Brain Research, 521 (1990) 167-174 Elsevier

167

BRES 15659

Behavioural modification of bulbospinal serotonergic inhibition and morphine analgesia Richard J. Milne and Gregory D. Gamble Department of Physiology, University of Auckland, Auckland (New Zealand) (Accepted 9 January 1990)

Key words: Morphine; Dose-response relationship; Serotonin; 5-Hydroxytryptophan; Parachlorophenylalanine; Analgesia

Habituation to the stress of sham nociceptive testing enhances a rat's sensitivity to noxious thermal stimuli and reduces the antinociceptive effect of a subsequent acute dose of morphine. Since serotonin (5-hydroxytryptamine, 5-HT) mediates stress responses, experiments were designed to elucidate the role of 5-HT in these phenomena. Intrathecal methysergide or 5,7-dihydroxytryptamine (5,7-DHT) reduced baseline tail-flick latencies of novice rats to those of habituated animals. Morphine dose-response relationships were fitted to a 4 parameter sigmoidal function. Baseline latencies of novice animals were increased by 5-hydroxytryptophan (5-HTP) and reduced by parachlorophenylalanine (PCPA) in both reflex tests and in the hot-plate test, but latencies of habituated animals were unchanged by either treatment. In both reflex tests, the maximum effect due to morphine was increased by 5-HTP and reduced by PCPA in novice but not in habituated animals. We conclude that the serotonergic component of morphine's bulbospinal action represents the stress of the testing environment rather than an essential part of morphine's action. INTRODUCTION Morphine produces analgesia by acting at both spinal and supraspinal sites. There is extensive evidence for a contributory role of serotonin (5-hydroxytryptamine, 5-HT) in the expression of morphine analgesia. Several lines of evidence have implicated descending bulbospinal pathways. In summary: midline brainstem raphe nuclei containing 5-HT and other neurotransmitters send massive projections to the spinal cord, partly via the dorsolateral funiculi; lesions of the dorsolateral funiculi attenuate morphine analgesia and reduce spinal monoamine levels; lesions of raphe nuclei reduce morphine analgesia; electrical stimulation of the rostral ventral medulla produces analgesia which has similar features to that produced by morphine injected at the same sites; antinociception induced by systemically administered morphine can be enhanced by systemic 5-HT precursors and intrathecally injected reuptake blockers, and blocked by systemic or spinal depletion or antagonism of 5-HT; 5-HT is released at spinal level after systemic administration of morphine; morphine enhances spinal 5-HT metabolism; morphine injected into nucleus raphe magnus or periaqueductal grey increases spinal cord 5-HT turnover and produces analgesia which can be antagonised by intrathecal 5-HT antagonists 2,3,5,11,13,26. Recently it has been proposed that the serotonergic

component of morphine's analgesic action may be related to stressors such as environmental novelty and/or physical restraint that are generally present during nociceptive testing 15-17'23. The evidence suggest that the serotonergic bulbospinal contribution to morphine analgesia may be expressed in animals that are stressed by restraint or by environmental novelty, but not in non-restrained animals which have been thoroughly habituated to the testing environment. However despite controversy surrounding the role of 5-HT in morphine analgesia, few authors have reported in any detail the degree of physical restraint used during nociceptive testing or the extent to which the animals have been habituated to the testing environment prior to testing. In preceding papers we have reported that repeated exposure of rats to the testing environment and procedures (habituation) reduces baseline withdrawal reflex latencies to those of spinal animals and substantially reduces the maximum effect attributable to morphine (Emax) 12'21. We have called this effect behavioural tolerance. Serotonin seems to be involved in the change in baseline latencies, since systemic depletion of 5-HT with parachlorophenylalanine (PCPA) reduced 50 °C hotplate baseline latencies substantially in novice but not habituated animals. Taken together, these findings suggest that habituation attenuates the serotonergic component of bulbospinal tonic descending inhibition. Since

Correspondence: R.J. Milne, Department of Physiology, University of Auckland, School of Medicine, Private Bag, Auckland, New Zealand. 0006-8993/90/$03.50 (~) 1990 Elsevier Science Publishers B.V. (Biomedical Division)

168 E , .... a n d E o a r e b o t h r e d u c e d by h a b i t u a t i o n it t h e r e f o r e

Manipulation of 5-HT

s e e m e d likely t h a t h a b i t u a t i o n w o u l d m o d i f y E . . . . via a

PCPA (350 mg/kg in sterile maize oil; Sigma) was administered daily for 3 days to deplete 5-HT levels selectively~s and to reduce metabolism of 5-HT in the dorsal spinal cord ~. 5-Hydroxytryptophan (5-HTP, Sigma, 80 mg/kg) was administered i.p. 1 h before the experiment to augment 5-HT metabolism and release.

b u l b o s p i n a l s e r o t o n e r g i c link. T h e e x p e r i m e n t s d e s c r i b e d in this p a p e r w e r e d e s i g n e d t o t e s t this h y p o t h e s i s by m a n i p u l a t i n g 5 - H T levels.

Spinalisation MATERIALS AND METHODS All experiments were performed on female George River Wistar rats (200-250 g). Animals were housed 8 to a cage in an animal facility until the commencement of the experiment and either used immediately in experiments ('novice' animals) or housed in the laboratory environment in smaller cages, 2-4 to a cage, and habituated to the testing procedures as described below.

Habituation procedures Animals were deliberately habituated to the experimental environment by a range of procedures over a period of at least 5 days as previously described21. Briefly, the animals were extensively handled by the experimenter and were not exposed to other persons or to sudden changes of temperature, extraneous noises or foreign odours; animals were allowed to escape from the experimenter's hand or the test equipment at will; and all animals were sham tested daily by dipping the tip of the tail and each hindpaw in turn into warm water. Previous experience had shown that repeated testing at 5 min intervals over a 25 min period does not change baseline latencies ~2. Some of the habituated animals had been used in earlier experiments which did not involve drugs or surgery.

Hot-plate test The assays used were the commonly used (55 °C) hot-plate test as well as the more sensitive (50 °C) hot-plate testel'ze to assay a supraspinally organised response. Licking one or other hind paw was used as the endpoint. A cut-off time of 20 s (55 °C hot-plate) or 30 s (50 °C hot-plate) was used to avoid thermal trauma to the paws. Hot-plate latencies were measured immediately prior to and at 15 min intervals subsequent to injection of morphine.

Reflex tests Nociceptive reflexes were assayed by dipping each hindpaw8'~2 and the tip of the tail 36 in turn into circulating water thermostatically controlled at 49 °C~2. The animal was hand held to minimise stress. Hot water rather than radiant heat was used in the tail-flick assay to provide a known and fully repeatable noxious stimulus. The endpoint was taken as sudden flexion of the hind limb or strong flick of the tail. A cut-off time of 30 s was used to avoid trauma to the paws or tail. The tail or paw was dried between tests. Reflex latencies were measured immediately prior to and at 15 min intervals subsequent to injection of morphine.

Intrathecal drug administration Under Nembutal anaesthesia a 7.5 cm length of fine polyethylene tubing was inserted through the atlanto-oecipital membrane in the dorsal subarachnoid space, with the tip placed just rostral to the lumbar enlargement (confirmed postmortem with dye injection)34. Animals were left 14 days to recover and a small number of animals which exhibited movement disorders were sacrificed immediately. Drugs were applied in a 10/A bolus of vehicle. Methysergide was given as 30/~g base in saline; 5,7-dihydroxytryptamine as 20/~g base with 20 mg/mi ascorbic acid in saline, following pretreatment 1 h earlier with 25 mg/kg s.c. desmethylimipramine to protect against norepinephrine depletion27.

Morphine administration Each animal received 2 of 6 doses (1, 2, 4, 6, 8 or 12 mg/kg calculated with respect to the free base). The drug was injected i.p. as a 0.5 ml bolus in saline on consecutive days, in randomised order.

In one group of animals the spinal cord was transected at the T~ ~ level under halothane anaesthesia. Care was taken to minimise bleeding by cautery. A 0.05 ml drop of 1% lignocaine solution with adrenaline (Astral was applied topically to the exposed dorsal surface of the cord, the head was ventroflexed to place the spinal cord under tension and the cord was crushed with jewellers' forceps, The cord spontaneously retracted 2-3 mm and the bony cavity was packed with absorbable gelatin sponge. Postmortem examination of most of the animals confirmed that no tissue remained in the spinal canal at that level. Animals in which reflexes did not return within 1 h were discarded from the study. To avoid overinflation of the urinary bladder, the water supply was withdrawn overnight and the bladder was expressed regularly. Autotomy was never observed. Other animals were sham spinalised, by performing a laminectomy and leaving the dura intact, to control for the effects of anaesthesia and surgery. Nociceptive reflexes were assayed in spinal and sham the spinal animals 1-4 days after spinalisation. This time interval is short enough to avoid changes in spinal 5-HT receptor sensitivity4.

Measurement of tail temperatures In experiments using intrathecal drugs, the temperature of the tail was measured using copper-constantan thermocouples (1.0 × 1.0 ram) taped securely to the ventral surface at a point located about 2 cm from the end of the tail. The outputs of the 3 locally built thermocouple amplifiers were directed via an analog to digital converter to a microprocessor and analysed online using software developed in the ASYST environment (MacMillan). Temperatures were sampled at 0.6 min intervals. Tail-flick measurements were made only when a particular temperature had remained stable for about 3 min.

Morphine dose-response analysis Data were fitted with a 4 parameter model sigmoid Ema x model as described previously21:

E(D)

=

E o +

Emax*DN EDs. N + D N

where E(D) is the integrated response after a morphine dose D, E0 is the integrated response in the absence of morphine, Ema~ is the asymptotic maximum integrated effect attributable to morphine, EDs0 is the dose producing a response of (E0+0.5* Emax), and N is a steepness factor (Hill coefficient). In order to evaluate the effect of the habituation procedures on the morphine dose-response relationship, animals were paired within tests as novice and habituated 21. 12 pairs of animals were randomly selected for the purpose of the data analysis. The dose-response relationships for each pair of animals were then modeled using a combined sigmoid Ema x model 21. Data were excluded from the analysis when the standard error of any estimated parameter was greater than 60% of the estimate. Parameter estimation was performed on an IBM 4341 mainframe computer using a non-linear least squares regression procedure (NLIN, SAS version 5) with the multivariate secant iterative method DUD. The variance of the response measurement was assumed to be the same at every dose level for both animals. The goodness of fit for each model was assessed using the Schwartz information criterion28 and differences between pairs were tested using the Wilcoxon signed pairs test 14.

Statistical analysis Time-effect relationships were analysed using a repeated mea-

169 sures ANOVA with corrections for deviations from sphericity where appropriate. Significant main and interaction effects were examined with Tukey's HSD test 14. Multiple a priori paired comparisons of computed parameters were made using multiple Student's t-test with the level of significance preserved through Dunn's procedure .4. A significance level of 5% was adopted throughout the analyses. All experiments were approved by the Animal Ethics Committee of the Auckland School of Medicine.

RESULTS

Intrathecal drugs If habituation attenuates the serotonergic component of tonic descending inhibition then we would predict that pharmacological removal of this component would reduce reflex latencies in novice animals but have less effect in habituated animals. 80 rats were implanted with intrathecal catheters and divided into 2 groups, one of which was habituated to the laboratory environment and procedures and the other remained in the animal care facility for the same period of time. Within each group of 40 animals, 4 subgroups of 10 animals received intrathecal applications of either methysergide (30 #g), 5, 7 - D H T or their respective vehicles. Tail-flick latencies (mean of 5 tests at 5 min intervals for each animal) of each of the 8 groups are illustrated in Fig. 1. The results of these experiments showed that preventing the release or blocking the subsequent action of 5-HT at spinal level reduced the tail-flick latency of novice animals to that of habituated animals, whereas neither drug treatment had any effect on latencies of habituated animals. Tail temperatures measured concurrently did not differ between any of the treatment or control groups, implying that the results were not due to changes in regional vasomotor tone [ANOVA; F4,56 = 0.54; P > 0.05]. Fig. 2 shows that for novice animals, morphine time-effect relationships, as assessed by tail-flick and leg-flexion tests, were clearly more prolonged in the

10

TAIL FLICK (49"C)

[ ] METHYSERGIDE(n=50) [ ] 5,7-DHT + DMI (m=SO) []

8 >(...) Z

• CONTROL(n=50)

CONTROL ( n = 5 0 )

6 4

.J

2

0

NOVICE

HABITUATED

Fig. 1. Effects of intrathecal application of methysergide, 5,7-DHT or the respective vehicles upon tail-flick latencies of novice and habituated animals (mean + S.E.M., 10 animals per group).

presence of 5-HTP compared to PCPA. In contrast, neither drug treatment affected habituated animals [ANOVA tail-flick, 5-HTP: F30,447 = 21.3; leg flexion, 5-HTP: /730,447 "~- 15.1; tail-flick, PCPA, F3o.447 = 13.8; leg flexion, PCPA: F30,447 = 16.3; P < 0.05]. To control for possible peripheral effects of PCPA we measured tail-flick latencies before and after PCPA administration in 7 acutely spinalised animals. In these animals, tail-flick latencies were unaffected by PCPA [t(138,0.05) = 0.43; P > 0.05; not illustrated]. Fig. 3 shows dose-response relationships computed for novice and habituated animals treated with either PCPA or 5-HTP, along with previously reported data obtained using untreated novice and habituated animals 21. Qualitatively, it is clear that manipulation of 5-HT levels with 5-HTP and PCPA had opposite effects in novice animals and that neither drug treatment had much effect upon morphine action in habituated animals. Quantitative comparisons with untreated animals are shown below (see Fig. 6 and Table I).

Hot-plate testing In another set of experiments 18 novice animals were randomly assigned to 2 equal groups which were treated with PCPA or 5-HTP. A further 18 animals were first habituated to the experimental environment and subsequently divided into 2 equal groups which were treated with PCPA or 5-HTP. Fig. 4 shows that PCPA and 5-HTP had opposite effects upon morphine time-effect relationships in novices but had little effect in habituated animals [ANOVA 5-HTP: F3o,987 = 5.64; P < 0.05; PCPA: F30,987 = 22.9, P < 0.05]. Prior experiments had shown that the vehicles used had no effect upon baseline latencies or responses to morphine. Fig. 5 shows dose-response relationships computed from integrated time-effect data illustrated in Fig. 4, along with previously reported data from untreated novice and habituated animals 21. The pattern of drug effects is similar to that found using reflex tests, in that depletion and enhancement of 5-HTP moved the curves in opposite directions.

Comparisons of model parameters To establish comparisons with untreated animals we have utilised previously reported data obtained under comparable conditions 2~. Fig. 6 shows computed model parameters for the treated and untreated animals and Table I summarises the statistical analysis. For novice animals 5-HTP increased E o and PCPA reduced E o as assessed in all tests. Similar patterns were found for the p a r a m e t e r Emax as assessed by both reflex tests. We have not included Ema. values for the hot-plate test because the 4 parameter model fitted this data poorly, especially

170

PCPA

Tail flick

30

ul v

-Habituated

Novice

20

>-,

¢1

J

01

I

i

•

I

i

•

Leg f l e x i o n Novice

3 0 1 Habituated

2oI 10

l

l, / ./.,T-..

O[

~,

t

[

•

5-HTP

I

I

•

Tail flick

30 " H a b i t u a t e d

Nov ice,~''~"--"~C'~.".I-~.. . . .

'

Z

2O

j

1o

;"

-

.'1// OI

-, ~ "