Introduction

Top Abstract Introduction Methods Results Discussion References

M6G is an analgesically active metabolite of morphine in humans (Hand et al., 1987) accounting for up to 10% of a systemically administered morphine dose (Yeh et al., 1977). M6G has been reported to be more potent than morphine as an antinociceptive agent in experimental animals, with the reported "apparent" potency being determined by factors such as route of administration, the species/strain of animal receiving M6G and the method of antinociceptive testing employed (Yoshimura et al., 1973; Pasternak et al., 1987; Abbott and Palmour, 1988; Paul et al., 1989; Sullivan et al., 1989). For example, following M6G administration to Sprague-Dawley (SD) rats by the intracerebroventricular (i.c.v.) route, M6G was reported to be 46 times more potent than morphine in the acid writhing test but 360 times more potent than morphine in the tail flick test (Frances et al., 1992). Furthermore, when administered by the i.t. and i.c.v. routes to mice, M6G was reported to be approximately 650 and 90 times the potency of morphine respectively, when assessed using the predominantly spinally mediated tail flick latency test (Paul et al., 1989). However, the available data are such that following systemic administration the comparative "apparent" potencies determined using the tail flick and hotplate latency tests, are difficult to assess.

The permeability of the BBB is very similar at the levels of the brain and the spinal cord in rats (Azzi et al., 1990; Smith and Shine, 1992). Thus, any differences in the "apparent" potency of systemically administered M6G using the tail flick latency test (principally a spinal reflex) compared with the hotplate latency test (predominantly involves supraspinal structures) would not be due to physiological differences in the integrity of the BBB of the spinal cord relative to that of the brain. As M6G appears to be 6-8 fold more potent relative to morphine when administered intrathecally than when both drugs are given by the i.c.v. route (Paul et al., 1989), this suggests that following systemic administration, M6G's antinociceptive effects would be mediated primarily by spinal rather than supraspinal mechanisms. However, there are currently no studies in the literature that have investigated this issue. There is also a lack of published information regarding the time-course of the antinociceptive effects evoked by i.c.v. M6G when assayed using the hotplate latency test. Therefore, this study was designed to evaluate the "apparent" levels of antinociception achieved in adult male Sprague-Dawley rats using both the tail flick and hotplate latency tests following individual single-dose administration of (i) M6G and (ii) morphine, by the intraperitoneal or the intravenous routes and (iii) M6G by the i.c.v. route.

	Methods

Top Abstract Introduction Methods Results Discussion References

Experimental animals. Ethical approval for this study was obtained from the Animal Experimentation Ethics Committee of The University of Queensland. Adult male Sprague-Dawley (SD) rats with a mean (± S.E.M.) weight of 305 (± 30) g for the i.p. and i.v. studies and 235 (± 3) g for the i.c.v. studies, were purchased from The University of Queensland Medical School Animal Breeding Facility. Rats were housed in a temperature controlled room (21°C ± 2°C) with a 12/12 hr light/dark cycle, with food and water available ad libitum.

Experimental procedures for i.p. and i.v. studies. Adult male SD rats underwent jugular vein cannulation whilst under 3% isoflurane: 97% oxygen general inhalational anaesthesia, using a calibrated Trilene vaporizer. The cannula was externalized through a subcutaneous tunnel to the back of the neck and was protected by a stainless steel spring. Following surgery, rats were allowed to recover overnight before experiments were initiated.

Experimental procedure for i.c.v. studies. Adult male SD rats were anaesthetized with a mixture of ketamine (20 mg) and xylazine (1.6 mg) administered by the i.p. route. A stainless steel guide cannula (21 G) was inserted stereotaxically to a depth of 3.5 mm (1 mm above the left lateral ventricle) in a position 1.5 mm to the left and 0.8 mm caudal to Bregma, according to the stereotaxic co-ordinates of Paxinos and Watson (1986). The cannula was secured in position with dental acrylic and the animal was kept warm during recovery. Rats were given a five day recovery period prior to i.c.v. drug injection. Rats were lightly anaesthetized with (50% :50%) CO₂:O₂ and drugs or saline (1 µL) were administered using a 10 µL Hamilton syringe. Initially, rats received either a low dose of an opioid other than morphine or M6G, or Angiotensin II (100 pmol/µL) as a preliminary check for cannula placement. Those rats that did not achieve an antinociceptive response (assessed using the tail flick latency test) after administration of the opioid or did not drink  5 mL of water within 10 min of i.c.v. Angiotensin II administration, were excluded from the study. Several days later, i.c.v. M6G (0.22 (n = 4), 1.25 (n = 5), 2.2 (n = 6) or 3.3 (n = 5) nmol) was administered and the degree of antinociception was assessed over a 3 hr experimental period. After completion of the experiment, malachite green dye (1 µL) was injected by the i.c.v. route to check for correct cannula placement. A lack of dye diffusion throughout the ventricular circulation of the brain or clumping of dye in the periventricular tissue indicated poor cannula placement, thereby precluding the inclusion of the data from such animals in the study. All doses of M6G administered to rats in this study were verified using high-performance liquid chromatography (HPLC) with electrochemical detection (Wright et al., 1994).

Antinociceptive testing. The degree of antinociception achieved was quantified using two methods (i) the hotplate latency test (Eddy and Leimbach, 1953), and (ii) the tail flick latency test (D'Amour and Smith, 1941). For the tail flick latency test, rats were placed inside restraining cages at least 5 min before tail flick latency determination. A constant heat intensity was applied to the dorsum of the lower third of the rat's tail and when the rat flicked its tail in response to the noxious thermal stimulus, both the heat source and the timer were stopped automatically. A maximum tail flick latency of 9 sec was permitted to minimize tissue damage to the rat's tail. For the hotplate latency test, a rectangular metal surface was heated to a temperature of 55 ± 0.5°C (Model 39D Analgesia meter IITC, Life science, California). The antinociceptive response was the latency observed from the time the rat was placed on the heated surface until the first overt behavioral sign of nociception such as (i) the rat licking a hind paw, (ii) vocalization, or (iii) an escape response. The timer was stopped by a foot-operated pedal and the rat was immediately removed from the hotplate. A maximum hotplate latency of 30 sec was used to prevent tissue damage to the rat's paws. Pre-dosing latencies were determined on at least three occasions (5 min apart with the three measurements being within ± 0.5 and ± 1 sec respectively for the tail flick and hotplate methods) before the administration of drugs or heparinised saline. Pre-dosing base-line latencies were 3-4 sec for the tail flick and hotplate latency tests. Antinociceptive testing was performed at 5, 15, 30, 45, 60, 90, 120 and 180 min after administration of M6G, morphine or saline (control rats). Hotplate latencies were determined approximately 30 sec after the tail flick latencies. To evaluate the possibility of sensitization of rats to the second thermal stimulus (i.e. whether hotplate values were altered by prior testing using the tail flick apparatus), the levels of antinociception observed in an additional group of rats (n = 3) that received i.v. morphine (10 µmol) were determined using the hotplate latency test only (data not shown).

Statistical analyses. The areas under the degree of antinociception vs. time curves (AUC's) were calculated using trapezoidal integration (table 1). All comparisons of the AUC data between experimental groups were performed using the Wilcoxon Rank-Sum test as implemented in the Minitab statistical analysis program, with a statistical significance criterion of P < .05. 

Materials. Polyethylene tubing (OD 1 mm, ID 0.5 mm) was purchased from Dural Plastics and Engineering Pty Ltd, (Sydney, Australia) and silastic tubing from Auburn Plastics and Engineering (Sydney, Australia). Dysilk Black Braided Siliconised Silk sutures were purchased from Dynek Pty Ltd, (Adelaide, South Australia).

Drugs. Isoflurane (Forthane^®) was purchased from Abbott Australasia, Pty Ltd (Sydney, Australia). Ketamine hydrochloride vials (100 mg/mL) were purchased from Parnell Laboratories Australia, Pty Ltd, (Sydney, Australia) and xylazine hydrochloride (Ilium Xylazil-20^®) vials (20 mg/mL) were purchased from Troy Laboratories Pty Ltd, (Sydney, Australia). Morphine sulphate ampoules (30 mg/mL) were purchased from David Bull Laboratories (Melbourne, Australia), and morphine-6-glucuronide (M6G) was purchased from Ultrafine Chemicals (Salford, UK). For the i.v. and i.p. studies, both morphine and M6G were diluted to the required concentrations with heparinised saline (1 I.U. mL¹) which was purchased from Astra Pharmaceuticals Pty Ltd, (Sydney, Australia). For the i.c.v. studies, M6G was diluted to the required concentrations with sterile normal saline purchased from Delta West Pty Ltd, (Perth, Australia). Angiotensin II was purchased from Sigma Chemical Company (Maryland, U.S.A).

	Results

Top Abstract Introduction Methods Results Discussion References

The time of onset of antinociception following systemic single-dose administration of both morphine and M6G was faster for i.v. than for i.p. dosing (compare figs. 1 and 2). This observation is consistent with the fact that following i.p. dosing, drugs such as morphine will undergo significant presystemic metabolism in the liver (Lukas et al., 1971). Importantly, the mean (± S.E.M.) area under the degree of antinociception vs. time curve (AUC) determined using the hotplate latency test (134 ± 18% MPE.hr) immediately following the tail flick test (n = 4) was not significantly different from the mean (± S.E.M.) AUC (164 ± 21% MPE.hr) achieved when antinociception was determined using the hotplate latency test alone (n = 3) for rats dosed with i.v. morphine (10 µmol).

View larger version (30K): [in this window] [in a new window]   Fig. 1.   A, Normalized mean (± S.E.M.) degree of antinociception (expressed as a percentage of the maximum possible effect (% MPE)) vs. time curve achieved using the hotplate (open symbols) and the tail flick (solid symbols) latency tests following single-dose i.p. administration of M6G (i) 1.9 µmol () or (ii) 3.8 µmol (). B, Mean (± S.E.M.) degree of antinociception (% MPE) vs. time curve achieved using the tail flick (solid symbols) and hotplate (open symbols) latency tests, following single-dose i.v. administration of M6G (i) 1.9 µmol () or (ii) 5.4 µmol ().

View larger version (33K): [in this window] [in a new window]   Fig. 2.   A, Mean (± S.E.M.) degree of antinociception (% MPE) vs. time curve achieved using the hotplate (open symbols) and the tail flick (solid symbols) latency tests following single dose administration of morphine (10 µmol) by either the i.v. () or i.p. () route. B, Mean (± S.E.M.) degree of antinociception (% MPE) vs. time curve achieved using both tail flick (solid symbols) and hotplate (open symbols) latency tests, following single-dose i.p. administration of morphine (i) 10 µmol () or (ii) 20 µmol ().

Figure 1A demonstrates the striking difference (P < .05) observed between the "apparent" levels of antinociception achieved in rats using the tail flick latency test compared with the hotplate method following single-dose administration of i.p. M6G (1.9 µmol). Similarly, following i.v. M6G (1.9 µmol) administration, high levels (>50% MPE) of antinociception were observed using the tail flick test whereas only "apparent" base-line levels of antinociception were observed 30 sec later in the same rats, using the hotplate latency test (Fig. 1B). Specifically, following i.v. M6G administration, high levels (>50% MPE) of antinociception were observed by 5 min after dosing when using the tail flick test whereas the "apparent" levels of antinociception achieved using the hotplate test 30 sec later in the same rats were not significantly different (P > .05) from pre-dosing base-line values or levels of antinociception achieved in control rats dosed with saline. Additionally, following i.v. administration of M6G, the mean (± S.E.M.) extent and duration of antinociception (area under the % MPE vs. time curve, AUC) observed using the tail flick latency test (267 ± 20.7% MPE.hr) was approximately twice the corresponding mean (± S.E.M.) AUC (129 ± 23% MPE.hr) observed following i.p. administration of the same dose of M6G (1.9 µmol). Similarly, doubling the i.p. M6G dose from 1.9 to 3.8 µmol resulted in a ~2-fold increase (P < .05) in the mean (± S.E.M.) value of the AUC from 129 (± 23) to 265 (± 13) % MPE.hr when assessed using the tail flick latency test, but had no effect on the "apparent" base-line levels of antinociception observed using the hotplate test (table 1). Increasing the i.v. dose of M6G from 1.9 to 5.4 µmol did not significantly increase the mean (± S.E.M.) AUC (267 ± 20.7 c.f. 268.3 ± 10% MPE.hr) when determined using the tail flick test because for both doses, maximum antinociception was observed for almost the entire 3 hr experimental period. By contrast, assessment of antinociception using the hotplate latency test in the same rats resulted in "apparent" base-line levels of antinociception (<20% MPE) following administration of both i.v. M6G doses (1.9 µmol and 5.4 µmol) (fig. 1B).

Similar results to those just described for M6G were also observed following both i.v. and i.p. administration of single doses of morphine (10 µmol) in that the mean (± S.E.M.) AUC's obtained (table 1) were consistently larger (P < .05) when antinociception was assessed by the tail flick method compared with the hotplate method (fig. 2A). This was particularly evident following the 10 µmol dose of i.p. morphine where the % MPE values observed using the hotplate test were not significantly different from predosing base-line values throughout the 3 hr experimental period. Although doubling the i.p. M6G dose from 1.9 to 3.8 µmol did not alter the base-line levels of antinociception observed using the hotplate latency test, a similar 2-fold increase in the magnitude of the i.p. morphine dose (from 10 to 20 µmol) evoked significant levels of antinociception using the hotplate latency test (fig. 2B).

Following i.c.v. M6G administration (0.22, 1.25, 2.2, and 3.3 nmol), the "apparent" magnitude and duration of antinociception observed were significantly greater (P < .05) when using the tail flick test relative to the hotplate test (fig. 3, A-C). Dose-dependent increases in antinociception were observed irrespective of the antinociceptive test employed, but these increases did not occur at the same rate for the two tests. Following administration of 0.22 nmol of M6G, low levels of "apparent" antinociception were observed when assessed using the tail flick test (AUC 26.7 ± 10.8% MPE.hr), whereas only base-line levels of antinociception were found when assessed using the hotplate latency test (AUC 2.3 ± 1.2% MPE.hr) (fig. 3, A and B). Although maximum antinociception (100% MPE) was observed using the tail flick test following i.c.v. administration of M6G in doses of 1.25, and 2.2 nmol (fig. 3A), the levels of antinociception observed in the same rats using the hotplate test did not exceed 40 and 60% MPE, respectively (fig. 3B). Additionally, following i.c.v. administration of 1.25 and 2.2 nmol M6G, the mean (± S.E.M.) extent and duration of antinociception (area under the % MPE vs. time curve, AUC) observed was 193.4 ± 37.5 and 221.1 ± 25% MPE.hr respectively, when assessed using the tail flick latency test, which are significantly greater than 28.7 ± 11.8 and 72.3 ± 22% MPE.hr, respectively, when assessed using the hotplate latency test (table 1). Following i.c.v. administration of 3.3 nmol of M6G, high levels of antinociception (>90% MPE) were observed using both the tail flick (AUC = 245.5 ± 25.8% MPE.hr) and hotplate (AUC = 96.2 ± 29% MPE.hr) latency tests. Following the administration of i.c.v. M6G in the dose range 1.25-3.3 nmol, the apparent duration of the antinociceptive effect was significantly shorter (reflected by the significantly lower AUC values) when assessed using the hotplate compared with the tail flick test.

View larger version (32K): [in this window] [in a new window]   Fig. 3.   A, Mean (± S.E.M.) degree of antinociception (% MPE) vs. time curve achieved using the tail flick (solid symbols) test, following single-dose i.c.v. administration of either 0.22 nmol (), 1.25 nmol (), 2.2 nmol () or 3.3 nmol () M6G. B, Mean (± S.E.M.) degree of antinociception (% MPE) vs. time curve achieved using the hotplate (open symbols) test following single-dose i.c.v. administration of either 0.22 nmol (), 1.25 nmol (), 2.2 nmol () or 3.3 nmol () M6G. C, Dose-response curves for i.c.v. M6G based on the area under the % MPE vs. time curve (AUC) data obtained using the tail flick (solid symbols) and hotplate (open symbols) antinociceptive tests.

In terms of behavior, there were no apparent differences between (i) control rats (dosed with i.v. or i.p. heparinised saline), (ii) rats that received M6G (1.9 µmol, 3.8 µmol and 5.4 µmol) by either the i.v. or i.p. routes, (iii) rats that received morphine (10 µmol) administered by the i.p. route or (iv) rats that received 0.22 nmol M6G i.c.v. However, rats that received i.v. morphine (10 and 20 µmol) or i.p. morphine (20 µmol) were noticeably sedated during the three hour study period. Rats that received i.c.v. M6G (1.25, 2.2 and 3.3 nmol) exhibited behavior dissimilar from that of control rats (i.c.v. saline). Specifically, these rats tended to have a rigid body posture, protruding eyes and they appeared to experience dose-dependent respiratory depression. Initially, rats that received i.c.v. 3.3 nmol M6G also displayed avoidance to a non-noxious stimulus and increased motor activity.

	Discussion

Top Abstract Introduction Methods Results Discussion References

Intracerebroventricular (i.c.v.) administration of M6G in a range of doses (0.22-3.3 nmol) produced significantly higher levels of "apparent" antinociception when assessed using the tail flick test relative to the hotplate test performed 30 sec later in the same rats. Additionally, systemic administration of M6G in a range of doses (1.9-5.4 µmol) given by the i.v. or i.p. routes, resulted in high levels of antinociception using the tail flick test whereas no significant antinociception was detected 30 sec later using the hotplate latency test in the same rats. Given that the tail flick and the hotplate tests have been reported to involve different nociceptive reflexes, it is important to review the relevant literature to more fully interpret our experimental data.

The tail flick test is regarded as a spinal reflex that is influenced by the activity of supraspinal structures (Irwin et al., 1951; Ramabadran and Bansinath, 1986; Pastoriza et al., 1996). Such a view is supported by the findings of Grossman and coworkers (1982) who showed that both the descending and the segmental control of the tail flick reflex are exerted in the region where tail flick afferents terminate in Laminae I and II of spinal cord segments S3 to Co3. In comparison, the hotplate test appears to require an intact central nervous system. In one study, bilateral lesions within the medial frontal cortex of the rat brain resulted in an increase (~80% ) in the mean hotplate latencies compared with non-lesioned control rats, whereas these lesions had no apparent effect on tail flick latencies, formalin test scores, or motor function (Pastoriza et al., 1996). There is also evidence that nociception can be modulated at the supraspinal level via ascending projections from the periaqueductal gray matter (PAG) (Morgan et al., 1989). In rats where the descending projections to the spinal cord were lesioned caudal to the PAG, antinociception elicited by electrical stimulation of the PAG decreased significantly (P < .01) from 50% to 14% using the tail flick latency test but increased, although not significantly, from approximately 46% to 60% (P > .05) using the hotplate latency test (Morgan et al., 1989). Thus, the available evidence tends to support the view that the tail flick test is appropriate for the detection of antinociception that is mediated predominantly by spinal mechanisms, whereas the hotplate test is appropriate for the detection of antinociception that is mediated predominantly by supraspinal structures.

Based on the above, our findings strongly indicate that spinal rather than supraspinal mechanisms primarily mediate the antinociceptive effects of systemically administered M6G. This interpretation of our data is supported by (i) our observation that following i.c.v. M6G dosing, the extent and duration of antinociception (AUC) was greater when determined using the Tail flick test compared with the Hotplate test (fig. 3, A and B) and (ii) previous studies that reported M6G had a higher relative potency compared with morphine following i.t. dosing (650 times) than i.c.v. dosing (90 times) in mice, using the tail flick test (Paul et al., 1989). Also our studies have shown that using the tail flick test, i.v. M6G appears to be approximately 5 times more potent than i.v morphine (compare figs. 1 and 2; table 1).

Traditional teaching suggests that M6G would be too hydrophilic (due to its glucuronide moiety) to pass through the BBB (Wu et al., 1997). However, studies based on relative HPLC retention times (Carrupt et al., 1991) or a potentiometric method (Avdeef et al., 1996) have shown that M6G is less than one order of magnitude more hydrophilic than morphine. Additionally as early as 1973, Yoshimura et al., showed unequivocally that radiolabelled M6G was able to cross the intact BBB in the rat brain. More recently, following systemic administration of M6G (s.c. 25 µmol/kg) to adult male Wistar rats, Aasmunstad and coworkers (1995) showed using in vivo brain microdialysis methods that M6G was detectable in the extracellular fluid of the CNS, with maximum concentrations occurring at approximately 50 min post-dosing. There are several possible mechanisms through which M6G may cross the BBB. Using computer-aided force field models, Carrupt and co-workers (1991) proposed that M3G and especially M6G are only slightly less lipophilic than morphine itself because these molecules have the ability to adapt their polarity to that of their medium by intramolecular folding. Thus in polar environments, it was proposed that the morphine glucuronides, M3G and M6G would exist predominantly in the "extended" form, with polar groups exposed. By contrast, in a lipophilic environment such as biological membranes, M6G would exist in the folded form, concealing its polar groups, thereby facilitating passive diffusion across the BBB. It has also been proposed (Prankerd, 1993) that two molecules of either M3G or M6G could form a double ion pair, in which the protonated tertiary nitrogen of one molecule pairs with the carboxylate group of the other and vice versa. This would have the effect of producing electronically "neutral" zwitterions, which could also passively diffuse across the BBB. Also an active transport process has been proposed from in vitro experiments in the choroid plexus (Hug, 1967). Finally it is also possible that M6G may cross the BBB by passing through the fenestrated epithelium of the third ventricles (Smith and Shine, 1992).

Although Gong and coworkers (1992) have reported previously that i.c.v. M6G administered to SD rats in a dose of 0.2 nmol evoked high levels of antinociception (>80% MPE) using both the tail flick and hotplate latency tests, we were unable to reproduce their results. Following i.c.v. administration of 0.2 nmol of M6G in our studies, rats did not achieve significant antinociception when assessed by either the tail flick or the hotplate latency tests. However, after increasing the i.c.v. M6G dose ten-fold to 2.2 nmol, 100% MPE was achieved using the tail flick test and  40% MPE was observed 30 sec later in the same rats using the hotplate latency test. The slopes of the sigmoidal dose-response curves were considerably different with a much steeper dose-response relationship being apparent for the tail flick latency data. These findings are consistent with the view that distinctly different mechanisms mediate the spinal and supraspinal antinociceptive effects of M6G.

Additionally, the time to achieve maximum antinociception following i.c.v. M6G dosing in our studies was  30-60 min, which is comparable with the median time (53 min) to achieve the maximum M6G concentration in rat brain microdialysate following subcutaneous M6G (25 µmol/kg) administration (Aasmundstad et al., 1995).

Our findings also show that systemically administered morphine evoked higher levels of "apparent" antinociception when assessed using the tail flick rather than the hotplate latency test (fig. 2, A and B; table 1). By contrast with M6G, a 2-fold increase in the i.p. morphine dose (from 10 to 20 µmol) evoked significant levels of antinociception using the hotplate test whereas 2-fold and 3-fold increases in the magnitude of the i.p and i.v. M6G doses, respectively, administered to rats in this study, evoked no significant antinociception using the hotplate test. Taken together these results indicate the apparent involvement of both spinal and supraspinal mechanisms in the antinociceptive effects of systemically administered morphine, particularly following administration of the larger morphine dose (20 µmol). Thus, our data do not support the previously reported view that spinal nociceptive processes are less sensitive to morphine and that following systemic morphine administration, supraspinal antinociceptive mechanisms predominate (Ling and Pasternak, 1983; Heyman et al., 1988). Rather, low dose systemic morphine (10 µmol) appears to act primarily by spinal mechanisms (remembering that i.p. morphine undergoes significant presystemic metabolism compared with i.v. morphine) with the contribution of supraspinal mechanisms becoming more evident as the systemic morphine dose is increased. However, similar findings to ours were reported by Langerman et al. (1995) following chronic i.v. infusion of morphine to SD rats at three different dosing rates (2, 4, and 6 mg kg¹ hr¹), where dissimilar levels of "apparent" antinociception were obtained depending upon whether the tail flick or the hotplate method was used for antinociceptive testing. The authors suggested that their findings were possibly due to the differential effects of morphine at spinal and supraspinal sites. Given that spinal/supraspinal synergy has been observed following concomitant i.c.v and i.t. morphine administration (Porreca et al., 1987; Pick et al., 1992; Pick et al., 1993; Miaskowski et al., 1993), it is highly likely that spinal/supraspinal synergy may underlie the antinociceptive effects observed in our studies following (i) i.p. administration of the higher morphine dose (20 µmol) and (ii) the administration of i.v. morphine (10 µmol) to rats in our studies.

Recently published studies have shown unequivocally that the integrity of the BBB is similar at the levels of the brain and spinal cord due to the presence of similar tight junctions between cells in the endothelial walls of the capillaries in the cerebral cortex, cerebellum and spinal cord (Azzi et al., 1990; Smith and Shine, 1992). Therefore our results showing that M6G appears to mediate its antinociceptive effects primarily through spinal mechanisms are not because M6G is more readily able to cross the BBB at the level of the spinal cord following systemic dosing. Rather, systemically administered M6G may reach supraspinal structures more rapidly than the spinal cord through the "leaky" fenestrated epithelium of the third ventricle (Smith and Shine, 1992).

As there was no significant difference (P > .05) between the % MPE values obtained in rats assessed using the hotplate test alone compared with rats assessed on the hotplate 30 sec after the tail flick test, this indicates that tail flick antinociceptive testing immediately prior to the assessment of antinociception using the hotplate test, did not significantly alter the subsequently determined hotplate latency values determined in our studies.

In summary, the results of our studies described herein clearly show that systemically and centrally administered M6G evokes antinociception predominantly via spinal mechanisms in the doses investigated, whereas the antinociceptive effects of systemically administered morphine involve both spinal and supraspinal mechanisms. It is important that this fact be recognized (i) when comparing experimental results obtained in different laboratories following administration of M6G to ensure that similar methods of antinociceptive testing have been employed and (ii) to ensure selection of the most appropriate method(s) for antinociceptive testing following M6G administration in any future studies. Our results also show that morphine appears to act predominantly via spinal mechanisms when administered systemically in relatively low doses (10 µmol) to rats with the contribution of supraspinal mechanisms becoming more significant when the magnitude of the systemic morphine dose is increased above 10 µmol. Furthermore, following systemic administration of morphine to the rat, the ED₅₀ doses determined using the hotplate latency test would be expected to be higher than those determined using the tail flick latency test, and these respective doses are not directly comparable.