Brain Research 887 (2000) 436–439
www.elsevier.com / locate / bres
Short communication
Magnesium and MK-801 have a similar effect in two experimental
models of neuropathic pain
*
`
Sophie Begon , Gisele Pickering, Alain Eschalier, Claude Dubray
´ ´ ´
INSERM EPI9904, Pharmacologie Fondamentale et Clinique de la Douleur, Laboratoire de Pharmacologie Medicale, Faculte de Medecine, B.P. 38, 63001 Clermont-Ferrand, Cedex 1, France
Accepted 19 September 2000
Abstract
Considering that magnesium and non-competitive NMDA receptor antagonists inhibit the opening of the channel linked to the NMDA receptor, we assessed their effects on mechanical hyperalgesia in two animal models of neuropathic pain (rats with a sciatic nerve ligature and diabetic rats). Our data show that magnesium reverses the hyperalgesia, as does MK-801. These results suggest that magnesium could be an alternative for the treatment of neuropathic pain in patients. 2000 Elsevier Science B.V. All rights reserved.
Theme: Sensory systems
Topic: Pain modulation: pharmacology
Keywords: Neuropathic pain; Hyperalgesia; Magnesium; MK-801
Current hypotheses about the pathophysiology of neuro- IASP guidelines for animal experiments [28]. For the pathic pain highlight the role of the N-methyl-D-aspartate experiment carried out in diabetic rats, Sprague–Dawley
(NMDA) receptor channel in the sensitization of the spinal male rats (Charles River, France) were injected intraperito-cord neurons which leads to chronic pain [8,26]. As a neally (i.p.) with streptozocin (STZ) (75 mg / kg) (Zanosar, matter of fact, many experimental studies as well as some Upjohn). From the 21st day onwards, several troubles of clinical trials reported the analgesic efficacy of non-com- nociception have been described [6], particularly a me-petitive NMDA antagonists such as dizocilpine maleate chanical hyperalgesia. The other experiment used a model (MK-801) or ketamine in neuropathic pain [5]. However, of peripheral mononeuropathy in rat. Sprague–Dawley with these drugs, a lot of adverse events, described in male rats (Charles River, France) were anesthetized and a patients [20], limit the therapeutic use of NMDA antago- chronic constrictive injury (CCI) of their right common nists. Electrophysiological and binding studies have shown sciatic nerve was performed according to the method that MK-801 is able to block the NMDA receptor channel described by Bennett and Xie [3].
[24] and furthermore, that the opening of this channel The CCI model develops a mechanical hyperalgesia
21
depends on the magnesium (Mg ) ion [18]. Therefore, we from day 12 after surgery [1]. The contralateral limb
21
hypothesized that Mg could have similar anti-hy- remained unoperated.
peralgesic properties as non-competitive NMDA antago- In these two experiments, the changes in mechanical nists with less adverse events. The aim of this study was to nociceptive thresholds were assessed by the paw pressure compare the anti-hyperalgesic effect of repeated dosages of test [21]. An increasing mechanical pressure was applied
21
Mg and MK-801 in two experimental models of neuro- by an analgesymeter (Apelex type 003920, Ugo Basil, pathic pain, the diabetic rat and the mononeuropathic rat. Italy) on a hind paw, until a vocalization was elicited These studies were conducted in accordance with the (vocalization threshold expressed in grams). For the CCI model, the measurement was carried out on the ipsilateral to the ligature and the contralateral paws. For each
*Corresponding author. Tel.: 133-4-7360-8033; fax: 1
33-4-7327-experiment, the vocalization threshold was assessed before
7162.
E-mail address: sophie [email protected] (S. Begon). the induction of the neuropathic hyperalgesia (baseline
]
S. Begon et al. / Brain Research 887 (2000) 436 –439 437
values), before the induction of treatment, when the hyperalgesia was developed (control values) and then twice daily (performed before each injection, at 8:00 am and 8:00 pm) throughout the period of treatment. No significant difference was observed between these two daily values of the vocalization thresholds; thus they have been averaged for each day.
For the two experiments, the treatments were given twice daily for 5 days. One group of animals (n58) received MK-801, dizocipline maleate (RBI / Sigma–Al-drich, France) (0.1 mg / kg, i.p., b.i.d.). For the control group (n58) animals were injected with saline (2 ml / kg, i.p., b.i.d.). A third group (n58) was treated with mag-nesium sulfate (MgSO ) (Sigma–Aldrich Chimie, France).4
The diabetic rats were given a slightly higher dosage of
21 21
Mg (200 mg / kg MgSO , i.e. 40 mg / kg Mg4 , i.p., b.i.d.) in comparison with the dosage used in
mononeuro-21
pathic rats (150 mg / kg MgSO , i.e. 30 mg / kg of Mg4 , i.p., b.i.d.). That choice was guided by published data [10]
21
reporting a Mg depletion in diabetic rats due to the polyuria. All the treatments were performed blind.
The data analysis was performed by a two-way ANOVA, followed by a PLS Fisher’s test to compare the treated
21
groups with Mg or MK-801 versus the control saline- Fig. 1. Effect of a 5-day treatment with MgSO (200 mg / kg, i.p., twice a
4
treated group (Statview 4.1 software). The area under the day) or MK-801 (0.1 mg / kg, i.p., twice a day) or saline, on mechanical
time-course curve (AUC) of the vocalization threshold was hyperalgesia in diabetic rats (n58 / group) (A). The two daily measure-ments of the vocalization threshold were averaged for each day. Data are
also determined throughout the treatment period. That
presented as means6S.E.M., ***P,0.001, **P,0.01, *P,0.05, versus
parameter was analyzed by a Student’s t-test to compare
the corresponding values of the control saline-treated group. Area under 21
the treated groups with Mg or MK-801 versus the the time-course curve (AUC) was calculated throughout 5 days of control saline-treated group. treatment (B), **P,0.01, versus the control saline-treated group.
In the model of diabetic neuropathy, the animals de-veloped a mechanical hyperalgesia according to the time
schedule described by Courteix et al. [6]. Three weeks assessed by the AUC values, show that Mg significantly after the induction of the diabetes, the decrease of the reversed (P,0.01) the mechanical hyperalgesia in the vocalization threshold (40.863.9, 41.862.9 and diabetic rats compared to the control saline-treated group, 45.061.6% in MgSO , MK-801 and saline-treated groups,4 with the same efficacy as does the NMDA antagonist respectively), as well as the baseline values (319.3610.8, treatment (Fig. 1B).
319.3616.2 and 312.9613.0 g for MgSO , MK-801 and4 In the experiment with the CCI model, a mechanical saline-treated groups, respectively) were very similar in hyperalgesia occurred 12 days after the sciatic nerve each group. The treatment with MgSO (200 mg / kg, i.p.,4 ligature (Fig. 2A). In comparison to the baseline values b.i.d.) increased progressively the vocalization threshold (272.3612.0, 257.869.3 and 263.4612.1 g for MgSO ,4
(Fig. 1A). After 3 days of treatment, this threshold was MK-801 and control saline-treated groups, respectively)
21
significantly higher (P,0.001) in the Mg -treated group before surgery, the vocalization threshold fell to 47.962.3, (354.2637.4 g) than in the control saline-treated group 50.162.3 and 48.262.4%, respectively, for each group of (192.6610.2 g). Thereafter, the vocalization threshold was animals. Systemic injection of MgSO (150 mg / kg, i.p.,4
leveled off until the end of the experiment without b.i.d.) reduced the mechanical hyperalgesia in mononeuro-significant differences with the baseline values. With the pathic rats. In comparison with the control saline-treated NMDA antagonist (0.1 mg / kg, i.p., b.i.d.), the mechanical group, the vocalization threshold became significant (P,
hyperalgesia was also totally reversed so that the vocaliza- 0.05) on the first day of treatment with MgSO4
tion threshold returned to the baseline values (Fig. 1A). In (138.165.5 versus 201.5619.0 g, respectively), then re-the MK-801-treated group re-the anti-hyperalgesic effect mained significant throughout all the experiment. The
21
438 S. Begon et al. / Brain Research 887 (2000) 436 –439
diabetic rats. Malcangio et al. [17] also found a significant anti-hyperalgesic effect which occurs 2 days after starting a similar treatment as the one we used. These authors found no effect of repeated dosages of MK-801 in normal rats.
In mononeuropathic rats, we showed curative effect of MK-801 (0.1 mg / kg, i.p., b.i.d.) on the induced hy-peralgesia. Similarly, Birch et al. [4] reported, as we did, such a curative effect in the CCI model. Nevertheless, most studies investigated a preemptive treatment with MK-801 by using fairly high dosages of MK-801 (0.5, 0.6 or 1 mg / kg / day, s.c. or i.p.) [7,15,22]. In these conditions MK-801 reduced mechanical and thermal hyperalgesia more rapidly and with a larger magnitude as observed in our experiments. On the counterpart, at these dosages, MK-801 often induces stereotypic behavior (head weaving and circling, jerky movements), loss of balance and ataxia [13,22,23] that we did not observe in our experiments.
The results obtained in our study show for the first time,
21
that repeated dosages of Mg totally or partially reverse the mechanical hyperalgesia both in diabetic rats and in CCI model. Similarly, after repeated systemic injection of
21
Mg , Xiao et al. [27] showed a prolonged effect on the heat-hyperalgesia in the CCI model.
21
Hallack et al. [12] showed in rat, an increase in Mg level in both cerebrospinal fluid and different brain re-gions, after i.p. injections of MgSO . That is in agreement4
21
with the possibility of Mg to cross the blood brain
21
barrier. Significant changes in Mg concentration were observed in the spinal cord, after a systemic injection, by
21
Feria et al. [9]. On the other hand, Mg is able to
1 11
modulate the opening of the Na / Ca channel of the NMDA receptor [18]. Several authors [4,17] give evidence that this type of receptor channel is involved in neuro-pathy-induced central sensitization. Thus, the
anti-hy-21
peralgesic effect of Mg in these two models of neuro-pathy, could result from a direct interaction of this ion with the activated NMDA receptor channel.
Fig. 2. Effect of a twice daily treatment, for 5 days, with MgSO (1504
mg / kg, i.p.) or MK-801 (0.1 mg / kg, i.p.) or saline in rats with a chronic As for the MK-801 treatment, we took care to avoid 21
constrictive injury of the sciatic nerve (n58 / group). The vocalization some adverse events due to the Mg injection, which threshold was assessed on the ipsilateral (A) and the contralateral (B)
could lead to a misinterpretation of the behavioral tests.
paws. The two daily measurements of the vocalization threshold were
Ishizaki et al. [14] after intrathecal administration of
averaged for each day. Data are presented as means6S.E.M., **P,0.01,
MgSO , described the occurrence of motor paralysis or
*P,0.05, versus the corresponding values of the control saline-treated 4
group. AUC was calculated throughout 5 days of treatment (C), ***P, ataxia. As a consequence, we used fairly low dosages of
0.001, **P,0.01, versus the control saline-treated group. MgSO in order to avoid these phenomena. Xiao et al. [27] 4
21
reported, as we did, that low dosages of Mg (30 and 40 model (Fig. 2B). AUC values (Fig. 2C), integrating the mg / kg) did not induce any adverse effects. In ours studies, effect of both compounds throughout the whole period of the fact that both tested drugs did not modify the
vocaliza-21
treatment, confirm the efficacy of Mg and MK-801 tion threshold for the contralateral paw in the CCI model,
21
treatment on the hyperalgesia compared to the control gives evidence that neither MK-801 nor Mg disturbed saline-treated group (P,0.001 and P,0.01, respectively). the behavioral assessment of the nociceptive sensitivity in The effects we observed with MK-801, taken as a rats. On the other hand, abdominal constrictions were reference NMDA antagonist, are fully in agreement with reported in mice, after an i.p. injection of MgSO4 (120
21
S. Begon et al. / Brain Research 887 (2000) 436 –439 439
diabetes mellitus and magnesium, Magnes. Res. 5 (3) (1992) 193–
with the assessment of pain threshold. That is why we
202.
deliberately assessed the vocalization threshold far from
[11] K. Gyires, Z. Toma, The use of the writhing test in mice for
the injection time to avoid them. screening different types of analgesics, Arch. Int. Pharmacodyn. 267 In these two models of experimental neuropathic pain, (1984) 131–140.
21
we showed that both MK-801 and Mg , which share [12] M. Hallak, Effect of parenteral magnesium sulfate administration on excitatory amino acid receptors in the rat brain, Magnes. Res. 11 (2)
some blocking properties on NMDA receptor channel, are
(1998) 117–131.
able to reverse the induced hyperalgesia with a very
[13] M. Irifune, T. Shimizu, M. Nomoto, T. Fukuda, Involvement of
similar degree of efficacy. The two treatments did not N-methyl-D-aspartate (NMDA) receptors in non-competitive NMDA induce any changes in the non-ligated paw in the CCI receptor antagonist-induced hyperlocomotion in mice, Pharmacol.
model. This suggests an action only in conditions of Biochem. Behav. 51 (2–3) (1995) 291–296.
[14] K. Ishizaki, M. Sasaki, S. Karasawa, H. Obata, T. Nara, F. Goto,
persistent pain and this is in line with recently published
The effect of intrathecal magnesium sulfate on nociception in rat
data emphasizing the role of NMDA receptors on the
acute pain models, Anesthesia 54 (1999) 241–246.
sensitization of dorsal horn neurons observed in similar [15] M. Kawamata, K. Omote, Involvement of increased excitatory
21
conditions [25]. Therefore, much attention has been placed amino acids and intracellular Ca concentration in the spinal dorsal
on NMDA receptor antagonists as a treatment for chronic horn in an animal model of neuropathic pain, Pain 68 (1996) 85–96. [16] W.S. Kingery, A clinical review of controlled clinical trials for
pain [5]. In patients suffering from neuropathic pain the
peripheral neuropathic pain and complex regional pain syndromes,
response to these treatments is often satisfactory in terms
Pain 73 (1997) 123–140.
of pain relief [16,19], but its clinical use is strongly limited [17] M. Malcangio, D.R. Tomlinson, A pharmacologic analysis of by frequent and very uncomfortable adverse events [2]. mechanical hyperalgesia in streptozotocin / diabetic rats, Pain 76
Our results suggest that it would be interesting to under- (1998) 151–157.
21 [18] M.L. Mayer, G.L. Westbrook, P.B. Guthrie, Voltage-dependent block
take clinical trials on Mg therapy in patients suffering
of magnesium of NMDA responses in spinal cord neurons, Nature
from painful peripheral neuropathies.
309 (1984) 261–263.
[19] K.A. Nelson, K.M. Park, E. Robinovitz, C. Tsigos, M.B. Max, High dose oral dextrometrophan versus placebo in painful diabetic
References neuropathy and postherpetic neuralgia, Neurology 48 (1997) 1212–
1218.
[20] J.W. Olney, Excitotoxic amino acids and neuropsychiatrics disorders, [1] N. Attal, F. Jazat, V. Kayser, G. Guilbaud, Further evidence for
Ann. Rev. Pharm. Toxicol. 30 (1990) 47–71. ‘pain-related’ behaviours in a model of unilateral peripheral
mono-[21] L.O. Randall, J. Sellito, A method for measurement of analgesic neuropathy, Pain 41 (1990) 235–251.
activity on inflamed tissue, Arch. Int. Pharmacodyn. 4 (1957) [2] M. Backonja, G. Arndt, K.A. Gombar, B. Check, M. Zimmermann,
409–419. Response of chronic neuropathic pain syndromes to ketamine: a
[22] G.D. Smith, J. Wiseman, S.M. Harrison, P.J. Elliott, P.J. Birch, preliminary study, Pain 56 (1994) 51–57.
Pretreatment with MK-801, a non-competitive NMDA antagonist, [3] J.G. Bennett, Y.K. Xie, A peripheral mononeuropathy in rat that
prevents development of mechanical hyperalgesia in a rat model of produces disorders of pain sensation like those seen in man, Pain 33
chronic neuropathy, but not in a model of chronic inflammation, (1988) 87–107.
Neurosci. Lett. 165 (1994) 79–83. [4] P.J. Birch, Clinical relevance of receptor pharmacology in the
[23] A. Vezzani, R. Serafini, M.A. Stasi, S. Caccia, I. Conti, R.V. Tridico, nociceptive pathway, Pain Rev. 2 (1995) 13–27.
R. Samanin, Kinetics of MK-801 and its effect on quinolinic [5] T.J. Coderre, Novel Aspects of Pain Management: Opioids and
acid-induced seizures and neurotoxicity in rats, J. Pharmacol. Exp. Beyond, in: J. Sawynok, A. Cowan (Eds.), Excitatory Amino Acid
Ther. 249 (1988) 278–283. Antagonists: Potential Analgesics for Persistent Pain, Wiley–Liss,
[24] E.H.F. Wong, A.R. Knight, G.N. Woodruff, [3H]MK-801 labels a New York, 1999, pp. 157–178.
site on the N-methyl-D-aspartate receptor channel complex in rat [6] C. Courteix, A. Eschalier, J. Lavarenne, Streptozocin-induced
brain membranes, J. Neurochem. 50 (1988) 274–281. diabetic rats: behavioural evidence for a model of chronic pain, Pain
[25] C.J. Woolf, R.J. Mannion, Neuropathic pain: aetiology, symptoms, 53 (1993) 81–88.
mechanisms, and management, Lancet 353 (1999) 1959–1964. [7] G. Davar, A. Hama, A. Deykin, B. Vos, R. Maciewicz, MK-801
[26] C.J. Woolf, S.W. Thompson, The induction and maintenance of blocks the development of thermal hyperalgesia in a rat model of
central sensitization is dependent on N-methyl-D-aspartic acid experimental painful neuropathy, Brain Res. 553 (1991) 327–330.
receptor activation; implications for the treatment of post-injury [8] A.H. Dickenson, V. Chapman, G.M. Green, The pharmacology of
hypersensitivity states, Pain 44 (3) (1991) 293–299. excitatory and inhibitory amino acid-mediated events in the
trans-[27] W.H. Xiao, G.J. Bennett, Magnesium suppresses neuropathic pain mission and modulation of pain in the spinal cord, Gen. Pharmacol.
responses in rats via a spinal site of action, Brain Res. 666 (1994) 28 (1997) 633–638.
168–172. [9] M. Feria, F. Abad, A. Sanchez, P. Abreu, Magnesium sulfate
[28] M. Zimmerman, Ethical guidelines for investigations of experimen-injected subcutaneously suppresses autotomy in peripherally
deaf-tal pain in conscious animals, Pain 16 (1983) 109–111. ferented rats, Pain 53 (1993) 287–293.
