BRAIN RESEARCH

ELSEVIER Brain Research 735 (1996) 177-187

R e s e a r c h r e p o r t

Galanin receptor antagonists attenuate spinal antinociceptive effects of DAMGO, tramadol and non-opioid drugs in rats

N o r m a S e l v e *, W e r n e r E n g l b e r g e r , E l m a r F r i d e r i c h s , H a g e n - H e i n r i c h H e n n i e s , W o l f g a n g R e i m a n n ,

B o b W i l f f e r t

Abteilung Pharmakologie, Griinenthal GmbH, Zieglerstr. 6, 52078 Aachen, FRG

Accepted 7 May 1996

Abstract

The involvement of endogenous galanin to antinociception elicited by intrathecally (i.t.) or systemically administered drugs from different chemical and therapeutic classes was investigated using the rat Randall-Selitto or the rat tail-flick test, in the absence or presence of the i.t. administered galanin receptor antagonists galantide and M-35. Antinociception elicited by i.t. tramadol (24 ~g), DAMGO (1 Ixg), clonidine (48 p,g), desipramine (6 ~g) or fenfluramine (60 Ixg) was attenuated by i.t. galantide (2 ~g); the attenuation reached significance at least at one time point. A partial antagonism by i.t. galantide was also observed against the antinociception of i.p tramadol (10 mg/kg), i.v. clonidine (1 mg/kg), i.p. desipramine (1 mg/kg), or i.p. dipyrone (1000 mg/kg), but antinociception by i.p. fenfluramine (30 mg/kg) was not affected. Using M-35 (2 Ixg i.t.), the antinociception of i.t. tramadol or DAMGO was attenuated, but no inhibition was observed when clonidine, desipramine or fenfluramine were used i.t. If drugs were administered systemically, only antinociception of i.p. fenfluramine but not that of i.p. tramadol, or i.v. clonidine, or i.p. desipramine or i.p. dipyrone was attenuated. In the rat tail-flick test, co-injection of either 2 Ixg i.t. galantide or M-35 with i.t. tramadol (12 ixg) almost abolished the antinociceptive effect, whereas the antinociception of systemically administered tramadol (4.6 mg/kg i.p.) was only partially attenuated by i.t. galantide and not affected by i.t. M-35. Binding studies in dorsal spinal cord tissue showed no affinity of galantide or M-35 to spinal p~-, or ~-, or K-opioid receptors and none of the other drugs interfered with the spinal galanin binding site. These data give further support of at least a partial galanin link in spinal processes of antinociception.

Keywords: Galanin receptor antagonist; Clonidine; Desipramine; Dipyrone; Fenfluramine; Antinociception

1. Introduct ion

Galanin is a widely distributed neuropeptide with multi- ple biological and pharmacological effects and is not a member of any known family of neuropeptides (see [2] for review). The 29-amino-acid peptide (30 in man) was dis- covered by V. Mutt and coworkers [45] and was named after the NH2-terminal glycine and COOH-terminal ami- dated alanine residues of the molecule. The sequences of galanin of different species are available now [2], which show homologous conservation of the N-terminal sequence confirming earlier studies of specific structural require- ments for receptor interactions using several galanin frag- ments [25].

A fragmentary list of the effects of galanin [2] includes inhibitory effects on glucose-induced insulin release, and hippocampal glutamate and scopolamine-induced acetyl-

* Corresponding author. Fax: + 49 (241) 569-2544.

0006-8993/96/$15.00 Published by Elsevier Science B.V, PI1 S0006- 8993(96)00543-4

choline release, and locus coeruleus firing rate, whereas growth hormone release and food intake are facilitated by galanin. Galanin modulates K + and N- and L-type Ca 2+ channels. G i and G o protein coupled receptors including the second messenger systems adenylate cyclase as well as phopholipase C are involved in galanin-mediated effects. Acetylcholine, serotonin, and norepinephrine are described as co-transmitters to galanin [2]. The existence of two distinct galanin receptor subtypes is discussed [3] and meanwhile a galanin receptor from pig brain has been purified [7] and a functional human galanin receptor has been sequenced and cloned [16].

The presence of galanin has been documented by im- munohistochemical techniques in the gastrointestinal sys- tem [11,30], and a detailed mapping of galanin-like neu- rons in the rat central nervous system is available [44]. Examination of the somatosensory system showed that galanin, under physiological conditions, exists in a few sensory neurons terminating in the dorsal horn of the

178 N. Seh:e el al. / Brain Research 735 (1996) 177 187

spinal cord. Analysis by indirect immunofluorescence dou- ble-staining revealed that subpopulations of calcitonin gene-related peptide-positive as well as substance P-posi- tive neurons were galanin-positive, too [23,48], and galanin may be involved in the processing of nociceptive inputs [48]. The reported effects of exogenously administered galanin on nociception, however, are contradictory, and antinociceptive effects as well as hyperalgesic effects were shown. Galanin caused antinociception in mice [35] and rats [8] when thermal nociceptive stimuli were used, but elicited hyperalgesia to innocuous mechanical stimuli in rats [8]. Weak antinociceptive efficacy of galanin was observed after both thermal or mechanical stimuli in rats with an intact somatosensory system [50], whereas prono- ciceptive effects of short duration by galanin were ob- served following mechanical stimuli to inflamed tissue [24,29,41]. Significant antinociception of the N-terminal 1-16 galanin fragment [51] and synergistic efficacy of galanin given to morphine or a CCK B receptor antagonist was observed in the spinal flexor reflex or the hot plate test in rats [47].

In view of the multiple actions of exogenous galanin the development of galanin receptor antagonists provided a tool for characterisation of the functional role of endoge- nous galanin [1,3,34]. Regarding nociception, investiga- tions using the high affinity galanin receptor antagonist M-35 [galanin-(1-13)-bradykinin-(2-9)-amide] suggest an endogenous galanin-mediated control of neuropathic pain following nerve injury [46,49]. The antagonism to mor- phine-induced antinociception in rats by two different galanin receptor antagonists, galantide [M-15, galanin-(l- 13)-substance P-(5-11)-amide] and M-35, has only re- cently been shown, and a galaninergic link in the spinal antinociceptive effects of morphine was assumed [40].

The aim of this study was to examine whether or not endogenous galanin is involved in spinal antinociception evoked by drugs other than morphine. Considering that morphine primarily activates fx-opioid receptors, but is active at the 8- and K-binding sites, too, the selective Ix-opioid subtype receptor agonist DAMGO [17] was used in the present investigation. A further drug chosen was tramadol, a centrally acting analgesic, which was shown to exert opioid (mainly Ix) and non-opioid modes of action which include norepinephrine re-uptake inhibition as well as indirect serotonergic effects at the supraspinal [10,36] and spinal level [39]. The study also included the non- opioid, partly adjuvant, analgesics, the e~2-adrenoceptor agonist clonidine, the noradrenaline re-uptake inhibitor desipramine, the indirect serotonergic mimetic fenflu- famine, and dipyrone, which belongs to the group of non-acidic non-steroidal ant i - inf lammatory drugs (NSAIDs) [43]. The possible involvement of endogenous galanin in the antinociception of these drugs was tested by use of the galanin receptor antagonists galantide and M-35.

Since the selected drags represent different pharmaco- logical classes and different modes of in vivo antinocicep-

rive efficacy, they were chosen as a tool for examination of the hypothesis that galanin is a more general link in spinal antinociception.

2. Materials and methods

2.1. Animals

Male Sprague-Dawley rats (Iffa Credo, Brussels, Bel- gium) were used with a weight of 140-170 g in the Randall-Selitto paw pressure test, 180-250 g in the tail- flick test and 550-590 g in binding studies. The rats were maintained on a 12 h light-dark cycle with free access to water and food at a temperature of 23-25°C and humidity of 60 _+ 10%. All experiments were conducted in a ran- domised manner during the light period between 7.30 and 18.00 h and were carried out in accordance with the German legislation for the use of experimental animals and with the European Communities Council Directive of 24th November 1986 (86/609/EEC).

2.2. Rat paw pressure test

Rats were made hyperalgesic by subplantar injection of 0.1 ml of a 20% suspension of Baker's yeast into one hind paw 4 h prior to the assessment of nociception according to the method of Randall and Selitto [37] and Gilfoil et al. [15]. The nociceptive thresholds of gently restrained rats were determined with a modified Basile analgesy meter (Ugo Basile, Milan, Italy); a pusher of 0.2 mm 2 tip diameter and a weight range of 0 to 450 g was used. All animals were handled and trained in the test situation two hours before drug administration. The supraspinal nocicep- tive reaction threshold was defined as occurrence of vocal- isation following increasing pressure applied to the dorsal side of the inflamed hind paw, cut-off value 450 g /0 .2 mm 2. The pain thresholds was determined once before and four times (15 or 20, 30, 45 and 60 min) after injection of the drugs.

2.3. Rat tail-flick test

The tail-flick test was performed as described by D'Amour and Smith [9] with minor modifications. Rat tails (about 2.5 cm from the root) were exposed to radiant heat from a light source (analgesia meter 2011; Rhema Labortechnik, Hofheim, Germany) at an energy setting of 25% of maximum. Maximum exposure time was limited to 30 s. Tail-flick latencies were measured twice before and 6 times after intrathecal injection.

2.4. Intrathecal injection

intrathecal injections were performed in analogy to the method described for mice by Hylden and Wilcox [21]

N. Selve et al./ Brain Research 735 (1996) 177-187 179

adopted for rats. Under light ether anaesthesia an 18 gauge needle attached to a Hamilton syringe was directly inserted into the spinal canal between vertebrae L1 and L2 or L2 and L3. An ipsilateral twitch of one of the extremities or the tail indicates the penetration of the spinal cavity. Drugs were injected in a volume of 5 Ixl.

2.5. Binding experiments

Interaction of galanin or galanin receptor antagonists with spinal Ix-, or g-, or K-opioid receptors and the interaction of tramadol with spinal galanin receptor was assessed. Rats were sacrificed by decapitation, and the dorsal half of the spinal cord was prepared as already described [38].

2.6. Interaction o f the different drugs with spinal galanin receptors

Preparation of crude synaptosomal membranes and binding was performed according to Bedecs et al. [4] with minor modifications. In brief, the dorsal part of the spinal cord was homogenized on ice in 5 ml ice-cold 0.32 mol/1 sucrose buffered with 5 mmol/1 HEPES (pH 7.4) with a teflon-glass homogenizer with 10 full up and down strokes at 700 rpm. The homogenate was diluted to 1 mg tissue weight/ml with HEPES buffer and centrifuged for 10 min at 1000 × g at 4°C. Subsequent centrifugation of the su- pernatant for 45 min at 10000 X g at 4°C yielded a pellet which was resuspended in a HEPES buffered medium (pH 7.4) to give a final protein concentration of 125-150 Ixg/ml in the assay. Protein concentration was determined according to Lowry et al. [27] with bovine serum albumin as standard. The medium contained (mmol/1): HEPES 5, NaC1 137, KC1 2.68, MgC12 2.05 and CaC12 1.68, supple- mented with 1 g / l glucose, 0.5 g/1 bovine serum albumin and 1 g/1 bacitracin. Assays were run in duplicate. In a total volume of 400 Ixl, the final concentration of [125i]galanin was 0.1 nmol/1. Incubation lasted for 30 min at 37°C until steady state was reached, and unspecified binding was determined in the presence of 1 Ixmol/1 of the galanin receptor antagonist galantide. The incubation was stopped by addition of 10 ml ice-cold medium fol- lowed by rapid filtration through Whatman G F / B glass fibre mats presoaked for 2 h in medium containing 0.5% polyethylenimine. Filters were washed 3 times with 5 ml ice-cold medium. The radioactivity on glas fibre disks was determined in a gamma counter.

2.7. Interaction with spinal tz-, or 6-, or K-opioid recep- to rs

Preparation of crude membranes and binding to Ix- or K-spinal opioid receptors was performed as follows: Imme- diately after preparation about 1 g rat spinal cord was homogenized on ice in 20 ml ice-cold 50 mmol/1 Tris HC1

buffer (pH 7.4) using a Polytron PT 3000 equipped with a DA-3012/2 cutting edge with 6000 rpm for 15 s. The homogenate was diluted to a volume of 100 ml giving about 10 mg tissue weight/ml Tris HC1 buffer (pH 7.4). After centrifugation for 15 min at 48000 × g at 4°C the pellets were resuspended in the same amount of 100 ml ice-cold Tris HC1 buffer by homogenization with 6000 rpm for 15 s. The homogenate was incubated at 25°C for 30 min and then centrifuged as above. The pellet was again resuspended in the same amount of Tris HC1 buffer and incubated at 25°C for 5 min before use. If not stated otherwise, 900 IXl of the suspension was used to give a protein concentration of 1.17-1.26 mg per assay (580-630 txg/ml). Protein concentration was determined according to Lowry et al. [27] with bovine serum albumin as stan- dard. All assays were run in triplicate with a final volume of 2 ml each. The final ligand concentration was close to its KD-value, i.e. 1 nmol/1 [3H]DAMGO for tx-opioid receptor binding and 1 nmol/1 [3H]CI-977 for K-opioid receptor binding in Tris HC1 buffer (pH 7.4 at 23°C) with or without the drug to be tested. Incubation lasted for 60 min at 25°C. Unspecific binding was determined in the presence of 0.1 mmol/1 levorphanol in the Ix-opioid recep- tor binding assay or in the presence of 1.0 Ixmol/1 brema- zocine in the K-opioid receptor binding assay. The incuba- tion was stopped by rapid filtration through Whatman G F / B glass fibre mats presoaked in 0.5% polyethylen- imine. Filters were washed 2-3 times with 5 ml ice-cold Tris HC1 buffer (pH 7.4) and then extracted with 10 ml Ready Protein ~ (Beckman) for at least 15 h. The radioac- tivity on filter disks was determined in a beta counter.

For the investigations of the interaction of galanin antagonists with spinal ~-opioid receptors, 50 mmol/1 Tris-HC1 buffer (pH 7.4) supplemented with 5 mmol/1 MgCI 2, 45 Ixmol/1 phenylmethyl-sulfonyl-fluoride and 0.45 mg /ml bovine serum albumine was used. The incu- bation time was 4 h at 25°C and the final protein concen- tration was adjusted to 370-430 Ixg/ml. [3H]-p-C1-DPDPE at a concentration of 0.25 nmol/1 was used as the ~ tracer and unspecific binding was determined in the presence of 1 mmol/1 naloxone.

2.8. Calculations and statistical analysis

Time latency to tail-flick or pressure eliciting vocalisa- tion were measured before and after drug administration. Prolongation of the latency or elevation of the threshold were calculated as percent of the maximal possible effect according to the following formula:

postdrug response - predrug response % M P E = × 100

cut offvalue - predrug response

Statistical evaluation included a test for general treatment differences by Kruskal-Wallis analysis, followed by post hoc comparisons using a nonparametric Wilcoxon rank

180 N. Selve et al. / Brain Research 735 (1996) 177 187

sum test statistic. Means_+ S.E.M. are given throughout with n = 10 per group, if not stated otherwise. ED~0 values were calculated by linear regression analysis, confi- dence intervals were estimated according to Litchfield and Wilcoxon.

In binding experiments, ICs0 values were calculated for a one binding site model identical to that used in the TOPFIT programme package by nonlinear adaptation of the data to the law of mass action. If not stated otherwise, ICs0 values are given as means _+ S.E.M. of 3 -6 indepen- dent experiments each.

2.9. Drugs

[3H]CI-977, spec. activity 43 Ci /mmol , [3H]DAMGO ([D-Ala2-NMePhe4-GlyS-ol]-enkephalin, = [3H]DAGO, = [3H]RX783006), spec. activity 60 C i /mmol (Amersham- Buchler, Braunschweig, Germany); [3H]-p-C1-DPDPE ([D- Pen 2,p-C1-Phe 4,D-Pen 5]-enkephalin), spec. activity 49 Ci /mmol , [J25I]galanin, porcine, spec. activity 2200 Ci /mmol (NEN, Bad Homburg, Germany); bremazocine- HCI (RBI, Natick, MA, USA); clonidine-HC1, DAMGO ([D-Ala2-NMePhe4-Gly 5-ol]-enkephalin), desipramine HC1, naloxone HC1 (Sigma, St. Louis, USA); dipyrone (Hoechst, Frankfurt, Germany); fenfluramine HC1 (Servier, Paris, France); galanin (rat), galanin (1-13)-bradykinin (2-9) amide (M-35), galanin (1-13)-substance P (5-11) amide (M-15, galantide; Bachem, Heidelberg, Germany); levor- phanol tartrate (Hoffmann-La Roche, Basel, Switzerland); morphine HCI (Merck, Darmstadt, Germany); tramadol HCI (Griinenthal, Aachen, Germany). All drugs were dis- solved in saline and administered in a volume of 10 m l / k g (i.p.) or 5 m l / k g (i.v.) or 5 Ixl/rat (i.t.).

3. Results

3.1. Antinociceptive effects of indiL, idual drugs

Dose dependency of the antinociceptive effect of the drugs was determined in order to define the individual dose eliciting 40-100% MPE, which was assumed to be suited for investigations of antagonism by the galanin receptor antagonists. None of the agonists or antagonists, given either intrathecally or systemically, altered the gross behavior of the rats. In addition, none of the different substances indicated any impairment of motor coordina- tion, when motor coordination of male Swiss mice (Cd- I(IRC)BR, 18-22 g) was tested using a standard rotarod procedure (n = 5 per group, data not shown).

Intrathecally or systemically administered tramadol elicited dose-dependent antinociception in the rat paw pressure test (Table 1).

Intrathecally administered DAMGO, clonidine, de- sipramine, or fenfluramine elicited dose-dependent antinociception in the rat paw pressure test (Table 1).

Ceiling effects were observed with desipramine and fenflu- ramine. Up to 60 ~xg/rat desipramine no higher effects were detectable than already obtained with 6 Ixg. Fenflu- ramine elicited antinociception with a shallow dose re- sponse curve and 60 txg was the highest dose tolerated without convulsions. Dipyrone up to 400 ~xg (about 1100 nmol) was not active.

Systemically administered clonidine, desipramine, fen- fluramine, or dipyrone elicited dose-dependent antinoci- ception in the rat paw pressure test (Table 1). Ceiling effects were observed when desipramine was given. Up to 50 m g / k g i.p. desipramine, a maximum of less than 40% MPE was obtained, which was already measured when 1 m g / k g was given.

Control experiments included i.t. saline (NaCI), or galantide or M-35 administration with or without pretreat- ment of the animals with i.p. saline and were performed in any of the different studies presented (Figs. 1-3). No increase in pressure threshold by saline or galantide or M-35 administration was detectable in the rat Randall- Selitto paw pressure test (Fig. 1 ; Fig. 3) except of one case when i.t. galantide 10 rain after i.p. saline induced slight but significant increase in pressure threshold 20 and 45 min after i.p. application (Fig. 1B). No increase in time latency was observed by saline or galantide or M-35 injection in the tail-flick test (Fig. 2).

3.2. Antagonism to antinociception by galanin receptor antagonists

3.2.1. Tramadol The effect of the intrathecally administered galanin

receptor antagonist galantide (2 p~g, about 0.75 nmol) on antinociception by either intrathecally or systematically administered tramadol in the rat paw pressure test with yeast-induced hyperalgesia is shown in Fig. 1. Pre-drug pressure for induction of the supraspinal reaction (vocalisa- tion) was 46.0 _+ 0.82 g /0 .2 mm 2 (n = 360) when applied to the inflamed paw; applied to the contralateral paw, pressure to induce vocalisation was 67.8 _+ 0.35 g /0 .2 mm 2 (n = 62) as measured in a comparable group.

Tramadol used in dose of 24 txg i.t. (about 80 nmol) elicited antinociception with a maximum of 58% MPE. Co-injection of 2 Ixg galantide with 24 txg tramadol reduced the antinociception of tramadol to about 35-40% of the effect of tramadol alone, reaching significance 45 rain after drug administration (Fig. 1A).

Tramadol used in a dose of 10 m g / k g i.p. administered 10 min prior to i.t. saline elicited antinociception with a maximum effect of 90% MPE (Fig. 1B). I.t. galantide given to i.p. tramadol-treated rats significantly reduced the antinociception of tramadol to about 50% of the effect of tramadol alone 45 and 60 min after i.p. drug administration (Fig. 1B).

Both sets of experiments were repeated using a different galanin receptor antagonist, M-35 (Fig. 1C,D), which was

N. Selve et al. / Brain Research 735 (1996) 177-187 181

100

75

LU 50

a'~ 25

100

75

LB 50

3 ;

~.~ 25

A lOO C

. . . . . .

75

UJ 50 0.

25

" ' ' ' " D "1"

I "--.~

I I I I I

0 15 0 15 30 45 60

min after i.t. injection

B D

I I I

30 45 60

rain after i.t. injection

T 100

• I : 1 .

- - . ~ * ~I~ 25 "1 " i

~------------~-- --I~ 0

l

I I I I I I I I

0 15 30 45 60 0 15 30 45 60

rain after i.p. injection min after I.p. injection

Fig. 1. A,C: antinociceptive effects of i.t. injected tramadol and interaction with the co-injected galanin receptor antagonists galantide (A) or M-35 (C) in the rat Randall-Selitto paw pressure test; ( • ) tramadol 24 t-~g i.t., ([]) tramadol 24 ~g i.t. and galantide 2 p~g i.t. (A) or M-35 2 t-~g i.t. (C), (O) NaCI i.t., (O) galantide or M-35 i.t. B,D: antinociceptive effects of i.p. injected tramadol and interaction with the galanin receptor antagonists galantide (B) or M-35 (D) in the rat Randall-Selitto paw pressure test. Tramadol was administered 10 min prior to the i.t. injection of either NaC1 or one of the galanin receptor antagonists; ( • ) tramadol 10 mg/kg i.p + NaC1 i.t., ([]) tramadol 10 mg/kg i.p. + galantide 2 ~zg i.t. (B) or M-35 (D) 2 ~g i.t., (O) NAC1 i.p. + NaC1 i.t., (O) NaC1 i.p.+ galantide 2 t-~g i.t. (B) or M-35 2 Ixg i.t. (D). Means + S.E.M. of 10 animals per group. * : Significant differences between tramadol groups in the presence or absence of galantide or M-35; p < 0.05.

g iven intrathecal ly in the same 2 ~zg dose (about 0.75

nmol) . A partial an tagonism to the ant inocicept ion o f

co- in jec ted t ramadol was observed reaching s ignif icant

levels 15 and 45 min after i.t. drug adminis t ra t ion s imilar

to the results obtained with galant ide (Fig. 1C). There was

no an tagonism detectable by i.t. M-35 to the ant inocicep-

t ive ef fec t o f i.p. adminis tered t ramadol (Fig. 1D), which

contrasts with the use o f galantide.

The ef fec t o f the intrathecal ly adminis tered galanin receptor antagonis t galant ide on ant inocicept ion by ei ther

intrathecal ly or sys temica l ly adminis tered t ramadol in the

ta i l - f l ick test is shown in Fig. 2. Pre-drug latencies for

induct ion o f the spinal noc icep t ive ref lex to the thermal

s t imulus were 7.2 _+ 0.21 s ( n = 150). I.t. t ramadol , 12 ~zg (about 40 nmol) , induced antinoci-

cept ion with a m a x i m u m of about 60% MPE. Co- in jec t ion

o f galant ide with t ramadol reduced the an t inoc icept ive

effect o f t ramadol to a lmost less than 16% M P E ; this

reduct ion reached signif icant levels 60, 90 and 120 min after i.t. drug adminis t ra t ion (Fig. 2A).

I.p. t ramadol , 4.64 m g / k g , el ici ted ant inocicept ion with

a m a x i m u m of about 90% M P E (Fig. 2B). I.t. galant ide

g iven to i.p. t ramadol- t rea ted rats s ignif icant ly reduced the

Table 1 EDs0 values of the antinociceptive effect of different intrathecally or systemically administered drugs in the rat paw pressure test using yeast- induced hyperalgesia

Drug EDs0

[ixg/animal] i.t. [mg/kg] i.p. or i.v.

DAMGO 0.3 [0.21-0.45] not investigated Tramadol 23 [16.6-36.7] 7 [5.3-9.6] i.p. Clonidine 16 [7.3-32.6] 0.3 [0.22-0.30] i.v. Desipramine > 60 > 50 i.p. Fenfiuramine > 60 20 [16.5-23.0] i.p. Dipyrone not effective up to 400 155 [113-215] i.p.

182 N. Selve et a l . / Brain Research 735 (1996) 177-187

antinociception of tramadol to about 65% of the effect of tramadol alone 20 and 30 min after i.p. drug administra- tion, and to less than 23% of the effect of tramadol 40, 70, 100, and 130 min after i.p. drug administration (Fig. 2B).

In analogy to the rat paw pressure test both sets of experiments were repeated using M-35 as a further galanin receptor antagonist (Fig. 2C,D), which was given intrathe- cally in the same 2 p~g dose (about 0.75 nmol). A partial antagonism to the antinociception of co-injected tramadol was observed reaching significant levels 30, 60, and 90 rain after i.t. drug administration, similar to the use of galantide (Fig. 2C). There was no antagonism detectable by i.t. M-35 to the antinociceptive effect of i.p. adminis- tered tramadol (Fig. 2D), which contrasts the results ob- tained with galantide. These results reflect those shown for M-35 in the supraspinal paw pressure test.

3.2.2. DAMGO and non-opioid drugs The effect of the intrathecally administered galanin

receptor antagonist galantide (2 p~g) on antinociception by either intrathecally or systemically administered opioid drugs was investigated in the rat paw pressure (Fig. 3). Since, in most cases, the antinociceptive effect of the drugs was of short duration, lasting 30 rain or less, data obtained 45 and 60 min after drag administration were not pre- sented here.

Co-injection of 2 p~g galantide with 1 ~g DAMGO (about 2 nmol) significantly reduced the antinociceptive effect of DAMGO by about 40%, 30 min after drug administration (Fig. 3). When co-injected with 48 ~g clonidine (about 200 nmol), galantide significantly reduced the effect of clonidine by about 22%, 15 min after drugs administration. The effect of desipramine, 6 ~g (about 20

100 A ~oo C

80

6O UJ DL

40

2O

-20

100

80

6O UJ IX.

4o

2O

0

- 2 0

80

6O UJ ¢L

40

2O

0

I I ! I I I

0 20 40 60 80 100 120

rain after i.t. injection

- 2 0

B 100

80

60 IJJ O_

40

20

- 2 0

I I I I I I

20 40 60 80 100 120

rain after i.t. injection

D

== = == == = = Z

I I I I I I I I I I I I

0 30 50 70 90 110 130 0 30 50 70 90 110 130

min after i.p. injection min after i,p. injection

Fig. 2. A,C: antinociceptive effects of i.t. injected tramadol and interaction with the galanin receptor antagonist galantide (A) or M-35 (C) in the rat tail-flick test; ( m ) tramadol 12 p~g i.t., ( D ) tramadol 12 p,g i.t. and galantide 2 ~g i.t. (A) or M-35 2 p~g i.t. (C), ( Q ) NaCI i.t., ( O ) galantide or M-35 i.t. B,D: antinociceptive effects of i.p. injected tramadol and interaction with the galanin receptor antagonist galantide (B) or M-35 (D) in the rat tail-flick test. Tramadol was administered 10 rain prior to the i.t. injection of either NaCI or one of the galanin receptor antagonists; ( m ) tramadol 4.64 m g / k g i.p + NaC1 i.t., ([:3) tramadol 4.64 m g / k g i .p .+ galantide 2 ixg i.t. (B) or M-35 2 ~tg i.t. (D), ( Q ) NaC1 i .p.+ NaCI i.t., ( O ) NaCI i .p .+ galantide 2 p~g i.t. (B) or M-35 2 p~g i.t. (D). Means _+ S.E.M. of 10 animals per group. * : Significant differences between tramadol groups in the presence or absence of galantide or M-35; p _< 0.05.

N. Seh, e et al. / Brain Research 735 (1996) 177-187 183

nmol), was significantly attenuated by 57% as measured 15 min after drug administration. The effect of fenflu- ramine, 60 ~g (about 260 nmol), was siginificantly re- duced by 72%, 15 min after drug administration.

Clonidine, 1 m g / k g i.v., elicited antinociception with a maximum of 85% MPE. Galantide, 2 Fg i.t., significantly reduced the effect of clonidine by about 40%, whereas the effect of 1 m g / k g i.p. desipramine was attenuated by

NeCl i . t . N a C l i # .

15rain 30rain 20rain 30rain

DAMGO 1 pg i.t. DAMGO 100

* not investigated systemically

0 j 15rnin 30rain 20rain 30rain

,oo C|onidine 48 pg i.t. Clonidine 1 mg/kg i.v.

°i1 lil 15min 30rain 20rain 30rain

°i] Desipramine 6 pg i.t. Desipramine 1 mg/kg i.p.

15rain $0min 20rain 30rain

Fenfluramine 60 IJg i.t,

15ml. 301nLft

Fenfluramine 30 mg/kg i,p.

20rain 30rain

°i] Dipyrone

not ac~ve up to 4 0 0 pg

Dipyrone 1000 mg/kg i,p.

lmi il 20rain 30rain

about 80% by galantide (Fig. 3), 20 min after systemic drug administration each. The antinociceptive effect of dipyrone with a maximum of 95% MPE was significantly reduced by 29 and 35%, 20 and 30 min after systemic drug administration (Fig. 3). Fenfluramine, 30 m g / k g i.p., in- duced antinociception with a maximum of 87% MPE. This effect was weakly attenuated by galantide, but the differ- ences between the i.t. saline versus i.t. galantide treated i.p. fenfluramine group did not reach significant levels (Fig. 3).

In any case of co-injected galantide, a partial antago- nism to the antinociception elicited by the different drugs was observed, reaching significance at least at one time point. When antinociception was induced by systemically administered clonidine, desipramine, or dipyrone, a partial antagonism of antinociception by i.t. galantide was also observed, this effect was significant at least at one time point.

DAMGO, clonidine, desipramine, fenfluramine, and dipyrone were also tested for antagonism to antinocicep- tion using M-35 as galanin receptor antagonist (Fig. 3) in the supraspinal rat paw pressure test.

The antinociceptive effect of intrathecally administered DAMGO, 1 p~g, was significantly attenuated by 52-69% by co-injection of M-35, 2 p~g, at any time point measured (Fig. 3). No antagonism was detectable when M-35 was co-injected to 48 p~g i.t. clonidine, 6 ~g i.t. desipramine, or 60 ~g i.t. fenfluramine (Fig. 3).

When drugs were administered systemically, antinoci- ception of fenfluramine, 30 m g / k g i.p., was significantly reduced to about 75% of its effect by i.t. M-35, 2 p~g, 20 and 30 min after drug administration. No antagonism was observed when i.t. M-35 was given to 1 m g / k g i.v. clonidine, 1 m g / k g i.p. desipramine, or 1000 m g / k g i.p. dipyrone treated animals (Fig. 3).

3.3. Binding experiments

3.3.1. Interaction o f different drugs with spinal galanin receptors

The drugs were tested for displacement of porcine [ 125I]galanin binding using crude synaptosomal membranes from rat dorsal spinal cord. Neither tramadol, nor cloni-

Fig. 3. Effect of the intrathecally administered galanin receptor antago- nists galantide (2 I.Lg) and M-35 (2 p~g) on antinociception of different intrathecally or systemically administered drugs in the rat paw pressure test with yeast-induced hyperalgesia measured 15 or 20 and 30 rain after drug administration. Open bars: antinociception elicited by the different drugs in the absence of galanin antagonists; cross-hatched bars: effect of the different drugs in the presence of galantide; hatched bars: effect of the different drugs in the presence of M-35. Means_+ S.E.M. of 10 animals per group; except of n = 20 for NaCI treated controls (without galanin antagonists), and n = 5 for desipramine plus M-35 treated groups. *" Significant differences between groups in the presence or absence of the galanin receptor antagonists; p _< 0.05.

184 N. Seh~e et al. / Brain Research 735 (1996) 177-187

dine, DAMGO, desipramine, dipyrone, or fenfluramine, 0.1 mmol/1, interfered with the [125I]galanin binding, whereas rat galanin or galantide or M-35 were active with ICs0 values in the nanomolar range (Table 2) as described previously [40].

nmol/1 for bremazocine [Table 2]. Neither tramadol, 10 ~mol /1 , nor galantide, 10 ~mol /1 , nor M-35, 10 i~mol/1, interfered with the binding of [3H]CI-977 to rat dorsal horn membranes.

3.3.2. I n t e rac t i on w i t h sp ina l t z-, o r 6-, o r K-op io id recep-

to rs

Crude membranes from rat dorsal horn were incubated with [3H]DAMGO to investigate p~-opioid receptor bind- ing. Maximal binding was 2350 dpm/assay , approxi- mately 15 f m o l / m g protein, and specific binding was 70% of total binding (n = 11). Tramadol concentration depen- dently displaced [3H]DAMGO binding with an ICs0 of 14 p, m o l / l (Table 2). As compared to tramadol, morphine showed a higher affinity to the ~-opioid receptor binding site by more than three orders of magnitude, and displaced [3H]DAMGO binding with an ICs0 of 5.4 nmol/1 (Table 2). Neither galantide nor M-35, l0 ~mol /1 , interfered with [3H]DAMGO binding.

Crude membranes from rat dorsal horn were incubated with [3H]-p-C1-DPDPE to investigate 8-opioid receptor binding. Bma x was determined as 32.1 + 7.1 f m o l / m g protein and K D as 0.63 _+ 0.198 nmol/1 (n = 5); specific binding was 53% of total binding (n = 12). Tramadol concentration dependently displaced [3H]p-C1-DPDPE binding with an ICs0 of 229.7 I~mol/1 (Table 2). Mor- phine concentration dependently displaced [3H]-p-C1- DPDPE binding with an ICs0 of 102 nmol/1 (Table 2), showing three orders of magnitude higher affinity to the dorsal horn 8-opioid receptor binding site as compared to tramadol. The galanin receptor antagonists, galantide and M-35, both used at a concentration of 10 ~M, altered the [3H]-p-C1-DPDPE binding in a different manner: weak enhancement by galantide and weak inhibition by M-35 (Table 2).

Crude membranes from rat dorsal horn were incubated with [3H]CI-977 to investigate K-opioid receptor binding. Maximal binding was 1560 dpm/assay , approximately 14 f m o l / m g protein, and specific binding was 30% of total binding (n = 12). Morphine and bremazocine concentra- tion dependently displaced [3H]CI-977 binding with an ICs0 of about 1 p~mol/1 for morphine and 0.4 _+ 0.27

4. Discussion

This is the first study to describe that drugs of various chemical and therapeutic origins induce antinociception after spinal or systemic administration obviously involv- ing, at least in part, endogenous galanin as a common link in the cascade of events necessary to propagate this antinociceptive effect.

Antinociception was measured mainly in the rat Ran- dall-Selitto test with vocalisation as nociceptive criterion, a reaction that involves higher structures in the brain (cf. [6]). Therefore, spinal effects of drugs do not only consist of effects on local reflexes, as observed in the tail-flick test, but provide evidence for a spinal modulation of ascending sensory impulses. The role of spinal endogenous galanin in antinociception was tested by local injection of two different galanin receptor antagonists, either coadmin- istered to intrathecally injected drugs or following sys- temic administration of the drugs.

Given alone, none of the galanin receptor antagonists showed any pro- or antinociceptive effect of its own, excluding a tonic effect of endogenous galanin in both test systems. However, galantide coadministered intrathecally with all the different antinociceptive drugs reduced the evoked antinociception. As regards tramadol, the in- hibitory effect by i.t. galantide was more prominent in the tail-flick test, where an almost complete inhibition oc- curred, as compared to the Randall-Selitto test, where only partial antagonism could be observed. Similar differences in the effects of galantide in both tests were reported by Reimann et al. [40] when morphine was used instead of tramadol. These observations may reflect the difference between a spinal reflex and a supraspinal reaction. In general, only partial antagonism by galantide could be observed in the Randall-Selitto test, which presumably may not be caused by kinetic effects or too low a dose used, as different time schedules and increasing doses of

Table 2 Inhibition of opioid binding to rat spinal cord membranes by tramadol or galanin receptor antagonists

Substance IC50 [nmol/1]Mean _+ S.E.M.

p. ~ K galanin

Ligand [ 3 H]DAMGO [ 3 H]p-CI-DPDPE [ 3 H]CI-977 [ 125 I]galanin

Tramadol 13,500 + 1,500 229,700 _+ 48,200 > 10 000 no inhibition at 100 o~M Galantide > 10,000 35% enhancement at 10 ~M > 10,000 2.7 _+ 0.8 M-35 > 10,000 25% inhibition at 10 p,M > 10,000 1.7 _+ 0.8 Morphine 5.4 + 1.87 102 _+ 16.6 > 100 no inhibition at 10 ~M Bremazocine not investigated not investigated 0.4 + 0.27 not investigated

N. Selve et al./Brain Research 735 (1996) 177-187 185

the galanin antagonist (2, 4 and 6 Ixg/animal, increasing volumes of 5, 7.5 and 10 Ixl had to be used; data not shown) not or only slightly enhanced the antagonistic effect. Therefore the results with spinally administered drugs suggest that besides their differences in terms of chemical structure and pharmacological mode of action their antinociceptive effects involve at least in part a galaninergic link.

Grossly similar results were obtained when another galanin receptor antagonist, M-35, was coadministered intrathecally. However, the antinociception of i.t. cloni- dine, i.t. desipramine or i.t. fenfluramine was not reduced by M-35. These differences in the efficacy of the antago- nists cannot be explained by our in vivo and in vitro results. They may be caused by the mixed agonist- antagonist properties of M-35 in contrast to the full recep- tor antagonist galantide [34].

The situation becomes more complicated when spinal administration of galanin receptor antagonists followed systemic drug administration, in order to investigate the spinal galaninergic part of the overall antinociceptive ef- fect of the drugs. Whereas antinociception caused by sys- temic administration of the drugs was partly inhibited by i.t. galantide in all cases except fenfluramine, M-35 failed to antagonize the antinociceptive effects of the drugs ex- cept fenfluramine. So in this set of experiments the results obtained with galantide were in contrast to those obtained with M-35. How can the results with systemically adminis- tered fenfluramine be explained, especially as they were not observed with i.t. fenfturamine? Serotonin release is facilitated by fenfluramine and extraneuronal serotonin acts on different 5-HT receptors. Activation of spinal 5-HT1A and 5-HT1B receptors differentially affect spinal sensory responses in the rat [31] whereby 5-HT1A mediates antinociception [52]. 5-HTIA receptors in different regions of the rat brain show a selective intramembrane interaction with galanin receptors shown autoradiographically [13,19,20] and resulting in a galanin-mediated reduction of 5-HT]A site affinity and an 8-OH-DPAT or imipramine induced enhancement of galanin receptor affinity [13,14,18-20,32]. Otherwise, brain galanin and 5-HTIA receptors are independently linked to different pools of G proteins [5] and therefore act functionally independent in spite of their anatomical neighbourhood. Systemically in- jected fenfluramine probably evokes mainly supraspinal antinociception which may involve activation of descend- ing inhibitory serotonergic pathways [12]. Then this part of its overall activity can be inhibited by the partial agonistic activity of M-35 [34] in two ways: by the reduction of 5-HT release from the terminals [14] and by the reduction of the affinity of the postsynaptic HT~A-receptors which may be located in the vicinity of these descending neurons. Locally injected fenfluramine will reach spinal 5-HT ter- minals less specifically, and therefore other than HTIA-re- ceptors may be involved in antinociception. The lack of the effect of M-35 against i.t. fenfluramine and systemi-

cally administered drugs in contrast to the full agonist M-15 [34] may be due to a lower efficacy. However, these assumptions lack direct proof and need further support.

Systemic administration of the non-opioids clonidine and fenfluramine resulted in powerful and long-lasting antinociception, whereas intrathecal administration of these drugs showed only weak effects of short duration in the Randall-Selitto test. Similar results were already described by Lund et al. [28], when i.t. desipramine or zimeldine were tested in the rat hot-plate. The observed effects were related to the supraspinal integration of the nociceptive impulses and the necessity of a supraspinal site of action for the antidepressants as analgesics. Only part of their effects is exerted at the spinal level. In the present study, the supraspinal vocalisation reaction served as the nocicep- tive criterion and therefore the same considerations may apply to the weak spinal efficacy of the non-oipiods. Prolonged spinal analgesia could be shown with the a2- adrenoceptor agonists oxymetazoline when the withdrawal reflex in the paw pressure test (in non-inflamed tissue) was used [42] and dexmedetomidine when the rat tail-flick was used [22]. Based on electrophysiological investigations following i.v. administration of dipyrone, the spinal cord was assumed to be the major site of action of dipyrone for suppression of inflammation-induced nociception [33]. Our direct approach, however, yielded no antinociceptive effi- cacy up to 400 p,g i.t. (1100 nmol). Systemically adminis- tered dipyrone, however, produced high and long-lasting antinociception which was partly inhibited by spinal galan- tide. So dipyrone can be assumed to act directly or indi- rectly at the spinal cord level, but only if systemically administered and only in conjunction with additional supraspinal activation.

Opioid receptors are involved in tramadol- or morphine-induced antinociception. The affinity of tra- madol and morphine to opioid receptor subtypes has been determined in supraspinal preparations [36]. The present study extends these investigations to the spinal cord. The absolute IC50 or K i values and the difference between the values for morphine and tramadot for Ix-, 8, and K-binding are found to be comparable to those already described in the brain [36]. Since Habert-Ortoli et al. [16] describe a 30% homology between the GALR1 (galanin receptor type 1) and the ~-opioid receptor, we tested whether galantide or M-35 interfered with 8- and, in addition, p,- and K-re- ceptor binding. Clearly, lack of any direct interaction of the galanin receptor antagonists with the Ix-ligand or K- ligand recognition site was shown. The slight enhancement of ~-binding by galantide and the displacement by M-35 is conspicuous and, in the case of the enhancement, difficult to explain; the high concentrations necessary, however, argue against pharmacological relevance of these effects. In order to look for a possible direct interaction of the drugs used for induction of antinociception, their affinity to the spinal recognition site of the galanin receptor was tested. The negative results for all drugs as presented here

186 N. Seh:e et al./Brain Research 735 (1996) 177-187

in conjunction with previously reported negative data for morphine [40] do not suggest a direct interaction. How- ever, an indirect interaction (e.g. secondary messenger interaction) cannot be excluded, and activation of galanin- ergic (inter)neurons may be caused via somatic or presy- naptic opioid, adrenergic or serotonergic receptors.

Biochemical and structural characterisation of galanin receptors in human hypothalamus and competition experi- ments of [125I]galanin binding to human hypothalamic membranes with native human, porcine, or rat galanin [26] suggested similarity of the galaninergic system between species so that functions of galanin observed in animals may be expected in humans, too. An important role of galanin in the control of the development of neuropathic pain was suggested by Verge et al. [46] since inhibition of endogenous galanin by infusion of M-35 enhanced the autotomy behaviour in rats following section of the sciatic nerve. However, it is not yet known in which place galanin is interposed in the antinociceptive pathway and how it mediates the antinociception of so different drugs as used in our study. The present work may add to strengthen the interest to evaluate the potential of galanin to alleviate acute and chronic pain states.

Acknowledgements

The authors wish to thank Ms. H. Fischer, Mr. G. Haase, Ms. W. Steffens and Mr. J. L~iufer for their skilful technical assistance.

References

[1] Bartfai, T., Fisone, F. and Langel, 0., Galanin and galanin antago- nists: molecular and biochemical perspectives, Trends Pharmacol. Sci., 13 (1992) 312-317.

[2] Bartfai, T., H~Skfelt, T. and Langel, U., Galanin - a neuroendocrine peptide, Crit. Ret,. Neurobiol., 7 (1993) 229-274.

[3] Bartfai, T., Langel, U., Bedecs, K., Andell, S., Land, T., Gregersen, S., Ahrdn, B., Girotti, P., Consolo, S., Corwin, R., Crawley, J., Xu, X., Wiesenfeld-Hallin, Z. and H6kfelt, T., Galanin-receptor ligand M40 peptide distinguishes between putative galanin-receptor sub- types, Proc. Natl. Acad. Sci. USA, 90 (1993) 11287-11291.

[4] Bedecs, K., Langel, 0. , Barfai, T. and Wiesenfeld-Hallin, Z., Galanin receptors and their second messengers in the lumbar dorsal spinal cord, Acta Physiol. Scand., 144 (1992) 213-220.

[5] Billecocq, A., Hedlund, P.B., Bolanos-Jim~nez, F. and Fillion, G., Characterization of galanin and 5-HTtA receptor coupling to adeny- late cyclase in discrete regions of the rat brain, Eur. J. Pharmacol. Mol. Pharmacol., 269 (1994) 209-217.

[6] Chapman, C.R., Casey, K.L. Dubner, R., Foley, K.M., Gracely, R.H. and Reading, A.E., Pain measurement: an overview, Pain, 22 (1985) 1-31.

[7] Chen, A., Fournier, A., Couvineau, A., Laburthe, M. and Amiranoff, B., Purification of a galanin receptor from pig brain, Proc. Natl. Acad. Sci. USA, 90 (1993) 3845-3849.

[8] Cridland, R.A. and Henry, J.L., Effects of intrathecal administration of neuropeptides on a spinal nociceptive reflex in the rat: VIP,

galanin, CGRP, TRH, somatostatin and angiotensin II, Neuropep- tides, 11 (1988) 23-32.

[9] D'Amour, F.E. and Smith, D.L., A method for determining loss of pain sensation, J. Pharmacnl. Exp. Ther., 72 (1941) 74-78.

[10] Driessen, B. and Reimann, W., Interaction of the central analgesic, tramadol, with the uptake and release of 5-hydroxytryptamine in the rat brain in vitro, Br. J. Pharmacol., 105 (1992) 147-151.

[11] Ekblad, E., RiSkaeus, ,~., H~kanson, R. and Sundler, F., Galanin nerve fibers in the rat gut: distribution, origin and projections, Neuroscience, 16 (1985) 355-363.

[12] Feinmann, C., Pain relief by antidepressants: possible modes of actions, Pain, 23 (1985) 1-8.

[13] Fuxe, K., yon Euler, G., Agnati, L.F. and Ogren, S.-O., Galanin selectivity modulates 5-hydroxytryptamine IA receptors in the rat ventral limbic cortex, Neurosci. Lett., 85 (1988) 163-167.

[14] Fuxe, K., Ogren, S.-O., Jansson, A., Cintra, A., H~irfstrand, A. and Agnata, L.F., lntraventricular injections of galanin reduces 5-HT metabolism in the ventral limbic cortex, the hippocampal formation and the fronto-parietal cortex of the male rat, Acta Physiol. Scand., 133 (1988) 579-581.

[15] Gilfoil, T.M., Klavins, I. and Grummbach, L., Effects of acetylsali- cylic acid on the edema and hyperesthesia of the experimentally inflamed rat's paw, J. Pharmacol. Exp. Ther., 142 (1963) 1-5.

[16] Habert-Ortoli, E., Amiranoff, B., Loquet, I., Laburthe, M. and Mayaux, J.-F. Molecular cloning of a functional human galanin receptor Proe. Natl. Acad. Sci. USA, 91 (1994) 9780-9783.

[17] Handa, B.K., Lane, A.C., Lord, J.A.H., Morgan, B.A., Rance, M.J. and Smith, C.F.C., Analogues of [3-LPH61 64 prossessing selective agonist activity at ix-opiate receptors, Eur. J. Pharmacol., 70 (1981) 531-540.

[18] Hedlund, P. and Fuxe, K., Chronic imipramine treatment increases the affinity of [125I]galanin binding sites in the tel- and diencephalon of the rat and alters the 5-HTIA /galanin receptor interaction, Acta Physiol. Scand., 141 (1991) 137-138.

[19] Hedlund, P.B., Aguirre, J.A., Narvaez, J.A. and Fuxe, K., Centrally coinjected galanin and a 5-HTaA agonist act synergistically to pro- duce vasodepressor responses in the rat, Eur. J. Pharmacol., 204 (1991a) 87-95.

[20] Hedlund, P., yon Euler, G. and Fuxe, K., Activation of 5-hydroxy- tryptaminelA receptors increases the affinity of galanin receptors in di- and telencephalic areas of the rat, Brain Res., 560 (1991b) 251 259.

[21] Hylden, J.L.K. and Wilcox, G.L. lntrathecal morphine in mice: a new technique, Eur. J. Phal~acol, 67 (1980) 313-316.

[22] Horvfith, G., Kovfics, M., Szikszay, M. and Benedek, G., Mydriatic and antinociceptive effects of intrathecal dexmedetomidine in con- scious rats, Eur. J. Pharmacol., 253 (1994) 61 66.

[23] Ju, G., HiSkfelt, T., Brodin, E., Fahrenkrug, J., Fischer, J.A., Frey, P., Elde, R.P. and Brown, J.C., Primary sensory neurons of the rat showing calcitonin gene-related peptide immunoreactivity and their relation to substance P-, somatostatin-, galanin-, vasoactive intestinal polypetide- and cholecystokinin-immunoreactive ganglion cells, Cell Tissue Res., 47 (1987) 417-431.

[24] Kuraishi, Y., Kawabata, S., Matsumoto, T., Nakamura, A., Fujita, H. and Satoh, M., involvement of substance P. in hyperalgesia induced by intrathecal galanin, Neurosci. Res. 11 (1991) 276 285.

[25] Lagny-Pourmir, I., Lorinet, A.M., Yanaihara, N. and Laburthe, M., Structural requirements for galanin interaction with receptors from pancreatic beta cells and from brain tissue of the rat, Peptides, 10 (1989) 757-761.

[26] Lorinet, A.-M., Javoy-Agid, F., Laburthe, M. and Amiranoff, B., Galanin receptors in human hypothalamus: biochemical and struc- tural analysis, Eur. J. Pharmacol. - Mol. Pharmacol., 269 (1994) 59 64.

[27] Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J., Protein measurement with the Folin phenol reagent, J. Biol. Chem., 193 (1951) 265-275.

N. SelL,e et a l . /Brain Research 735 (1996) 177-187 187

[28] Lund, A., Tj¢lsen, A. and Hole, K., Antinociceptive effect of intrathecally administered desipramine and zimeldine in rats, Neu- ropharmacology, 9 (1990) 819-823.

[29] Lundeberg, T., Meiter, B., BjiSrkstrand, E. and Uvn~is-Moberg, K., Oxytocin modulates the effects of galanin in carrageenan-induced hyperalgesia in rats, Brain Res., 608 (1993) 181-185.

[30] Melander, T., HSkfelt, T., R~ikaeus, ,~., Fahrenkrug, K. and Mutt, V., Distribution of galanin-like immunoreactivity in the gastro-in- testinal tract of several mammalian species, Cell Tissue Res. 239 (1985) 253-270.

[31] Murphy, R.M. and Zemlan, F.P., Selective serotoninlA/1B agonists differentially affect spinal nociceptive reflexes, Neuropharmacol- ogy, 29 (1990) 463-468.

[32] Narvfiez, J.A., Diaz, Z., Aguirre, J.A., Conz~.lez-Bfiron, S., Yanai- hara, N., Fuxe, K. and Hedlund, P.B., Intracisternally injected galanin-(l-15) modulates the cardiovascular responses of galanin- (1-29) and the 5-HTjA receptor agonist 8-OH-DPAT, Eur. J. Pharmacol., 257 (1994) 257-265.

[33] Neugebauer, V., Schaible, H.-G., He, X., Li.icke, T., Giindling, P. and Schmidt, R.F., Electrophysiological evidence for a spinal antinociceptive action of dipyrone, Agents Actions, 41 (1994) 62-70.

[34] Ogren, S.O., Pramanik, A., Land, T. and Langel, U., Differential effects of the putative galanin receptor antagonists M15 and M35 on striatal acetylcholine release, Eur. J. Pharmacol., 242 (1993) 59-64.

[35] Post, C., Alari, L. and Htikfelt, T., Intrathecal galanin increases the latency in the tail-flick and hot-plate tests in mouse, Acta Physiol. Scand., 132 (1988) 409-419.

[36] Raffa, R.B., Friderichs, E., Reimann, W., Shank, R.P., Codd, E.E. and Vaught, J.L., Opioid and nonopioid components independently contribute to the mechanism of action of tramadol, an 'atypical' opioid analgesic, J. Pharmacol. Exp. Ther., 260 (1992) 275-285.

[37] Randall, L.O. and Selitto, J.J., A method for measurement of analgesic activity on inflamed tissue, Arch. Int. Pharmacodyn. Ther., 111 (1957) 409-419.

[38] Reimann, W. and Schneider, F., Presynaptic et 2-adrenoceptors mod- ulate the release of [3H]noradrenaline from rat spinal cord dorsal horn neurones, Eur. J. Pharmacol., 167 (1989) 161-166.

[39] Reimann, W. and Hennies, H.-H., Inhibition of spinal noradrenaline uptake in rats by the centrally acting analgesic tramadol, Biochem. Pharmacol., 47 (1994) 2289-2293.

[40] Reimann, W., Englberger, W., Friderichs, E., Selve, N. and Wilffert, B., Spinal antinociception by morphine in rats is antagonised by galanin receptor antagonists, Naunyn-Schmiedeberg's Arch. Phar- macol., 350 (1994) 380-386.

[41] Satoh, M., Kuraishi, Y. and Kawamura, M., Effects of intrathecal antibodies to substance P, calcitonin gene-related peptide and galanin on repeated cold stress-induced hyperalgesia: comparison with car- rageenan-induced hyperalgesia, Pain, 49 (1992) 273-278.

[42] Sherman, S., Loomis, C., Milne, B. and Cervenko, F., Prolonged spinal analgesia in the rat with the ot-adrenoceptor agonist oxymeta- zoline, Eur. J. Pharmacol., 140 (1987) 25-32.

[43] Shimada, S.G., Otterness, I.G. and Stitt, J.T., A study of the mechanism of action of the mild analgesic dipyrone, Agents Ac- tions, 41 (1994) 188-192.

[44] Skofitsch, G. and Jacobowitz, D.M., Immunohistochemical mapping of galanin-like neurons in the rat central nervous system, Peptides, 6 (1985) 509-546.

[45] Tatemoto, K., RiSkaeus, A., J~Smvall, H., McDonald, T.J. and Mutt, V., Galanin - - a novel biologically active peptide from porcine intestine, FEBS Lett., 164 (1983) 124-128.

[46] Verge, V.M.K., Xu, X.-J., Langel, U., HiSkfelt, T., Wiesenfeld-Hal- lin, Z. and Bartfai, T., Evidence for endogenous inhibition of autotomy by galanin in the rat after sciatic nerve section: demon- strated by chronic intrathecal infusion of a high affinity galanin receptor antagonist, Neurosci. Lett., 149 (1993) 193-197.

[47] Wiesenfeld-Hallin, Z., Xu, X.-J., Hughes, J., Horwell, D.C. and H~Skfelt, T., PD134308, a selective antagonist of cholecystokinin type B receptor, enhances the analgesic effect of morphine and synergistically interacts with intrathecal galanin to depress spinal nociceptive reflexes, Proc. Natl. Acad. Sci. USA, 87 (1990) 7105- 7109.

[48] Wiesenfeld-Hallin, Z., Batfai, T. and H~Skfelt, T., Galanin in sensory neurons in the spinal cord, Front. Neuroendocrinol., 4 (1992) 319-343.

[49] Wiesenfeld-Hallin, Z., Xu, X.-J., Langel, U., Bedecs, K., HtSkfelt, T. and Bartfai, T., Galanin-mediated control of pain: enhanced role after nerve injury, Proc. Natl. Acad. Sci. USA, 89 (1992) 3334-3337.

[50] Wiesenfeld-Hallin, Z., Xu, X.-J., Hao, J.-X. and H~Skfelt, T., The behavioural effects of intrathecal galanin on tests of thermal and mechanical nociception in the rat, Acta Physiol. Scand., 147 (1993) 457-458.

[51] Xu, X.-J., Wiesenfeld-Hallin, Z., Fisone, G., Bartfai, T. and Htikfelt, T. The N-terminal 1-16, but not C-terminal 17-29, galanin frag- ment affects the flexor reflex in rats, Eur. J. Pharmacol., 182 (1990) 137-141.

[52] Xu, W., Qiu, X.C. and Han, J.S. Serotonin receptor subtype in spinal antinociception in the rat, J. Pharmacol. Exp. Ther., 269 (1994) 1182-1189.