ) and the x sites ([3H]ethylketocyclazocine, 1.15 n M ) in homogenates of rut brain or guinea-pig cerebellum (1.4 mg membrane protein), at 24°C in 50 mM TrisjClpH 7.4 The inhibitory constant Ki and the pseudo-Hill coefficient (h) were calculated from IC50 values using the Cheng and Prussof relation. Reported values are the mean of 2, 3 or 4 experiments carried out in duplicate and values in parentheses are SEMs. The selectivity ratio at the p sites is expressed as the ratio Ki(6)/Ki(p)

Ligand [3H] [DAla’, MePhe4, [3H] [DPenz’5]- [3H]Ethylketo- Selectivity Gly-olS]Enkephalin Enkephalin cyclazocine ratio

Ki h Ki h Ki h

nM nM nM

_ _ _ ~ ~ -

Dermorphin 0.70 (0.02) 2.11 (0.03) 61.7 (4.1) 1.17 (0.08) > 5000” - 88.1 > 5000‘ -

[~Ala’, MePhe4, Gly-ol’1Enkephalin 1.06 (0.09) 0.89 (0.04) 89.2 (6.5) 1.13 (0.06) >5000 - 84.1

[~Pen’.~]Enkephalin 441 (31) 0.98 (0.04) 2.15 (0.08) 0.98 (0.03) > 5000 - 0.0048 Ethylketocyclazocine - - - - 0.77 (0.04)” 1.14 -

0.52 (0.05)‘ 1.09 -

~~

a Inhibition of [3H]ethylketocyclazocine binding in the rat brain homogenates in the presence of 200 nM [DAla’, MePhe4, Gly-ol’lcnke- phalin (p blocker) and 200 nM [~Pen~~~lenkephal in (6 blocker).

Inhibition of [3H]ethylketocyclazocine binding in homogenates of guinea-pig cerebellum

sake of homogeneity under similar experimental conditions, with the same preparation of rat brain membranes, purified radiolabels and unlabeled compounds to characterize the binding systems used.

Scatchard (Fig. 3A) and Eadie-Hofstee plots (not shown) of [3H] [DAh2, MePhe4, Gly-o15]enkephalin saturation bind- ing isotherm (0.075 - 10 nM) was linear and suggested the presence of a single population of high-affinity sites. The maximal number of binding sites, B,,,, and the dissociation constant, Kd, were determined from least-squares linear re- gression of Scatchard or Eadie-Hofstee analysis of the satu- ration isotherm. The Kd was 1.18 f 0.10 nM using the Scatchard plot and 1.10 f 0.09 nM from the Eadie-Hofstee plot, and the binding capacity, B,,,, 114.8 3.0 and 110.6 4 2.7 fmol/mg membrane protein, respectively (Tab- le 1).

As shown in Fig. 3B, the Scatchard plot of the saturation isotherm of the [3H] [ ~ P e n ~ ~ ~ l e n k e p h a l i n (0.05 - 10 nM) pre- sents apparent homogeneity of binding (correlation coef- ficient = 0.94). Specific binding showed a linear relationship with tissue concentration (correlation coefficient = 0.96) up to 3.5 mg membrane protein/assay (not shown). However, high amounts of non-specific binding were observed even at

low protein concentrations and produced some scatter in the experimental data. The SD (Erad) was around 9%. Estimation of the binding parameters of [’HI [~Pen~,~]enkephal in gave the following values: Kd = 1.16 nM and B,,, = 52.6 fmol/ mg protein using Scatchard analysis; Kd = 1.07 nM and B,,, = 51.1 fmol/mg protein using Eadie-Hofstee analysis (Table 1).

The x opioid receptors were specifically labeled in prep- arations of rat brain membranes by using the non-selective x opioid [3H]ethylketocyclazocine in the presence of saturating concentration (200 nM) of non-radioactive [DAla’, MePhe4, Gly-o15]enkephalin and [~Pen~,~]enkephal in to block binding of the radioligand to p and 6 opioid receptors. Under these conditions, Scatchard analysis of specific [3H]ethylketo- cyclazocine binding revealed a biphasic curve (Fig. 3 C). Com- puter-assisted nonlinear regression analysis of the data afforded a best fit for an isotherm calculated for two binding components; the first, x,, with a Kd = 0.38 0.05 M and a B,,, = 13.5 1.7 fmol/mg protein, the second, x2 , with a Kd = 12.15 + 1.81 nM and a B,,, = 86.8 7.2 fmol/mgpro- tein.

The homogeneity of [DAh2, MePhe4, Gly-o15]enkephalin and [uPen2,’]enkephalin binding in the rat brain was con-

630

1 0 0

5 0

0,

-

-

- 9 -7 - 5 - 3 -11

1 0 0

5 0

0 -

A

-

-

I " " " " ' ~ " " ' ~ 1

-1 1 - 9 - 7 - 5 - 3

l 1 " " " " " l

-11 - 9 -7 - 5

0 5 0 l o o

B ( f rnol i r n g l

Log dose / M Fig. 4. Opioid receptor binding selectivity qf [DAla2, MePhe4, Gly-ol'/enkephalin, [~Pen ' ,~]enkephul in , ethylketocyclazocine and dermorphin in [he rut bruin. (A) Inhibition of [3H] [~Ala ' , MePhe4, Gly-ol']enkephalin specific binding (0.5 nM) to rat brain membranes (0.7 mg/ml membrane protein) by increasing concentrations of unlabeled [uAla', MePhe4, Gly-o15]enkephalin ( O ) , [DPen2'5]enkephdh (H) and dermorphin (V). (B) Inhibition of [3H] [~Pcn'.~]enkephalin specific binding (0.5 nM) to rat brain membranes (0.7 mg/ml membrane protein) by increasing concentrations of unlabeled [~Pen'.~]enkephalin (fl), [DAIa', MePhe4, Gly-o15]enkephalin (V) and dermorphin (0). (C) Inhibition of [311]ethylketocyclazocine specific binding (1 .I 5 nM) to rat brain membranes (0.7 mg/ml membrane protein) by increasing concentrations of ethylketocyclazocine (0) and dermorphin ( x ). Expcriments were conducted in the presence of 200 nM [DAla', MePhe4, Gly-ol']enkephalin and [~Pen'.~]enkephalin as p- and 6-receptor blockers, respectively. The percentage of specific binding was calculated as 100 x (B,-Bn)/(Bo-Bn), where B, and B, are respectively the amount bound in thc presence or in the absence of competing compound, and B, is the non-specific binding, i.e. the binding in the presence of 1000 nM unlabeled [~Ala ' , MePhe4, Gly-o15]enkephalin or [~Pen'.']enkephalin or the binding in the presence of 1000 nM ethylketocyclazocine plus 200 nM [DAIa', MePhe4, Gly-ol'lenkephalin and [~Pen'.~]enkephalin, respectively. The solid lines are theoretical fits to a simple binding isotherm determined by non-linear least-squarcs regression analysis. The data shown are for a single representative experiment. Values of IC50 were determined by regression analysis based upon 2-5 independant experiments carried out in duplicate. (D) Scatchard analysis of [3H]dermorphin specific binding to rat brain membranes (1.4 mg membrane protein) at 24'C in Tris/Cl pH 7.4 in the presence of 100 nM unlabeled [~Pen'~~]enkephaIin (*) or [DAla', MePhe4, Gly-ols]enkephalin (0). Non-specific binding was measured in the presence of 1000 nM unlabeled dermorphin plus 10 nM [~Pen'.~]enkephalin or [DAla', MePhe4, Gly-o15]enkephalin, respectively. Value of Kd and B,,, were calculated from the slope and intercept estimated by linear least-squares analysis of the data

firmed by displacement experiments of each unlabeled peptide against its 3H-labeled analog. The displacement curves showed (Fig. 4) a pseudo-Hill coefficient close to unity and both could be fitted to a simple competitive model assuming

only one population of binding sites. The apparent Ki values obtained for the two peptides were in close agreement with their Kd values determined by Scatchard analysis of saturation binding isotherms (Table 2).

63 1

In the presence of 200 nM of each of these two peptides, Receptor binding selectivity ojdermorphin in the rat bruin non-labeled ethylketocyclazocine competed for the binding of [3H]ethylketocyclazocine to rat brain membranes with a K, value of 0.77 nM (Fig. 4). At this radioligand concentration, labeliiig of the high-affinity and low-density x1 sites is favoured as shown by the close agreement between values of Ki and Kd (Table 2). Further experiments were conducted in the absence of p and 6 blockers in homogenates of guinea pig cerebellum. Whereas neither p nor 6 sites were found in this tissue, it has been shown to contain an homogenous popu- lation of x-opioid sites [42]. In a membrane preparation of guinea-pig cerebellum, the specific binding of 1.15 nM [3H]ethylketocyclazocine was readily inhibited by non-labeled ethylketocyclazocine with a Ki value of 0.52 f 0.05 nM (not shown), which is a value close to the Kd of x1 receptor subtype as determined in the rat brain by Scatchard analysis of satu- ration binding isotherms.

Whereas unlabeled [DAla’, MePhe4, Gly-ol’lenkephalin was very potent in its ability to block [3H] [~Ala’, MePhe4, Gly-o15]enkephalin binding to rat brain membranes, [~Pen’~~]enkephalin was a very weak inhibitor, having a Ki value of 441 31 nM. A nearly opposite result was obtained with inhibition of [3H] [~Pen’,’]enkephalin by [DAla’, MePhe4, Gly-01’1enkephalin (Fig. 4). The latter displaced specifically bound [3H] [DPen’.’]enkephalin with a K, value of 89.2 & 4.1 nM, indicating a very low potency. The competi- tive effects of [DAla’, MePhe4, Gly-ol’lenkephalin and [~Pen’~~]enkephalin against 1.15 nM [3H]ethylketocyclazo- cine were studied in guinea-pig cerebellum preparations. No displacement was found by using 5000 nM of either peptide (Table 2).

Value of the ratio of Ki values for unlabeled [DAla’, MePhe4, Gly-ol’lenkephalin or [~Pen’~’]enkephalin to inhibit [3H] [DPen2ss]enkephalin versus their ability to inhibits [3H] [DAla’, MePhe4, Gly-ol’lenkephalin binding are in accord- ance with the reported high selectivity of [~Ala’, MePhe4, Gly-o15]enkephalin and [~Pen’,’]enkephalin for the p and 6 opioid receptors, respectively [39, 411.

Saturation analysis of [3H]dermorphin binding to rat brain tissues

The saturation isotherm of the specific binding of [3H]dermorphin in rat brain membranes was characterized over the concentration range of 0.065-7.5 nM. The specific binding was linear with tissue concentration up to 4 mg mem- brane protein/assay (correlation coefficient = 0.99, P < 0.1 YO). Very low non-specific binding was detected. For instance, at a tissue concentration of 0.7 mg/ml membrane protein per assay and 0.2 nM [3H]dermorphin (20000 cpm per assay), the total binding represented 2060 cpm, 90 cpm accounting for non-specific binding. The degradation of [3H]dermorphin in the crude rat brain membranes prepared for binding assay was very slow since only 6% hydrolysis occurred after a 5-h incubation at 24°C as evaluated by the procedure of Gay and Lahti 1361.

Scatchard representation of the data showed (Fig. 3D) that a single high-affinity class model is sufficient to character- ize the specific interaction with a Kd value of 0.46 * 0.03 nM and a maximal binding capacity B,,, of 92 f 4.5 fmol/mg membrane protein (Table 1). The SD (Erad) was around 5%. Eadie-Hofstee analysis of the data generated essentially ident- ical values (Table 1).

Dermorphin, when tested for its effect on the binding of the highly selective p opioid peptide [3H] [DAla’, MePhe4, Gly-o15]enkephalin (0.3 nM) in rat brain preparations, com- petitively inhibited in a concentration-dependent manner with a pseudo-Hill coefficient close to unity and a Ki of 0.70 kO.02 nM (Table2). This value is close to that of unlabeled dermorphin under similar conditions (Table 2) and, furthermore, similar to the Kd value of dermorphin determined from saturation experiments. In contrast, dermorphin was found to be 88 times less active in competing with the binding of the 6 probe [3H] [~Pen**~]enkephalin under the same con- ditions (Fig. 4), the Ki being 61.7 nM, a value close to that determined with unlabeled [~Ala’, MePhe4, Gly-o15]enke- phalin under similar experimental conditions (Table 2). Ac- cordingly, 5 pM dermorphin was necessary to displace fully the specific binding of 0.9 nM [3H] [~Pen’~’]enkephalin to rat brain membranes. Confirmation of this selectivity pattern was obtained when the saturation isotherms of [3H]dermorphin in rat brain membranes were studied: in the presence of 100 nM unlabeled [~Pen’~’]enkephalin the binding param- eters remained almost unchanged (Kd = 0.74 nM, B,,, = 106 fmol/mg protein), whereas the specific interaction was completely abolished with 100 nM [DAla’, MePhe4, Gly- o15]enkephalin.

The cross-reactivity of dermorphin towards the x opioid receptors was investigated in the rat brain membranes prep- aration in the presence of 200nM [DAla’, MePhe4, Gly- ol’lenkephalin and [~€‘en’~’]enkephalin as p and 6 blockers, respectively. Whereas unlabeled ethylketocyclazocine com- peted for the binding of [3H]ethylketocyclazocine with a Ki of about 0.8 nM (Fig. 4), dermorphin was found to be virtually inactive (Ki > 5000 nM). Similar results were obtained in homogenates of guinea pig cerebellum (Table 2).

As reported in Table 2, the Ki values of dermorphin are 61.7 nM and 0.7 nM on 6 and p binding sites, respectively are indicative of a selectivity ratio Ki(6)/Ki(p) = 88. Under identical experimental conditions, [DAla2, MePhe4, Gly- ol’]enkephalin exhibited a selectivity index of 84 (Table 2). Thus, both the binding affinity and the selectivity ratio of dermorphin for the p opioid receptor are similar to those exhibited by the purest synthetic probe available up to date.

Displacement oJ” 3H]derrnorphin binding by opiates and op io id pep t ides

In order to demonstrate further the selectivity of dermorphin for the p receptors, a wide range of opiates and opioid peptides were tested for their ability in displacing 0.5 nM [3H]dermorphin from rat brain membrane prep- arations (Fig. 5). The displacement curves obtained could all be fitted to a simple competitive model assuming a single population of binding sites. The pseudo-Hill coefficients (Table 3) accounted for the homogeneity of the binding curves. As expected, p-selective ligands were very potent in blocking [3H]dermorphin binding to rat brain membranes. In particular, unlabeled morphine, naloxone and [DAla’, MePhe4, Gly-ollenkephalin selectively displaced bound [3H]dermorphin with K , values close to their Kd values as determined from saturation experiments. Note that nonla- beled dermorphin and [DAla’, MePhe4, Gly-ol’]enkephalin are almost equipotent in inhibiting [3H]dermorphin binding (Table 3). In contrast, &selective ligands [~Pen’.~]enkephalin, [DPen’, Pen5]enkephalin, [DThr’Ienkephalin-Thr and [DAla’,

632

A

- 1 1 -9 -7 - 5 -3

Log dose / M Fig. 5. Ability qf opiates and opioidpeptides to displace 0.5 n M (3H]dermorphinfrom rat brainpreparations. Details of the calculation of specific binding arc described in Table 3. (A) /I-endorphin (m); [Met’lenkephalin (n): morphine ( A ) ; dermorphin (0); dynorphin A (0) . (B) [DAla’, MePhe4, Gly-o15]enkephalin ( 7 ) ; naloxone (0); [~Pen~ .~ ]enkepha l in ( 1); [DThr2]enkephalin-Thr (0); ethylketocyclazocine (+); [~Ala’,~Leu~]]enkephalin (m); [~Ser’]enkephalin-Thr (1); [DPen’, Pen5]enkephalin (N). The data shown are for a single representative experiment. IC50 values were determined by regression analysis based upon 2 - 5 independent experiments carried out in duplicate

Table 3. Potency of’opiates and opioid ligands in inhibiting the binding of (3H]dermorphin (0.5 n M ) in homogenates of rat brain (1.4 mg membrane protein) at 24°C in TrisjClpH 7.4 The inhibitory constant (Ki) and the pseudo-Hill coefficient (h) were calculated from ICs0 using the Cheng and Prussof relation. Reported values (k SEM) arc the mean of 2, 3 or 4 experiments carried out in duplicate

Ligand Ki h

P-Endorphin Dermorphin [DAla’, MePhe4, Gly-o15]Enkephalin Naloxone Dynorphin A Morphine Ethylketocyclazocine [DAla’, ~Leu’IEnkephalin [~Ser’IEnkep halin-Thr [Met5]Enkephalin [~Thr’IEnkephalin-Thr [DPefl’, Pen ’1 Enkephalin [~Pen’ ,~ ]En kephalin

nM 0.34 (0.01) 1.13 (0.05) 0.73 (0.02) 0.97 (0.03) 0.85 (0.03) 1.02 1.58 (0.06) 1.14 (0.04) 1.87 (0.13) 1.14 (0.07) 2.34 (0.09) 1.14 (0.03) 2.36 (0.16) 1.03 (0.07) 5.75 (0.15) 1.14 (0.04) 6.23 (0.23) 0.94 (0.06)

9.2 (0.3) 0.91 (0.02) 16.4 (1.26) 1.24 (0.09) 348 (26) 0.97 (0.06) 595 (41) 0.89 (0.03)

uLeu’]enkephalin had low tendency to bind to [3H]dermorphin high-affinity sites (Fig. 5). Furthermore, their K, values (Table 3) were in excellent agreement with those previously determined by using [3H][~Ala2, MePhe4, Gly-

o15]enkephalin as selective p radiolabel (Tables 2 and 3). Note that [ ~ P e n ~ ~ ~ l e n k e p h a l i n is the least potent ligand in displacing bound [3H]dermorphin and is only marginally less potent than [DPen’, Pen5]enkephalin (Table 3). The non-selective x ligand, ethylketocyclazocine, exhibited a good potency inhibiting [3H]dermorphin binding, a finding in accordance with its low selectivity for x sites and high residual affinity for p sites [41].

Having observed the profound differences in the inhibitory potency of selective ligands on [3H]dermorphin binding, sev- eral naturally occurring opioid peptides were examined for their ability to compete with [3H]dermorphin for the p recep- tors (Fig. 5) . The apparent Ki values are listed in Table 3. The rank of potency observed was: P-endorphin > dermorphin > dynorphin A > [Met’lenkephalin. This rank order parallels closely that determined by several other groups using [3H] [ ~ A l a , MePhe4, Gly-ol’lenkephalin as radiolabel. Note that dynorphin A has a rather high inhibitory potency despite its potential breakdown at 24°C in these rat brain membrane preparations.

Distribution of (3H]dermorphin receptor site in the rut bruin us shown by in vitro autoradiography

The distribution of [3H]dermorphin sites in the rat brain was visualized by in vitro autoradiography of thaw-mounted sections. On rat brain slices, as in the membrane preparations, specific binding routinely represented > 96% of the total bind- ing at a ligand concentration of 1.5 nM. In addition, binding parameters and specificity profile of [3H]dermorphin on slide-

633

3H DERM

A B C

+ DPDPE D

+ DAGO E

+ DERM F

Fig. 6. The distribution of [3H]dermorphin sites in the rat brain visualized by in vitro autorudiography of thaw-mounted sections. Sections of brain labeled with [3H]dermorphin either alone (A-C) or plus [~Pen’.~]enkephalin (D), [DAIa’, MePhe4. Gly-o15]cnkephalin (E) or unlabeled dermorphin (F)

Table 4. Regional distribution of[3H]dermorphin binding site densitks in coronal sections of rat brain as determined by semi-quantitative autoradiography Sections of rat brain were labeled in vitro with [3H]dermorphin as described in Methods. Only semi-quantitative density values arc re- ported in this table and are as follows: + + + +, very dense; + + +, dense; + +, moderate; f , light and 0, undetectable

Region Receptor density

Telencephalon Olfactory tubercles Cortex layer I Cortex laycr IV Cortex layers 11, V, VI Nucleus accumbens Caudate putamen (patches) Globus pallidus Hippocampus, pyramidal layer Posteromcdial cortical amygdaloid nucleus

Dien wphalon Medial ha ben ula Thalamus:

mediodorsal ventrolateral ventromedial lateral posterior posterior nuclear group medial geniculate nucleus

Mescwmphalon Superior colliculus Inferior colliculus Interpeduncular nucleus Substantia nigra Medial tcrminal nucleus of accessory tract

+ + ++I I+ + + + + + I+++ + + + + + + + + +++I

+ + + + 0 + + + + +++ + + + I++

+ + + + + + + + + + + + + +ll

mounted sections were similar to those obtained with mem- brane homogenates. For instance, binding of [3H]dermorphin was completely displaced by 10 nM [~Ala’, MePhe4, Gly- o15]enkephalin. In contrast up to 96% of the labeling still remained in the presence of 100 nM [~Pen~~’]enkephalin (Fig. 6). This suggests that the visualized sites are probably similar to those detected in membrane preparations studies. Autoradiograms of selected coronal sections of rat brain are shco~wn in Fig. 6 and semi-quantitative data are summarized in Table 4. In striatum, patches of high density were observed, surrounded by a less intense diffuse labeling. Dermorphin sites also occurred at high density in the medial aspect of the nucleus accumbens. The habenula, the superior colliculi and the interpeduncular nucleus were also heavily labeled. In the thalamus, very dense labeling was selectively localized in the mediodorsal and ventromedial nuclei. Dense labeling was ob- served in the posterior nuclear group, the lateral posterior and medial geniculate nuclei while no site was detectable in the ventrolateral nucleus. Moderate densities were observed in olfactory tubercles and the globus pallidus. In the neocortex, layers I and IV contained medium levels of receptor sites, while other layers were less densely labeled. The corpus callosum lacked [3H]dermorphin sites.

DISCUSSION

The aim of this work was to provide a complete evaluation of the selectivity and affinity of dermorphin for each of the p, 6 and 1.1 binding sites in the rat brain. This was achieved in the following ways. (a) The homogeneity of [3H]dermorphin binding sites was evaluated by equilibrium measurements covering a wide range of concentrations. (b) The ability of dermorphin to inhibit the binding of the highly p-selective

634

ligand [3H] [DAla’, MePhe4, Gly-ol’lenkephalin and that of the 6 probe [3H] [ ~ P e n ~ ~ ~ l e n k e p h a l i n was analysed; the cross- reactivity of dermorphin towards the x opioid receptors was also investigated by using [3H]ethylketocyclazocine, a non- selective x ligand, in the presence of p and 6 blockers. (c) The ability of highly selective receptor-type ligands to displace bound [3H]dermorphin was compared to that of dermorphin in displacing bound [3H] [DAla’, MePhe4, Gly-ol’lenkephalin and [3H] [~Pen’.’]enkephalin [3H]ethylketocyclazocine from the p, 6 and K sites, respectively. (d) The inhibition of [3H]dermorphin binding by a series of opioid ligands of known selectivity and affinity for the morphine receptor was ana- lysed. (e) Finally, in vitro autoradiography of sites revealed by [3H]dermorphin within the brain showed that the distribution of [3H]dermorphin binding sites correlated well [43 - 461 with the p specificity of this peptide (Fig. 6 and Table 4).

Results from these analyses clearly demonstrate that dermorphin appears to be the only naturally occurring opioid peptide that combines high affinity, strict selectivity and very potent agonist activity towards the morphine (p) receptor. In particular, dermorphin is more potent in vivo than any natural or synthetic opiates or opioid peptides in inducing long lasting analgesia. For instance, p-endorphin, the most potent of the mammalian opioid peptides is about five times less active than dermorphin by the intracranial route (Fig. 2), both peptides being still highly active through intravenous administration (unpublished results). By comparison with selective synthetic opioids, dermorphin combines a similar selectivity ratio and an affinity twice as high as [DAla’, MePhe4, Gly-ol’lenke- phalin, the p-ligand most often used for in vitro experiments [50]. Morphiceptin and the morphiceptin-derived ligand, PL 17, display a selectivity for the p sites almost identical to that of dermorphin but an affinity 600 and 20 times lower, respectively [47]. In addition, the very low non-specific binding exhibited by [3H]dermorphin makes this peptide a highly ap- propriate radioactive probe for quantitative autoradiography and for testing the existence of postulated preceptor subtypes that could mediate different physiological functions [48, 491. Further, this readily obtainable, natural peptide with high peptidase resistance should be highly appropriate for in vivo investigations.

Due to its D-amino acid residue and to the lack of enkephalin core, dermorphin is unique both among peptides synthesized by animal cells and, particularly, among all the hitherto known opioid peptides. Dermorphin thus represents an excellent model to investigate the relationships between structure/conformation, receptor selectivity and analgesic ac- tivity. For instance, superimposition of the three-dimensional structure of dermorphin on morphine, [L-Ala2]dermorphin and [~Ala’, MePhe4, Gly-ol’lenkephalin, both in aqueous solution and once bound to the p receptor, should allow accurate identification and characterization of determinants of high-affinity binding to the morphine receptor.

This work was supported in part by funds from the Fondation pour la Recherche Mkdicale Francaise and the Ministire de la Recherche et de I’Eneignement Supirieur (87/H/0303). Thc expert technical assis- tance o f J . J. Montagne is acknowledged.

REFERENCES 1. Anastasi, A,, Erspamer, V. & Endan, R. (1986) Arch. Biochem.

2. Erspamcr, V., Falconieri-Erspamer, G. & Ceil, J. M. (1986) Comp. Biophys. 125, 57-68.

Biochem. Physiol. 85 C, 125 - 137.

3. Erspamer, V. & Melchiorri, P. (1980) Trends Pharmacol. Sci. I ,

4. Erspamer, V. & Melchiorri, P. (1983) in Neuroendocrine perspec- tives (Miiller, E. E. & McLeod, M. M., eds) vol. 2, pp 37 - 106, Elscvier, Amsterdam.

391 -395.

5. Zasloff, M. (1987) Proc. Nut1 Acad. Sci. USA 84, 5449 -5453. 6. Erspamer, V., Melchiorri, P., Broccardo, M., Falconieri-

Erspamer, G., Falaschi, P., Improta, G., Negri, L. & Renda, T. (1981) Peptides 2, 7-16.

7. Erspamer, V. (1983) Trends Neurosci. 6 , 200-201. 8. Feurle, G. E., Carraway, R. E., Rix, E. & Knauf, W. (1983) J .

Clin. Invest. 76, 156- 162. 9. Hoffman, W., Bach, T. C. & Kreil, G. (1983) EMBO J . 2, 11 1 -

124. 10. Broccardo, M., Erspamer, V., Falconieri-Erspamer, G., Impota,

G., Linari, G., Melchiorri, P. & Montecucchi, P. C. (1981) Br. J. Pharmacol. 73, 625-631.

11. Montecucchi, P. C., de Castiglione, R., Piani, S., Gozzini, L. & Erspamer, V. (1981) Znt. J . Peptide Protein Res. 17, 275-283.

12. Richter, R., Egger, R. & Kreil, G. (1987) Science 238, 200-202. 13. Nakanishi, S., Inoue, A., Kita, T., Chang, A. C. Y., Cohen, S . &

Numa, S. (1979) Nature 278, 423-429. 14. Kakidani, H., Furutani, Y., Takahashi, H., Noda, M., Morimoti,

Y., Hirose, T., Asai, M., Inayama, S . , Nakanishi, S. & Numa, S. (1982) Nature 298, 245-250.

15. Noda, M., Furutani, Y., Takahashi, H., Toyosato, H., Hirose, T., Tnayama, S., Nakanishi, S . & Numa, S. (1982) Nature 295,

16. Negri, L., Melchiori, P., Erspamer, V. & Falconieri-Erspamer, G.

17. Buffa, R., Solcia, E., Magnoni, E., Rindi, G., Negri, L. &

38. Tsou, K., Wang, F. S . , Wang, S. H. & Tang, Y. Q. (1985)

19. Mor, A., Delfour, A., Amiche, M., Sagan, S., Nicolas, P., Grassi,

20. Castellano, C. & Pavone, F. (1985) Behuv. Neurosci. YY, 1120-

21. Stevens, C. W. & Yaksh, T. L. (1986) Bruin Res. 385, 300-304. 22. Guglietta, A., Irons, B. J., Lazarus, L. H. & Melchiorri, P. (1987)

Endocrinology 120, 2137-2141. 23. Salvadori, S., Sarto, G. & de Tomatis, R. (1982) Int. J . Peptides

Protein Res. 19, 536- 542. 24. Darlak, K., Grzonka, Z., Janicki, P., Czlonkowski. A. &

Gumulka, S. W. (1983) J . Med. Chem. 26, 1445- 1447. 25. De Castiglione, R. & Perseo, G. (1983) Int. J . Peptide Protein

Res. 21, 471 -474. 26. Cervini, M. A., Rossi, A. C., Perseo, G. & de Castiglione, R.

(1985) Peptides 6 , 433 -437. 27. Sasaki, Y., Matsui, M., Fujita, H., Hosomo, M., Taguchi, M.,

Suzuki, K., Sakurada, S., Sato, T., Sakurada, T. & Kisara, K. (1985) Neuropeptides 5, 391 -394.

28. Kisara, K., Sakurada, S., Sakurada, T., Sasaki, Y., Sato, T., Suzuki, K . & Watanabe, H. (1986) Br. J. Pharmacol. 87, 183- 189.

29. Westphal, M., Hammonds, R. G. & Li, C. H. (1985) Peptides 6, 149 - 152.

30. Rossi, A. C., de Castiglione, R. & Perseo, G. (1986) Peptides 7, 755-759.

31. Nicolas, P., Delfour, A., Boussetta, H., Morel, A,, Rholam, M. & Cohen, P. (1986) Bioclwm. Biophys. Res. Cornmun. 140, 565- 573.

32. D’Amour, F. E. & Smith, D. L. (1941) J. Phurmacol. Exp. Ther. 72,74 - 79.

33. Amichc, M., Delfour, A., Morgat, J. L., Roy, J., Houvet, J. & Nicolas, P. (1987) Biochem. Biophys. Res. Commun. 148,1432- 1439.

34. Lowry, O., Rosebrough, N. J., Farr, A. L. &Randall, R. J. (1951) J. B i d . Cliem. 193, 265 -275.

35. Cheng, Y. C. & Prussof, W. H. (1973) Biochem. Phurmucol. 22,

202 - 206.

(1981) Peptides 2, 45-49.

Melchiorri, P. (1982) Histochemistry 76, 273-276.

Neuropeptides 5, 449 -452.

J. & Pradelles, P. (1988) Neuropeptides 13, 51 -57.

1127.

3099- 3108.

635

36. Gay, D. D. & Lahti, R . A. (1981) Int. J . Peptide Protein Res. 18,

37. Nicolas, P. & Li, C. H. (1985) Proc. Nut1 Acad. Sci. U S A 82,

38. Kosterlitz, H. W. & Paterson, S. J. (1981) Br. J . Pharmacol. 73, 299P.

39. Mosberg, H. I., Hurst, R., Hruby, V. J., Gee, K., Yamamura, H. I., Galligan, J. J. & Burks, T. F. (1983) Proc. Nut1 Acad. Sci.

40. Akiyama, K., Gee, K. W., Mosberg, H. I., Hruby, V. J. & Yamamura, H. I. (1985) Proc. Natl Acad. Sci. USA 82,2543- 2547.

41. Kosterlitz, H. W. & Paterson, S. J. (1980) Proc. R. Soc. Lond. B,

42. Rohson, L. E., Foote, R. W., Maurer, R. & Kosterlitz, H. W.

43. Quirion, R. R., Bowen, W., Herkenham, M. & Pert, C. B. (1 982)

102-110.

3 1 78 - 31 8 1.

USA 82, 2543 -2547.

210, 11 3 - 122.

(1984) Neuroscience 12, 621 -627.

Cell. Mol. Neurobiol. 2, 333 - 345.

44. Tempel, A. & Zukin, S. R. (1987) Proc. Natl Acad. Sci. USA 84,

45. Mansour, A., Khachaturian, €I., Lewis, M. E., Akil, H. & Watson,

46. Waksman, G., Hamel, E., FourniC-Zaluski, M. C. & Roques, B.

47. Blanchard, S. G., Lee, P. H. K., Pugh, L. W. W., Hong, J. S. &

48. Pasternak, G. W. (1982) Life Sci. 31, 1303-1306. 49. Rothman, R. B., Jacobson, A. E., Rice, K. C. & Herkenham, M.

50. Patcrson, S. J., Robson, L. E. & Kosterlitz, H. W. (1984) Peptides

51. Amiche, M., Sagan, S., Mor, A., Delfour, A. & Nicolas, P. (1989)

52. Mor, A., Delfour, A., Sagan, S., Amiche, M., Pradelles, P.,

4308 - 4312.

S. J. (1988) Trend.7 Neurosci. I / , 308-314.

P. (1986) Proc. Natl Acad. Sci. U S A 83, 1523- 1527.

Chang, K. J. (1987) Mol. Pharmacol. 31, 326-333.

(1987) Peptides8, 1015-1021.

6,147-188.

Mol. Pllnml?lncol. 35, 774 - 779.

Rossier, J . & Nicolas, P. (1989) FEBS Lett. 255, 269-274.