Secondary Logo

Journal Logo


Intrathecal anaesthesia alters intracellular Ca2+/Mg2+ homeostasis in the spinal cord neurones of experimental rats

Bahar, M.*; Berman, S.; Chanimov, M.*; Weissgarten, J.; Averbukh, Z.; Cohen, M. L.*; Grinshpon, Y.*

Author Information
European Journal of Anaesthesiology: April 2001 - Volume 18 - Issue 4 - p 231-237
  • Free



The mechanism of action of local anaesthetics used for epidural and subarachnoid spinal anaesthesia has been of great theoretical and practical interest to clinicians for a long time. The effects of local anaesthetics on the peripheral nerve system and on sensory ganglion neurones have been amply investigated [1,2]. On the other hand, their impact on spinal cord cells has not yet been fully elucidated [3]. It is well accepted that during subarachnoid spinal and epidural anaesthesia, drug molecules diffuse rapidly into the spinal cord [3,4]. The cells of the spinal cord become extensively exposed to anaesthetic drugs, thus changing their membrane permeability and neuronal excitability [3,4].

In recent years, ample in vitro evidence has accumulated concerning the importance of free Mg2+ ions for a variety of living cell functions, including neural cell signal transmission [5, 6, 7, 8, 9]. In our previous investigations, we examined the in vivo role of Mg2+ in intrathecal anaesthesia. We were able to induce a state of sustained spinal anaesthesia and general sedation with both sensory and motor block of the hind limbs in experimental rats by intrathecal administration of 6.3% MgSO4 in a 20-μL bolus [10]. The effect was stable and long lasting, approximately 1 h post injection [11]. When haemodynamics of electrolytes after intrathecal MgSO4 administration were studied, a modest rise in serum Mg2+ and no statistically significant changes in serum Ca2+, Na+ or K+ concentrations were found [11].

Serum electrolyte content is by no means a reliable index of intracellular electrolyte status. Serum levels might remain within normal range, despite their intracellular concentrations being far from normal [12,13]. The aim of our present investigation was therefore to study the dynamics of intracellular electrolyte content in spinal cord neurones of experimental rats, after intrathecal anaesthesia induced by MgSO4. Lidocaine, a different, non-electrolyte, and more conventional anaesthetic agent, was also examined in the same experimental setting.

Because our main goal was to examine the longitudinal dynamics of intracellular electrolyte balance rather than transient transmembrane flux shifts and to determine the total intracellular Ca2+ and Mg2+ concentrations with maximal accuracy and precision, the atomic absorption spectrophotometry methods were employed for this investigation.

Materials and methods

Animals and surgical procedures

This study received the approval of the Ethics Committee of the Assaf Harofeh Medical Center and was carried out according to regulations of the local Committee for Experimentation on Animals. Eighty, 2-month-old, Sprague-Dawley rats (weighing 250-300g) were used in this study. Ten animals were sacrificed for preliminary pilot procedures: to select the optimal time intervals for sacrifice after MgSO4 and lidocaine injections, and to establish methodology of spinal cord cell isolation for subsequent intracellular electrolyte determinations.

Injection procedures

Spinal cannulation and injection procedures were performed according to a protocol described in our previous studies [10,11]. In the present study, the catheters were inserted at the lumbar 2-3 level and slid rostrally for a distance of 20-25 mm. The correct location of the catheter tip in the subarachnoid space was verified by seeing cerebrospinal fluid issuing through the lumen of the catheter. This was confirmed 24 h before the experiment by a standard procedure of injecting a bolus of 20μL 2% lidocaine sufficient to cause transient hind limb paralysis. This volume and concentration of lidocaine serve as reliable indices to determine the right location of the catheter tip [10,11]. The animals received intrathecal injections of a 20 μL bolus of 6.3% MgSO4, 4% lidocaine or 0.9% NaCI (normal saline) as a control injection. Osmolarity of 6.3% MgSO4 solution was 286 mOsm kg−1, and of 4% lidocaine was 281 mOsm kg−1, both iso-osmolar to rat plasma.

Experimental design

Study groups

Control group.

Ten animals were injected with 20 μL of 0.9% saline. All were sacrificed by overdose of halothane 30 min after the intrathecal injection. In our preliminary studies, different time intervals were proven to be irrelevant for the control group intracellular electrolyte content. The spinal cords were removed immediately after sacrificing and placed in separate Petri dishes in phosphate-buffered saline (PBS, pH 7.4), until clear of blood.

MgSO4 injection.

Three groups of experimental rats, 10 animals in each, were given 20μL of 6.3% MgSO4. After 30 min, 2 h and 24h, respectively, the rats were sacrificed by overdose of inhaled halothane. Their spinal cords were immediately removed and placed in separate Petri dishes in PBS, as already described.

Lidocaine injection.

Three additional groups of animals, 10 per group, were administered 20μL of 4% lidocaine. The animals were sacrificed 15 min, 30 min and 1 h after the injection. Their spinal cords were also removed immediately and placed in separate Petri dishes in PBS solution.

Spinal cord cell isolation for intracellular electrolyte measurement

Spinal cords were removed intact by cutting the vertebrae with scissors. The portion of the spinal cord adjacent to the tip of the catheter and for a distance of 5mm above and below this point was excised and discarded, to eliminate the possible traumatic effect of the injection. The spinal cord above the excised segment and the cord below this point were then submitted separately for analyses. The tissue samples (mean weight 200-250 μg) were repeatedly washed in PBS prepared without Mg2+ or Ca2+ salt addition. When clear of blood, the samples were cleaned of small vessels and connective tissue, incubated for 20 min in 0.1% collagenase, placed in separate Petri dishes in 3mL PBS and minced through a 100μ stainless-steel mesh. The isolated cell samples were collected in 15-mL polystyrene conical test tubes and washed twice by centrifugations at 600g for 15 min. The cell viability was assessed by 0.1% Eosin exclusion. Only samples with viability exceeding 95% were used for subsequent electrolyte determinations. One millilitre of ultrapure (atomic absorption spectrophotometry-AAS-purity grade) concentrated nitric acid was added to each cell pellet. The samples were left overnight at room temperature and then finally digested by boiling the test tubes in a water bath at 90°C for 1 h.

Intracellular electrolyte determination

Ca2+ and Mg2+ concentrations were measured in an atomic absorption spectrophotometer (SpectrAA 800, Varian, Australia). Matrix diluent was prepared as follows: 5g lanthanum oxide were dissolved in 80 mL of concentrated hydrochloric acid and diluted with water to 1 L. All the chemicals were of AAS purity grade, and only reverse osmosis purified water was used for all preparations. Standard solutions were prepared by serial dilutions from stock CaCl2 and MgCl2 1 mg dL−1 solutions, using the same matrix diluent. Standard curves were built so as to fit the expected concentration ranges evaluated in our preliminary studies. The assay included background correction and validation of the results with internal and external quality controls. Na+ and K+ concentrations were determined by flame photometry methods. Total protein was determined using the Bradford assay [14].

Data presentation and statistical analysis

The results were presented as the means ± SD of 10 experiments, in ng mg−1. The statistical differences between the results were evaluated by the ANOVA Kruskall-Wallis test.


The results are presented graphically in Figs 1, 2, 3 and 4. Figures 1 and 2 show the mean Mg2+, Ca2+, Na+ and K+ concentrations in the thoracic and lumbar segments of the spinal cord after the MgSO4 injection. Figures 3 and 4 demonstrate the respective results obtained after the lidocaine administration.

Fig. 1.
Fig. 1.:
Total intracellular electrolyte concentration in thoracic spinal cord neurones after intrathecal anaesthesia by 6.3% MgSO4 solution. Ca, calcium, in ng mg−1 total protein; Mg, magnesium, in ng mg−1 total protein; Na, sodium, in ng × 10−2 mg−1 total protein; K, potassium, in ng × 10−4 mg−1 total protein; 30 min, 2 h, 24 h, time intervals after intrathecal MgSO4 injection.
Fig. 2.
Fig. 2.:
Total intracellular electrolyte concentration in lumbar spinal cord neurones after intrathecal anaesthesia by 6.3% MgSO4 solution.
Fig. 3.
Fig. 3.:
Total intracellular electrolyte concentration in thoracic spinal cord neurones after intrathecal anaesthesia by 4% lidocaine.
Fig. 4.
Fig. 4.:
Total intracellular electrolyte concentration in lumbar spinal cord neurones after intrathecal anaesthesia by 4% lidocaine.

Figure 1 presents the dynamics of electrolyte balance in the thoracic portion of the spinal cord of rats injected with 6.3% MgSO4. A statistically significant rise in total intracellular Ca2+ is evident after 30′ (172±24.4 ng mg−1 protein vs. control 23±3.1 ng mg−1 protein, P<0.05). It remains significantly elevated 2 h later, and after 24 h is still significantly higher than in the control group, which was subjected to sham ‘anaesthesia’ by saline injection (121 ± 14.1 and 108 ± 11.9 ng mg−1 protein vs. control 23± 3.1 ng mg−1 protein, P<0.05). A rise in total intracellular Mg2+ concentration is also apparent 30 min post injection (34.9 ± 6.6 and 34.0 ± 6.1, respectively, vs. control 21 ± 5.8 nm mg−1 protein, P=NS), but the difference reaches statistical significance only after 24 h (49 ± 5.4 vs. 21 ± 5.8 ng mg−1 protein, P<0.05). No statistically significant changes in Na+ or K+ content are evident, though the concentrations tend to increase 30 min after the injection [(56 × 10−2 ± 19.1 × 10−2) ng mg−1 protein vs. control (30 × 10−2± 11.3 × 10−2) ng mg−1 protein] and then decline gradually to 36 × 10−2± 10.9 × 10−2 ng mg−1 protein (P=NS in each comparison).

Figure 2 shows changes in intracellular electrolyte status in the lumbar segment of the spinal cord after MgSO4 injection. A significant rise in Ca2+ is evident 30 min after MgSO4 administration, and is sustained after 2 h and 24 h (222 ± 23.2, 229 ± 18.8 and 176 ± 14.9, respectively, vs. control 43±8.1 ng mg−1 protein, P<0.05 in each comparison). A slight rise in Mg2+ concentration is detectable in this part of the spinal cord, though it does not reach statistical significance (27 ± 4.9 ng mg−1 protein vs. 23 ± 4.3 ng mg−1 protein, P=NS). No significant changes in Na+ and K+ are evident 30′ after the injection, the decrease becoming detectable 24 h later. At any given time interval Na+, K+, Ca2+ and Mg2+ content in the lumbar segment (Fig. 2) is not different from that in the thoracic portion of the spinal cord (Fig. 1).

Figures 3 and 4 show the results obtained in spinal cord cells after 4% lidocaine injection. Taking into consideration the different time responses to lidocaine, as well as its shorter life span, we selected different time intervals for electrolyte status determinations after lidocaine injection, in order to match anaesthetic activities of the two agents pharmacokinetically.

Figure 3 shows the changes in electrolyte status of the thoracic cells after lidocaine injection. Similarly to the results obtained with MgSO4, total intracellular Ca2+ is significantly elevated 15 min and 30 min after the injection and is still significantly different from the control value 1h later (69±7.3, 64±4.8 and 53 ± 4.6 ng mg−1 protein vs. control 33.4 ± 2.6 ng mg−1 protein, P<0.05 in each comparison). Mg2+ tends to increase 15 min after the injection, though no statistically detectable changes in Mg2+ concentration are apparent (27±4.1 vs. 23±3.8 ng mg−1 protein, P=NS). Na+, on the other hand, declines gradually and significantly from control 23.3× 10−2 ng mg−1 protein to 11 × 10−2 ng mg−1 protein, after lidocaine administration (P<0.05). K+ content is also gradually decreased 15 min and 30 min after 4% lidocaine injection, though this decline reaches statistical significance only after 1 h.

Figure 4 shows the effect of lidocaine injection on the lumbar part of the spinal cord. Thirty minutes after the injection, intracellular Ca2+ rises significantly above the control value (94±6.9 ng mg−1 protein vs. 23±3.8 ng mg−1 protein, P< 0.01) and then decreases slowly after 1 h (46±7.7 ng mg−1 protein). No statistically significant changes in Mg2+ content are evident (23.5 ± 3.9 ng mg−1 protein at the end of the study vs. control 22.5±4.1 ng mg−1 protein, P=NS). Na+ and K+ concentrations do not change significantly, though tend to be lower than control values 1 h after the 4% lidocaine injection.


In our previous investigations performed on chronically cannulated rats, we were able to establish a steady state of spinal anaesthesia by intrathecal injection of 6.3% MgSO4 in a 20-μL bolus [10,11]. We achieved a reversible sensory and motor blockade of the lower half of thetrunk and hind limbs, sedation and non-reactivity to pain. We were also able to demonstrate an absence of neurotoxicity on histological examination after serial MgSO4 injections [10,11]. The effects were sustained for approximately 1 h after the MgSO4 administration and enabled the performance of abdominal surgery on experimental rats without pain [10,11]. Furthermore, we studied the effect of intrathecally administered MgSO4 on serum concentrations of Mg2+, Ca2+, Na+ and K+ ions. Serum Mg2+ levels were slightly increased 1h to 2h after MgSO4 injection. No significant changes in serum Ca2+, Na+ or K+ were evident [11]. Apparently, these data demonstrate that intrathecally administered MgSO4 has little or no effect on electrolyte homeostasis. Recent investigations and our own experience suggest that this is not always the case. Serum electrolyte levels often prove to remain within the normal reference range in spite of significant intracellular changes [12,13]. Thus, in the present investigation, our goal was to study the dynamics of electrolyte status in spinal cord cells after intrathecal administration of 6.3% MgSO4 and to compare it with the effect of 4% lidocaine, a more conventional intraspinal anaesthetic of non-electrolyte origin. The first approach to total intracellular measurement of Ca2+ in nervous system cells was recently made by Golub and his colleagues [15] utilizing a costly and complicated method of inductively coupled plasma spectrochemical analysis. To the best of our knowledge, ours is the first report of a method for intracellular Ca2+ and Mg2+ determination in spinal cord neurones by atomic absorption spectrophotometry. Contrary to our expectations, the results obtained after MgSO4 administration did not demonstrate significant changes in intracellular Mg2+ concentration. The increase in intracellular Mg2+ was short-lived and did not reach statistical significance. However, total intracellular Ca2+ rose dramatically and, while decreasing in the later stages of the experiment, still demonstrated a statistically significant difference from control values. Four per cent lidocaine administration produced a very similar effect. Lidocaine, a widely used conventional anaesthetic agent, was chosen for this study as one sharing some common patterns of action with MgSO4, namely producing sensory and motor block, sedation and non-reactivity to pain after spinal intrathecal injection. Nevertheless, lidocaine differs from MgSO4 by its non-electrolyte character and much shorter period of action, thus being the drug of choice for local anaesthesia when a strong and prompt effect is desired. Similarity of effects after both MgSO4 and lidocaine administration tempts to suggest a common pathway of action, starting with massive entrance of Ca2+ to the spinal cord cells. The thought is not new. Several mechanisms have already been proposed at the nerve cell membrane level [16, 17, 18, 19, 20]. For example, Ca2+ influx through voltage-gated Ca2+ channels is considered to be a common starting mechanism of action for all opioid analgesics [21]. It has been proven that, seconds after acute treatment, opioids produce a rapid fall in neuronal Ca2+ levels. Then, an adaptive gradual increase in intracellular Ca2+ takes place. This elevation of intracellular Ca2+ is thought to downregulate the acute opiate effects, i.e. to oppose tolerance development [21]. Moreover, elevated intracellular Ca2+ is thought to play a certain role as an additional analgesic [21]. With respect to spinal subarachnoid anaesthesia, local anaesthetic agents proved to exert their pain-suppressive activity through voltage-sensitive ion channels [1, 2, 3,18]. Lidocaine, bupivacaine and others suppress pain transmission by blocking voltage-gated Na+ and K+ channels in peripheral nerves as well as those in other nerve cell types [8,19,20]. Bupivacaine, mepivacaine and lidocaine affect voltage-dependent Na+ and K+ channels in the dorsal horn neurones of the spinal cord [19,20]. All these in vitro findings are in keeping with our present results obtained in vivo, namely, the gradual decline in total intracellular Na+ and K+ content after both lidocaine and MgSO4 administration, eventually resulting in a depletion of their intracellular stores. Less is known about the mode of action of conventional local anaesthetics as well as MgSO4 where voltage-sensitive Ca2+ channels are involved. It has been well established that most mammalian cells, including the cells of the nervous system, possess Mg2+ transport mechanisms [1, 2, 3]. These were shown to be effective and rapid [22], whereas the subsequent rise of intracellular Mg2+ proved to be moderate and transient [22,23]. The identity and mechanisms of action of Mg2+ transporters are still not clear. It has been shown, though, that Mg2+ does regulate Ca2+ ion entry into the neural cells through voltage-dependent Ca2+ ion channels and that after changes in Mg2+ concentrations in extracellular fluid, no significant corresponding change in intracellular Mg2+ concentration occurs [8]. In a different study, Kato and his colleagues observed transient elevation of Mg2+ after depolarization of cultured dorsal root ganglion neurones which was accompanied by a significant rise of intracellular Ca2+ and could be inhibited in a Ca2+-depleted environment [9]. Intracellular Mg2+ peak was sharp, short-lived and transient, while the following Ca2+ increase demonstrated a broad time course plateau with a very slow recovery to the resting state [9]. The findings of our present investigation, namely, a significant and prolonged rise in total intracellular Ca2+ content accompanied by a very modest and delayed Mg2+ elevation, are consistent with these in vitro observations. According to another line of in vitro investigations, Mg2+ may affect Ca2+ homeostasis in nervous tissue cells by yet another mechanism, namely through ligand-gated Ca2+ ion channels, by a selective activation of NMDA (N-methyl-D-aspartate) receptors. It is usually assumed that Mg2+ acts as a blocker of Ca2+ permeable NMDA-gated channels [24, 25, 26]. Removal of extracellular Mg2+ induces cultured rat cortical neurones to undergo Ca2+ oscillations. These oscillations are carried on through NMDA receptors and blocking the latter would inhibit Ca2+ fluxes [27]. However, in a recent study, Reichling et al. demonstrated an intracellular Ca2+ increase in the neuronal cells from rat dorsal horn after activation of NMDA receptors in the presence of physiological or even slightly increased extracellular Mg2+ concentrations [28].

In conclusion, the results of our present study demonstrate a sustained total intracellular Ca2+ elevation apparently provoked by a transient increase in Mg2+ and accompanied by a gradual Na+ and K+ depletion in the spinal cord neurones after intraspinal anaesthesia by MgSO4 and lidocaine. According to these findings and in agreement with numerous in vitro studies, a suggestion of a possible common mechanism of induction of intrathecal anaesthesia, resembling, though not identical with that proposed for opioid drugs [21], seems to be plausible. Significant and sustained increase in intracellular Ca2+ could, at least in part, explain the motor and sensory blockade effect exerted not only by MgSO4 [11] and lidocaine, but by all local anaesthetic drugs without exception used for spinal subarachnoid and epidural anaesthesia [1]. In our previous investigations using MgSO4 for intrathecal injection, we observed, together with motor and sensory block, also analgesic and sedative effects [10,11]. At this point, one could only speculate about the role of increased intracellular Ca2+ with respect to these observations.


1 Evers SA. Cellular and molecular mechanisms of anaesthesia. In: Barach PG, ed. Clinical Anaesthesia, 3rd edn. Philadelphia: Lippincott-Raven, 1996: 119-133.
2 Butterworth JF, Strichartz GR. Molecular mechanisms of local anesthesia: a review. Anesthesiology 1990; 72: 711-374.
3 Dishewski A, Hempelmann G, Vogel W, Safronov BV. Blockade of Na+ and K+ currents by local anaesthetics in the dorsal horn neurons of the spinal cord. Anaesthesiology 1998; 88: 172-179.
4 Urban BJ. Clinical observations suggesting a changing site of action during induction and recession of spinal and epidural anaesthesia. Anaesthesiology 1973; 39: 496-450.
5 Anderson W, Lewis D, Swartzwelder H, Wilson W. Magnesium-free medium activates seizure-like events in the rat hippocampal slices. Brain Res 1986; 398: 215-219.
6 El-Behery Puil E. Effects of hypomagnesemia on transmitter actions in neocortical slices. Br J Pharmacol 1990; 101: 1006-1010.
7 Grubbs R, Maguire M. Magnesium as a regulatory cation: criteria and evaluation. Magnesium 1987; 1: 113-127.
8 Zhang A, Shou-Hui-Fan Cheng TPO, Altura BT, Wong RKS, Altura BM. Extracellular Mg++ modulates intracellular Ca++ in acutely isolated hippocampal CA1 pyramidal cells of the guinea-pig. Brain Res 1996; 728: 204-208.
9 Kato H, Gotoh H, Kajikawa M, Suto K. Depolarization triggers intracellular magnesium surge in cultured dorsal root ganglion nerves. Brain Res 1997; 779: 329-333.
10 Bahar M, Rosen M, Vickers MD. Chronic cannulation of the intradural or extradural space in the rat. Br J Anaesth 1984; 56: 405-410.
11 Bahar M, Chanimov M, Grinshpun Y, Koifman I, Cohen ML. Spinal anaesthesia induced by intrathecal magnesium sulfate. An experimental study in a rat model. Anaesthesia 1996; 51: 627-633.
12 Stendig-Lindberg G, Harsat A, Graff E. Magnesium content of mononuclear cells, erythrocytes and 24 hurine in carefully screened apparently healthy Israelis. Eur J Clin Chem Clin Biochem 1992; 30: 833-836.
13 Abraham AS, Meshulam Z, Rosenmann D, Eylath U. Influence of chronic diuretic therapy on serum, lymphocyte and erythrocyte potassium, magnesium and calcium concentrations. Cardiology 1988; 75: 17-23.
14 Bradford M. A rapid and sensitive method for the quantification of microgram quantities of protein using the principle of protein-dye binding. Ann Biochem 1976; 72: 248-254.
15 Golub M, Bin Han Keen C. Developmental patterns of aluminium and five essential elements in the central nervous system of the fetal and infant guinea pigs. Biol Trace Element Res 1996; 55: 241-251.
16 Judge S. Effect of general anaesthetics on synaptic ion channels. Br J Anaesth 1983; 55: 191-200.
17 Miller KW. The nature of the site of general anaesthesia. Int Rev Neurobiol 1981; 27: 1-61.
18 Maclver MB, Kendig JK. Anaesthetic effects on resting membrane potential are voltage-dependent and agent specific. Anaesthesiology 1991; 74: 83-88.
19 Guo X, Castle N, Chernoff D, Strichartz G. Comparative inhibition of voltage-gated cation channels by local anaesthetics. Ann NY Acad Sci 1991; 625: 181-199.
20 Castle NA. Bupivacaine inhibits the transient outward K+ current but not inward rectifier in rat ventricular myocytes. Pharmacol Exp Ther 1990; 255: 1038-1046.
21 Way EL. Future vistas. In: Kuhar MJ, Pasternak GW, eds. Analgesics: Neurochemical, Behavioral and Clinical Perspectives. New York: Raven Press, 1984: 313-333.
22 Cefaratti C, Romani A, Scarpa A. Characterization of two Mg++ transporters in sealed plasma membrane vesicles from rat liver. Am J Physiol 1998; 275: C995-C1008.
23 Terada H, Hayashi H, Noda N, Saton H, Katon H, Yamasaki N. Effects of Mg++ on Ca++ waves and Ca++ transients on rat ventricular monocytes. Am J Physiol 1996; 270: H907-H914.
24 Mayer M, Westbrook G, Guthrie P. Voltage-dependent block by Mg++ of NMDA responses in spinal cord neurons. Nature 1984; 309: 261-263.
25 Mayer M, Westbrook G. Permeation and block of N-methyl-D-aspartic acid receptor channels by divalent cations in mouse cultured central neurons. J Physiol 1987; 394: 501-527.
26 Nowak L, Bregestovski P, Ascher P, Herbet A, Prochiantz A. Magnesium gates glutamate-activated channels in mouse central neurons. Nature 1984; 307: 462-465.
27 Wang X, Gruenstein E. Mechanism of synchronized Ca++ oscillations in cortical neurons. Brain Res 1997; 767: 239-249.
28 Reichling D, MacDermott AB. NMDA receptor mediated calcium entry in the absence of AMPA receptor activation in rat dorsal horn neurons. Neurosci Lett 1996; 204: 17-20.

ANAESTHESIA CONDUCTION; ANAESTHESIA SPINAL; HOMEOSTASIS; water-electrolyte imbalance; ION TRANSPORT; calcium signalling; magnesium signalling; SPECTROPHOTOMETRY; atomic absorption

© 2001 European Society of Anaesthesiology