Magnesium deficiency reveals the neurotoxicity of Δ‐9‐tetrahydrocannabinol (THC) low doses in rats

ARTICLE

Auteur(s) : P. Bac¹, N. Pages², C. Herrenknecht¹, P. Maurois¹, J. Durlach³

¹ Laboratoire de Pharmacologie, Faculté de Pharmacie, Paris XI, 92296 Châtenay Malabry; ² Laboratoire de Toxicologie, Faculté de Pharmacie, Strasbourg, 67400 Illkirch Graffenstaden; ³ SDRM, UPCM, Paris VI, 75252 Paris cedex 5

Introduction

Rat muricidal behaviour (MB) is a striking interspecific agressive behaviour that leads rats to prey on any mouse brought into their cage and to kill it. MB may be observed in natural killer rats (NK) as a result of innate habits [1]. In all strains, 10-15% rats are NK, whereas their congeners are not [2]. This behaviour may be also induced by long term isolation [3, 4], chemicals (p. chlorophenylalanine, THC) [5-13], nutritional deficiencies (thiamine, tryptophan, magnesium) [14-18] or various regional brain lesions [19-24]. Acute (single i.p. 11 mg THC /kg b.w) or chronic administration of cannabis extracts or THC may induce behavioural alterations, mainly characterized by a hyperaggressiveness, in rats [7-12]. Individual housing and chronic starvation were reported as basic features for THC to induce aggressiveness in the rat [13]. Similarly, we reported that a severe magnesium deficiency induced by a 50 ppm magnesium-deficient diet was also responsible for MB in rats [12, 18]. Intraspecific aggressiveness was also observed in mice presenting a low blood magnesium level [25]. It is worth noting that both THC and magnesium deficiency may decrease the serotoninergic function, which plays an inhibitory role in killing behaviour. Nevertheless, a lot of other mechanisms of neural hyperexcitability can be also concerned [26-29]. Consequently, it may be assumed that any decrease in magnesium plasma levels may potentiate the neurotoxicity of xenobiotics.
The aim of the present paper was to compare MB induced by a single 11 mg THC /kg b.w. in rats fed a normal magnesium diet (1700 ± 100 ppm) or by lower doses of THC (2, 4 and 8 mg/kg) in magnesium-deficient rats suffering a severe or moderate magnesium-deficiency to that of natural killer rats (NK).

Material and methods

Rat diets

Male Long Evans, 7 weeks old rats (Janvier, France) were used. They were fed on either UAR standard diet (1700 ± 100 ppm) or 50 or 150 ppm UAR Mg-deficient diets, for a period of 42 days. Room temperature was maintained at 21 ± 1°C with a dark:light cycle of 12:12 hrs. Care and treatment of rats were according to the guidelines for animal care.

Muricide assay

The determination of the NK rat percentage has been previously described [11]. Briefly, NK rats were selected without isolation and starvation, by leaving a new mouse with each rat as follows: 24 h on Day 0, 4 h on day 1, and 10 min on day 2. The rats that repeatedly killed the mouse were considered as NK.
The naive rat groups were subjected to a 24 h period of starvation and a 48 h period of isolated housing. Three hours before treatment, they were fed ad libitum. The muricide assay was repeated six times every hour with each rat by introducing a new OF1 mouse (IFFA Credo, France) (body weight 22-26 g) into the cage every 60 min. All the rats were tested for muricidal activity, 1, 2, 3, 4, 5 and 6 hours after the injection. The MB occurrence and the muricidal pattern (attack latency, attack on the living mouse and attack on the dead mouse expressed as seconds) were investigated in both experiments and compared to that of NK. In the present paper, results are reported as percentage of the mean values measured in NK.

1^st experiment

One group of 40 rats with a normal magnesium status was i.p. injected with 11 mg/kg THC (1% aqueous solution in Tween 80) under constant volume (2.5 ml/kg). Preliminary assays were done on smaller rat groups (n = 20) by injecting in the same conditions 2, 4 or 8 mg/kg THC.

2^nd experiment

Two groups of 160 magnesium-deficient rats were fed for 42 days a 50 or 150 ppm magnesium-deficient diet. At the end of the deprivation period, they were i.p. injected with 0, 2, 4 or 8 mg/kg THC in the same conditions as previously indicated. Finally, one group (n = 40) fed 50 ppm magnesium-diet was injected 11 mg/kg THC.

Results

Percentage of natural killer rats and MB pattern

In the Long Evans strain used, 11% rats were NK, the remaining 89% being non killer rats. NK rats, whatever the assay number presented a similar pattern. As soon as the mouse was placed into the rat's cage, the MB steps were as follows. First (attack latency), the rat observed the mouse and approached it slowly (14 ± 0.3 s), then (attack on the living mouse), it attacked the mouse directly over the neck until death (24 ± 0.3 s). Finally, the rat continues attacking the dead mouse for a short period of time (62 ± 0.5 s).

MB in naive rats injected once 11 mg THC/kg b.w.

One hour after a single injection of 11 mg THC/kg b.w., 70% of the naive rats tested, in conditions of starvation and isolated housing, exhibited MB. Even after taking off the percentage of NK, it appeared that 59% of naive rats became muricidal after a single dose of THC. One hour after the first injection, their MB pattern significantly differed from that of NK. The durations of all three phases were notably increased, mainly the attack on the dead mouse which was 5-fold increased (302 ± 8 s vs 62 ± 0.5 s). Similar trends, somewhat attenuated, were observed in the following assays, 2 and 3 h after injection. Thereafter, from the 4^th assay, MB became more efficient. The two first phases were as in NK rats whereas the attack on the dead mouse was still slighly increased. Finally, on the 6^th assay, all the naive rats that became muricidal presented the same behaviour as NK rats.
A single injection of lower doses (2, 4, 6 or 8 mg/kg b.w.) failed to induce additional MB.

Magnesium deficiencies alone

Rats were fed severe or moderate (50 ppm or 150 ppm) Mg-deficient diets for 42 days. These treatments resulted, at the end of the deprivation period, in 4- and 2- fold decreases in magnesium plasma levels respectively [12]. The naive rats fed 150 ppm did not show additional MB as compared to NK. In contrast, 100% of the rats fed 50 ppm became muricidal. Their MB differed from that of NK only with regard to the 2^nd phase which duration was about twice increased (56.2 ± 0.4 s vs 25.6 ± 0.6 s). Interestingly, this MB remained quite identical over the 6 successive assays.

THC in magnesium deficient rats

Severe magnesium deficiency

Rats were injected either 2, 4, 8 or 11 mg THC /kg b.w.. A 100% MB appeared as could be expected from the previous assay, in the 50 ppm magnesium-deficient rat group without THC treatment, for the three lower doses. The MB common features in those rat groups were an almost instantaneous mouse killing and a furious destructive behaviour on the dead mouse. On the 6^th assay, the time elapsed from the moment of the mouse introduction into the cage and the mouse death was lower than the attack latency alone in NK rats. This results both from a shortening of the attack latency duration which reached a constant value of 3.5 ± 0.2 s from the 2^nd assay and from a decrease in the mouse killing duration (9.4 ± 1.8 s), both reflecting an increased aggressiveness. An apparent discrepancy appeared in the 4 and 8 mg/kg treated rat groups, in the 1^st and 2^nd assays which were significantly increased as compared to NK or to rats exposed to 0 and 2 mg THC /kg b.w.. This effect was attributed to an inadequate localization and holding of the prey linked to THC-induced motor uncoordination in the two first hours following injection. Conversely, the rats tore the mouse to pieces more and more furiously for period which increased on every new assay. For the same reason, the higher dose (11 mg/ kg) was responsible for an important increase in the attack latency (range: 62.3 ± 8.1 s in the1^st assay - 31.5 ± 4.3 s in the 6^th) and in the attack on the living mouse (range: 55.6 ± 6.2 – 38.5 ± 5.1 s from the 1^st to the 6^th assay respectively) over the 6 successive assays. As indicated, both phases decreased slowly from the 1^st to the 6^th assay but remained in any case longer than with the lower doses. However, as with the low doses, the duration of the attack on the dead mouse was immediately severe (286.5 ± 12.4 s on the 1^st assay) and increased to values higher than 10 minutes on the 4^th, 5^th and 6^th assays.

Moderate magnesium deficiency

As previously, the injection of either 2, 4 or 8 mg THC/kg b.w. resulted in a severe alteration of MB pattern represented by shortened periods of the attack latency (which reached less than 5 sec from the 4^th assay with 4 mg THC/kg and from the 1^st assay with 8 mg/kg). Even 2 mg/kg induced a decrease in the attack latency (9.2 ± 0.7 sec on the 6^th assay) as compared to NK. On the 6^th assay, whatever the dose, the attack on the living mouse decreased slightly (18.2 ± 0.3 s) as compared to NK (24.5 ± 0.3 s) and significantly (p < 0.001) as compared to untreated magnesium-deficient mice (54.4 ± 1.1 s). Finally, the 3^rd phase enhanced progressively and reached a 2 or 3-fold increase in the three treated groups, in a THC-dose dependent manner, on the 6^th assay.

Discussion

MB may be either an innate behaviour in NK rats, or, more frequently, the consequence of various stresses or xenobiotics leading to severe neurotransmitter imbalances responsible for a marked interspecific aggressivity. In the Long Evans strain used, 11% of the rats were NK, a percentage which is agreement with the usual rates measured in other strains [2]. The presence of an intrusive mouse, by inducing a great stress in the rat, reveals its agressiveness. In the present study, we used OF1 mice, a strain rather aggressive by itself which is more stressing for the rat than passive strains like Balb/c, for instance [18]. As reported before, it seems that the rat responds to the social stress linked to the mouse presence and behaviour by killing it. The latency duration corresponds to the maximal acute stress duration the rat can indure before attacking the mouse [18]. In natural killer rat, the attack on the living mouse is efficient resulting in a rapid killing of the mouse. The mouse death is generally followed by an attack on the dead mouse which does not exceed 1 minute. The measurement of each successive phase duration brings consequently interesting information about the rat behaviour. In the present paper, we studied the consequences of (i) a single i.p. injection of THC various doses, a chemical known to induce MB after a single dose of 11 mg/kg, in conditions of isolation and starvation [3, 10]; (ii) a severe magnesium deficiency induced by 50 ppm magnesium-deficient diet which may also induce MB [11, 12] and a moderate magnesium deficiency (150 ppm magnesium-deficient diet) which does not induce MB by itself but provokes probably alterations of the central neurotransmission [12], (iii) injections of 11 mg THC/kg b.w. or lower doses, usually considered as non toxic, in rats suffering a severe or moderate magnesium deficiency.

According to Fujiwara [10] we showed that about 60% of naive Long Evans rats became muricidal 1 hour after a single injection of 11 mg THC/kg b.w.. As compared to NK, the first muricide attack was greatly disorganized, poorly efficient and was followed by a wild prolonged attack on the dead mouse, reflecting a THC-enhanced aggressiveness [6]. However, this MB pattern underwent changes throughout the six 1h-delayed consecutive MB assays, without further THC injection. On the 6^th assay, MB resembled that of NK, leading us to assume that a true killing behaviour was progressively acquired by 59% of the tested population (which is non killer in standard conditions) for the 6 successive assays whereas the aggressiveness disappeared with the physiological elimination of THC. Lower doses (2, 4, 8 mg/kg) assayed on the same rats fed a normal 1 700 ± 100 ppm magnesium diet failed to induce MB.

We reported that magnesium deprivation in rats may lead to two levels of magnesium deficiency according to the magnesium content of the diet (150 or 50 ppm Mg-deficient diets respectively) [11, 12]. In 150 ppm magnesium deficient rats, the disorders are linked to an insufficient magnesium intake and are reversed by oral physiological magnesium supplementation. In 50 ppm Mg-deficient group, the repetition of the assays leads to a magnesium depletion and the disorders, related to a dysregulation of the magnesium metabolism control, requires specific correction [18, 31] which is generally very difficult [18]. Usually, the various magnesium salts used in therapy are efficient during the treatment period but their protective effect generally disappears rapidly after the end of treatment [18]. In the present assays, all the 50 ppm magnesium-deficient rats became muricidal. Their MB was similar over the 6 successive assays and resembled that of NK with exception of the attack on the living mouse which was twice as long, possibly because of the psychomotor hyperexcitation linked to magnesium depletion resulting in less effective hunting. In contrast, 150 ppm magnesium-deficiency failed to induce MB, but could initiate neurotransmitter alterations without behavioural consequences.

As could be expected from the previous assays, THC injections in magnesium deficient mice resulted in severe behavioural damages. The lower doses, generally considered as non toxic, induced, in the 50 ppm magnesium-deficient group, severe behavioural alterations. As a whole, it appears that, in parallel with the repetition of the assays, leading in part to the elimination of THC and in part to a magnesium depletion, rats became “super killers”, unable to withstand the intruder more than 3 seconds, and killing it faster and faster. On the 6^th assay, the time elapsed from the introduction of the mouse and its death by the rat was less than 8 seconds. The second characteristic of the behaviour is that rats suffered true fury fits with prolonged attacks on the dead mouse, ten-times longer than in NK and 9-times longer than in untreated 50 ppm magnesium-deficient rat groups. This compulsive effect of the rats on the dead mouse can be attributed neither to hunger, since the rats were fed 3 hours before the experiment, nor to remaining blood or bones since rats were transferred after each assay in a new clean cage. Finally the association of doses of THC, as low as 2 mg/kg, to a magnesium depletion induced a dramatic increasingly psychotic behaviour characterized by a hyperaggressiveness with uncontrolled fury fits.

Even though slighly attenuated, the same pattern was observed when low THC doses were injected in 150 ppm magnesium-deficient rats. As a whole, the MB pattern was all the more severe as the magnesium deficiency was severe and the THC dose was higher. A single 11 mg THC /kg b.w. was so disturbing that the rats did not attack the mouse so quickly and efficiently as with the lower doses, but the attack on the dead mouse still increased, in dramatic conditions. It may be assumed that the neurotransmission alteration was so important that the mouse did not initiate instantaneous reactions. However, the increase in the attack on the dead mouse to values higher than 10 minutes indicate a severe nervous damage with a terrifying hyperaggressiveness, which, at least for lower doses did not disappear with THC elimination, but, on the contrary, increased with the repetition of the stresses.

Conclusion

It appears that low magnesium levels may reveal the neurotoxicity of low doses of THC and reciprocally that low doses of THC may be at risk in rats suffering a severe but also a moderate magnesium deficiency. The observed neurotoxic potentialisation is probably linked to parallel neurotransmitter alterations, including for instance serotonin, catecholamines and/or GABA [26-30].

Finally, a moderate magnesium deficiency linked to an insufficient nutritional intake is relatively frequent in human beings (about 20% in the total population) [31]. Consequently, in view of our results in rats, it may be assumed, that, if the same occurs in humans, this deficiency may aggravate in cannabis users both immediate damage and long term psychic sequelae induced by THC.

References

1. Karli P. The Norway rat's killing response to the white mouse: An experimental analysis. Behaviour 1956; 10: 81-103.

2. Valzelli L. Psychopharmacology of aggression: An overview. Int Pharmacopsychaitry 1981; 16: 39-48.

3. Valzelli L, Garattini S. Biochemical and behavioural changes induced by isolation in rats. Neuropharmacology 1972; 11: 17-22.

4. Hatch AM, Balazo T, Wiberg CS, Grice HC. Long term isolation stress in rats. Science 1963; 142: 507-8.

5. Vergnes M. Induction du comportement d'agression Rat-Souris par la p-chlorophenylalanine: rôle de l'amygdale. Physiol Behav 1980; 25: 353-6.

6. Palermo Neto J, Carlini EA. Aggressive behaviour elicited in rats by Cannabis sativa: effects of p-chlorophenylalanine and DOPA. Eur J Pharmacol 1972; 17: 215-20.

7. Ueki S, Fujiwara M, Ogawa N. Mouse-killing behaviour (muricide) induced by delta-9-tetrahydrocannabinol in the rat. Physiol Behav 1972; 9: 585-7.

8. Miczeck KA. Does THC induce aggression ¿ Suppression and induction of aggressive reactions by chronic and acute Δ9-THC treatments in laboratory rats. In: Braude MC, Szara S, eds. Pharmacology of marihuana, vol. 2. New York: Raven Press; 1976: 499-514.

9. Segawa T, Bando S, Hosakawa M. Brain serotonin metabolism and Δ9-tetrahydrocannabinol-induced muricidal behaviour in rats. Jpn J Pharmacol 1977; 27: 581-2

10. Fujiwara M, Ueki S. The course of aggressive behaviour induced by a single injection of Δ9-tetrahydrocannabinol and its characteristics. Physiol Behav 1979; 22: 535-9.

11. Bac P, Pagès N, Herrenknecht C, Paris M. Measurement of the three phases of muricide behaviour induced by Δ9-tetrahydrocannabinol in isolated, fasting rats. Physiol Behav 1998; 63: n^o 5, 815-20.

12. Bac P, Pagès N, Herrenknecht C, Dupont C, Maurois P, Vamecq J, Durlach J. THC aggravates rat muricide behavior induced by two levels of magnesium deficiency. Physiol Behav 2002; 77: 189-95.

13. Carlini EA, Hamaoui A, Martz MW. Factors influencing the aggressiveness elicited by marihuana in food deprived rats. Br J Pharmacol 44: 794-804.

14. Vorhees CV. Thiamin deficiency induced muricide behaviour in rats. Physiol Behav 1979; 23: 211-4.

15. Onodera K, Kisara K, Ogura Y. Effect of 5-hydroxytryptophan on muricide response induced by thiamin deficiency. Arch Int Pharmacodyn 1979; 240: 220-7.

16. Onodera K, Ogura K, Kisara K. Characteristics of muricidal induced by thiamin deficiency and its suppression by antidepressants or intraventricular serotonin in the rat. Brain Res 1981; 169: 139-53.

17. Gibbons JL, Barr GA, Bridger WH, Leibowitz JF. Manipulations of dietary tryptophan: effects on mouse killing and brain serotonin in the rat. Brain Res 1979; 169: 139-53.

18. Bac P, Pagès N, Herrenknecht C, Teste JF. Inhibition of mouse-killing behaviour in magnesium-deficient rats: effect of pharmacological doses of magnesium pidolate, magnesium aspartate, magnesium lactate, magnesium gluconate and magnesium chloride. Magnes Res 1995; 8: 37-45.

19. Ueki S, Nurimoto S, Ogawa N. Characteristics in emotional behaviour of the rat with bilateral olfactory bulb ablations. Folia Psychiatr Neurol Jap 1972; 26: 227-37.

20. Grant LD, Cosuna DV, Grossman SP, Freedman DX. Muricide after serotonin depleting lesions of midbrain raphe nuclei. Pharmacol Biochem Behav 1977; 1: 105-10.

21. Yamamoto T, Ueki S. Characteristics in aggressive behaviour induced by midbrain raphe lesions in rats. Physiol Behav 1977; 19: 105-10.

22. Gomita Y, Ueki S. Effects of long term isolation on aggressive behaviour and excitability of the rat with olfactory bulbectomy. Life Sci 1979; 24: 2455-60.

23. Albert DJ, Walsh ML, Ryan J, Siemens Y. Mouse killing in rats: a comparison of spontaneous killers and rats with lesions of the medial hypothalamus or the medial accumbens nucleus. Physiol Behav 1982; 29: 989-94.

24. Shibata S, Yamamoto T, Ueki S. Differential effects of medial, central and basolateral amygdaloid lesions on four models of experimentally-induced aggression in rats. Physiol Behav 1982; 28: 289-94.

25. Frances H, Monier C, Henrotte JG. Behavioural differences between mice selected for low (MGL) and high (MGH) red blood cell magnesium levels. Magnes Res 1993; 6, 3: 305-6.

26. Valzelli L. Serotoninergic inhibitory control of experimental aggression. Pharmacol Res Commun 1982; 14: 1-13.

27. Abe Y, Tadano T, Yonezawa A, Kisara K. Suppressive effect of intraventricular injected Dopamine and Nomifensine on muricide induced by thiamine deficiency. Pharmacol Biochem Behav 1987; 26: 77-82.

28. Barr GA, Gibbons J, Bridger WH. Neuropharmacological regulation of mouse killing by rat. Behav Biol 1976; 17: 143-59.

29. Durlach J, Bac P, Bara M, Guiet-Bara A. Physiopathology of symptomatic and latent forms of central nervous hyperexcitability: a current general scheme. Magnes Res 2000; 13, 4: 293-303.

30. Garcia-Cairasco N, Oliveira JAC, Wakamatsu H, Bueno STB, Guimaraes FS. Reduced exploratory activity of audiogenic seizures susceptible rats. Physiol Behav 1998; 64: 671-4.

31. Durlach J. (2001): Importance and clinical forms of chronic primary magnesium deficiency in human beings. In Advances in Magnesium Research: Nutrition and Health. Eds Y. Rayssiguier, A. Masur & J. Durlach, pp 13-20. Londres: John Libbey.