Anticonvulsant effects of magnesium sulfate on hippocampal seizures: Therapeutic implications in preeclampsia-eclampsia David B. Cotton, MD, Cynthia A. Jannsz, BS, and Robert F. Berman, PhD Detroit, Michigan OBJECTIVE: The objective of this study was to determine whether magnesium sulfate has central anticonvulsant effects. STUDY DESIGN: In three experiments we investigated the anticonvulsant properties of magnesium sulfate on the hippocampus because of its high density of N-methyl-D-aspartate receptors and link to clinical epilepsy. Seizure activity was elicited in the hippocampus of rats or in in vitro hippocampal brain slices. The effects of magnesium on seizure activity were determined. RESULTS: (1) Intraperitoneal magnesium sulfate (270 or 360 mg/kg) failed to block hippocampal seizures but reduced electroencephalographic amplitude and seizure duration. (2) Injection of magnesium sulfate directly into a seizure focus blocked seizures and elevated seizure thresholds. (3) Low magnesium levels produced epileptiform activity in an in vitro hippocampal brain slice preparation; the epileptiform activity was reversed by increasing the magnesium concentration. CONCLUSIONS: Magnesium sulfate has central anticonvulsant activity on hippocampal seizures, implicating the N-methyl-D-aspartate receptor in eclamptic seizures and in the therapeutic efficacy of magnesium sulfate. (AM J OBSTET GVNECOL 1992;166:1127-36.)
Key words: Magnesium sulfate, hippocampus, seizures, eclampsia
Magnesium sulfate has been used for >60 years to prevent and treat eclamptic seizures. I -3 In spite of magnesium sulfate's empiric success in obstetrics, the literature of other medical disciplines has decried its use on the grounds that it does not have any central inhibitory action and merely masks seizures via peripheral neuromuscular blockade!-7 However, on the basis of the well-documented clinical response of eclamptic seizures to magnesium sulfate in the absence of any neuromuscular blockade, Pritchard" has proposed that magnesium sulfate does have central effects. Experimental work in dogs, cats, and subhuman primates also provides support for a central mechanism of action. Borges and Gucer" showed that intravenous magnesium sulfate infusion suppressed neuronal burst firing and decreased interictal spike frequency in a penicillin-induced epilepsy model. Suppression occurred at serum magnesium levels well below those required for complete neuromuscular blockade; Borges and Gucer concluded that magnesium sulfate does have a central effect. In a similar model of penicillin-induced epileptic foci in cats Koontz and Reid!O were unable to duplicate From the Department of Obstetrics and Gynecology, Wayne State University-Hutzel Hospital. Presented by invitation at the Tenth Annual Meeting of the Gynecological and Obstetrical Society, Carlsbad, California, September 57, 1991. Reprint requests: David B. Cotton, MD, Hutzel Hospital-Wayne State University, Department of Obstetrics and Gynecology, 4707 St. Antoine, Detroit, MI 48201. 616135626
Borges and Gucer's9 findings. In Koontz and Reid's model a saline-infused group had the same rapid decrement in seizure activity as the group receiving magnesium sulfate. Koontz and Reid II later performed an experiment with the same seizure model to determine the effect of pretreatment with magnesium sulfate; again, they could not document any anticonvulsant effect of magnesium sulfate. They concluded that parenteral magnesium sulfate had no demonstrable anticonvulsant effect in the penicillin-induced seizure focus model in cats. In another experiment, Hilmy and Somjen l2 demonstrated a fourfold to sevenfold greater deposition of magnesium in peripheral tissues as opposed to brain tissues after intravenous magnesium administration. These investigators concluded that magnesium acts primarily through neuromuscular blockade and that a barrier mechanism protects the central nervous system from high levels of magnesium. Others have suggested that in eclampsia any apparent central effects of magnesium sulfate must be the result of disruption of the blood-brain barrier and leakage of magnesium into the brain. 4 • !O Donaldson 4 expanded on these observations and concluded that magnesium sulfate does not treat eclamptic convulsions and adds substantial risks to mother and baby. Because of these disparate conclusions regarding magnesium sulfate's mechanism of action, we examined in rats the effects of magnesium sulfate on seizures initiated in the hippocampus, a brain region with a low seizure threshold and a high density of N-methyl-o1127
1128 Cotton, Janusz, and Berman
aspartate receptors. A study of hippocampal seizures should provide a sensitive index of the possible central nervous system effects of magnesium sulfate. The high density of N-methyl-n-aspartate receptors is also of interest because these receptors have been linked to various models of epilepsy and are blocked by magnesium." Material and methods Surgical preparation. Male Long-Evans rats (Harlan Sprague-Dawley Inc., Indianapolis) were individually housed in stainless steel cages in an environmentally controlled vivarium under 12-hour light-dark cycles. Animals had free access to food throughout the experiments. Twelve animals (275 to 300 gm) were surgically anesthetized (60 mg/kg sodium pentobarbital intraperitoneally) and stereotaxically implanted with chronically indwelling bipolar electrodes (MS-303/2, Plastics One Inc., Roanoke, Va.) aimed for the dorsal hippocampus. The stereotaxic coordinates were 4.0 mm posterior bregma, 2.0 mm lateral from midline, and 4.0 mm ventral to the superior surface of the level skull. The electrode assembly was chronically attached to the skull with three stainless steel jeweler's screws and acrylic cement. Animals were allowed 1 week of postsurgical recovery before further treatments were begun. An additional group of 5 male Long-Evans rats (275 to 300 gm) were surgically implanted with chemitrodes into the dorsal hippocampus as described above. The chemitrode, which permitted electric stimulation and intracranial drug injection at the same brain site, consisted of a bipolar electrode (MS303/3, Plastics One Inc.) cemented to a 23-gauge stainless steel cannula. A 30-gauge obturator plugged the cannula until the time of injection. The ends of the electrode extended 0.5 mm past the end of the cannula. The coordinates for implantation were the same as those described above. A I-week postsurgical recovery period was again used. A final group of 10 male Long-Evans rats (200 to 260 gm) were used for the in vitro hippocampal brain slice experiments as described below. Experimental design and protocol Experiment 1. In this experiment the effects of intraperitoneal injections of magnesium sulfate at 270 and 360 mg/kg on hippocampal epileptiform afterdischarges were determined. Rats implanted with bipolar electrodes in the dorsal hippocampus were placed into a Plexiglas acrylic plastic recording ch:lmber and leads from a polygraph (model 7, Grass Instruments Co., Quincy, Mass.) were attached to the implanted electrodes. Baseline electroencephalographic activity was recorded from the hippocampus via the implanted electrodes. One-second trains (100 Hz, I-millisecond pulse duration, biphasic, square waves) of low-intensity electric stimulation were delivered
April 1992 Am J Obstet Gynecol
through the implanted electrodes beginning at a current intensity of 10 /LA. Stimulation intensity was increased in steps of 10 /LA until a brief, 5- lO-second epileptiform afterdischarge was recorded from the hippocampus immediately after stimulation. The level of stimulation that evoked an afterdischarge was defined as the afterdischarge threshold; this threshold was used for all drug studies. Electric stimulation was delivered by a stimulator (model S88, Grass Instruments) equipped with two constant-current units (Photoelectric Stimulus Isolation Unit, constant current output, Grass Instruments). An oscilloscope (502A, Tektronix, Inc., Portland, Ore.) was used to monitor the amplitude of the electrical stimulation. Twenty-four hours after afterdischarge threshold determination, the animals were returned to the recording chamber. Three to 5 minutes of baseline electroencephalogram was first recorded, with recording continuing throughout the session. Animals were then injected intra peritoneally with either 270 or 360 mg/kg magnesium sulfate or a volume of saline solution equal to that used for the 360 mg/kg dose on one of three test days. The magnesium sulfate doses were chosen on the basis of an earlier study that reported the anticonvulsant effects of magnesium sulfate in rats. Ii The order of drug administration was randomized across days. Ten minutes after injection animals were stimulated through the implanted electrodes at the afterdischarge threshold current. Epileptiform afterdischarge duration, frequency and electroencephalographic amplitude before and after the seizure were recorded. Finally the duration of postictal electroencephalographic depression was determined. All drug injections were separated by a minimum of 48 hours. Magnesium sulfate was made up in sterile saline solution at a concentration of 90 mg/ml. Serum magnesium levels were measured immediately before and at timed intervals after intraperitoneal injection of magnesium sulfate in 10 animals. Each rat was surgically anesthetized and implanted with a chronic, indwelling jugular catheter. The catheters were constructed of polyethylene tubing, filled with heparinized saline solution to reduce obstruction of the cannula, and sealed with a metal pin. Catheters were flushed twice daily with heparinized saline solution, and blood samples were taken 3 days after surgery. In brief, rats were i~ected intraperitoneally with either 180, 270, or 360 mg/kg of magnesium sulfate. At 0, 5, 15, 30, or 60 minutes after injection the pin was removed from the catheter and an 18-gauge needle and syringe inserted into the end of the catheter. Then 0.5 ml samples of blood were collected in 1 ml Microtainer vials (Becton-Dickinson, Rutherford, N.J.). Magnesium levels in 0.7 /LI serum samples were analyzed colorimetrically (Eastman Kodak Co., Rochester, N.Y.).
Volume 166 Number 4
Experiment 2. Direct intracranial injections of magnesium sulfate were made into the dorsal hippocampus through the chronically implanted chemitrodes to determine the effects of magnesium sulfate on seizure activity directly at the seizure focus. In brief, the afterdischarge threshold was determined for these animals as described for experiment I. Again, an electroencephalographic recording was taken continuously throughout this experiment. Animals were then injected with I fl.l of saline solution or I fl.l of saline solution containing an empirically chosen dose of 50 fl.g of magnesium sulfate. 15 Five minutes after chemitrode injection animals were stimulated at the afterdischarge threshold, and the afterdischarge duration and spike frequency were determined. Chemitrode injections were made with a 10 fl.l microsyringe (Hamilton Co., Reno, Nev.) attached to a 30-gauge i~ection needle cut to fit the implanted chemitrode cannula. Injections were made over a 60-second period; the injection needle was left in place an additional 60 seconds to ensure drug delivery to the hippocampus as previously described. IS If drug injection blocked the afterdischarge, current levels were raised in steps of 10 fl.A until an afterdischarge was elicited. Experiment 3. Ten rats (200 to 260 gm) were decapitated and their brains quickly removed. The brain was placed on its dorsal surface; the hippocampus was rapidly dissected on ice at 4° C as previously described. '6 The hippocampus was placed onto a tissue slicer (Stoelting Co., Wood Dale, Ill.), and 400 fl.m slices were cut at a 70 degree angle perpendicular to its septotemporal axis. Slices were immediately placed into cold, oxygenated (95% O 2 /5% CO 2 ) artificial cerebrospinal fluid consisting of 2.4 mmollL calcium chloride, 1.3 mmollL magnesium sulfate, 5 mmollL potassium chloride, 120 mmollL sodium chloride, 1.24 mmollL sodium biphosphate, 26 mmollL sodium bicarbonate, and 10 mmollL dextrose. The magnesium-free medium was prepared by omitting magnesium sulfate. All slices were transferred to artificial cerebrospinal fluid within 7 minutes from the time of death to ensure viability of the tissue. Slices were gently transferred to the recording chamber. The recording chamber was constructed of transparent Plexiglas acrylic plastic. The atmosphere within the chamber was humidified, oxygenated, and heated to 34° C. A brain slice well on which nylon netting was stretched was placed into the chamber and contained 2 ml of artificial cerebrospinal fluid. Hippocampal slices were placed on the netting. Under these condition stable electrophysiologic activity can be maintained in brain slices for 8 to 12 hours. Slices were allowed to equilibrate to the chamber for I hour before electrophysiologic recordings were begun. Extracellular recordings were made with glass micropipettes filled with 2 moll L sodium chloride and with
Anticonvulsant effects of magnesium sulfate
1129
impedances between 1 to 4 megohms. The tip of the recording electrode was placed into the pyramidal c;ell layer of the CA3 subregion of the hippocampus at a depth of approximately 70 to 100 fl.m with a hydraulic micropositioner (model 650, David Kopf Instruments, Wood Dale, Ill.). A concentric bipolar, tungsten microstimulating electrode was used to antidromically stimulate the Schaffer collateral axons in the stratum radiatum of CA3. Extracellular field potentials recorded from pyramidal cells in CA3 were amplified lOOO-fold (model DAM-50, World Precision Instruments, New Haven, Conn.), visualized on a digital storage oscilloscope (model 4094C, Nicolet, Madison, Wisc.) and stored on disk for later analysis. Electric stimulation (single O.l-millisecond monophasic squarewave pulses,S to 30 V) was delivered by a stimulator (model S-48, Grass Intruments) and PSIU-6 stimulus isolation unit. Stimulation of the Schaffer collateral fibers was delivered every 60 seconds, and the intensity was adjusted until a stable extracellular field potential was obtained. After 15 to 20 minutes of superfusion with normal artificial cerebrospinal fluid (i.e., magnesium-containing), magnesium-free artificial cerebrospinal fluid was introduced. Evoked field potentials continued to be recorded every 60 seconds. After 20 to 25 minutes in a magnesium-free artificial cerebrospinal fluid magnesium was added back to the superfusion solution at a concentration of 90 and then 180 fl.moll L. The occurrence of epileptiform bursting under magnesium-free conditions was recorded, as were the effects of the readdition of magnesium sulfate. Histology. At the end of experiments I and 2 animals were injected with an overdose of sodium pentobarbital (120 mg/kg intra peritoneally) and perfused transcardially with 0.9% saline solution followed by 10% formalin. The brains were then frozen-sectioned at 20 fl.m, mounted on microscope slides, and stained with cresyl violet. The precise locations of the implanted electrodes and chemitrodes were then verified microscopically under 40 x magnification. Data analysis. Data were analyzed by a one-way analysis of variance. Post hoc testing was performed when appropriate with the Newman-Keuls test. Values are expressed as mean ± SEM. The significance level was set at p < 0.05.
Results Experiment 1. Intraperitoneal injections of magnesium sulfate produced moderate levels of ataxia, muscular weakness, and lethargy at 270 mg/kg, with severe effects seen at the 360 mg/kg dose. The onset of muscular weakness was within 2 minutes after injection, and animals typically showed complete recovery within 40 to 60 minutes. The effects of intraperitoneal injections of magne-
1130 Cotton, Janusz, and Berman
April 1992 Am J Obslet Gynecol
14
~
-
Ol
12
E
.!!1 Q) > Q)
10
E
8
'in Q) c
6
....J
:::J
Ol
«l
~
E
:::J
4
~
Q)
CJ)
2
.•.
--{}--
-._ _.
180 270 360
O~----~-----'------~-----r----~----~ o 20 40 60
Time After Injection (min) Fig. 1. Serum magnesium levels over a period of I hour after intraperitoneal injection of 180, 270, or 360 mg / kg of magnesium sulfate.
Table I. Summary of effects of intraperitoneal injections of magnesium sulfate on hippocampal electroencephalographic seizure threshold, seizure duration, and hippocampal electroencephalographic amplitude Group
Saline solution (n = 11 ; mean ± SEM) Magnesium sulfate 270 mg/kg (II = II ; mean ± SEM) Magnesium sulfate 360 mg / kg (II = II ; mean ± SEM)
*P
Key words: Magnesium sulfate, hippocampus, seizures, eclampsia
Magnesium sulfate has been used for >60 years to prevent and treat eclamptic seizures. I -3 In spite of magnesium sulfate's empiric success in obstetrics, the literature of other medical disciplines has decried its use on the grounds that it does not have any central inhibitory action and merely masks seizures via peripheral neuromuscular blockade!-7 However, on the basis of the well-documented clinical response of eclamptic seizures to magnesium sulfate in the absence of any neuromuscular blockade, Pritchard" has proposed that magnesium sulfate does have central effects. Experimental work in dogs, cats, and subhuman primates also provides support for a central mechanism of action. Borges and Gucer" showed that intravenous magnesium sulfate infusion suppressed neuronal burst firing and decreased interictal spike frequency in a penicillin-induced epilepsy model. Suppression occurred at serum magnesium levels well below those required for complete neuromuscular blockade; Borges and Gucer concluded that magnesium sulfate does have a central effect. In a similar model of penicillin-induced epileptic foci in cats Koontz and Reid!O were unable to duplicate From the Department of Obstetrics and Gynecology, Wayne State University-Hutzel Hospital. Presented by invitation at the Tenth Annual Meeting of the Gynecological and Obstetrical Society, Carlsbad, California, September 57, 1991. Reprint requests: David B. Cotton, MD, Hutzel Hospital-Wayne State University, Department of Obstetrics and Gynecology, 4707 St. Antoine, Detroit, MI 48201. 616135626
Borges and Gucer's9 findings. In Koontz and Reid's model a saline-infused group had the same rapid decrement in seizure activity as the group receiving magnesium sulfate. Koontz and Reid II later performed an experiment with the same seizure model to determine the effect of pretreatment with magnesium sulfate; again, they could not document any anticonvulsant effect of magnesium sulfate. They concluded that parenteral magnesium sulfate had no demonstrable anticonvulsant effect in the penicillin-induced seizure focus model in cats. In another experiment, Hilmy and Somjen l2 demonstrated a fourfold to sevenfold greater deposition of magnesium in peripheral tissues as opposed to brain tissues after intravenous magnesium administration. These investigators concluded that magnesium acts primarily through neuromuscular blockade and that a barrier mechanism protects the central nervous system from high levels of magnesium. Others have suggested that in eclampsia any apparent central effects of magnesium sulfate must be the result of disruption of the blood-brain barrier and leakage of magnesium into the brain. 4 • !O Donaldson 4 expanded on these observations and concluded that magnesium sulfate does not treat eclamptic convulsions and adds substantial risks to mother and baby. Because of these disparate conclusions regarding magnesium sulfate's mechanism of action, we examined in rats the effects of magnesium sulfate on seizures initiated in the hippocampus, a brain region with a low seizure threshold and a high density of N-methyl-o1127
1128 Cotton, Janusz, and Berman
aspartate receptors. A study of hippocampal seizures should provide a sensitive index of the possible central nervous system effects of magnesium sulfate. The high density of N-methyl-n-aspartate receptors is also of interest because these receptors have been linked to various models of epilepsy and are blocked by magnesium." Material and methods Surgical preparation. Male Long-Evans rats (Harlan Sprague-Dawley Inc., Indianapolis) were individually housed in stainless steel cages in an environmentally controlled vivarium under 12-hour light-dark cycles. Animals had free access to food throughout the experiments. Twelve animals (275 to 300 gm) were surgically anesthetized (60 mg/kg sodium pentobarbital intraperitoneally) and stereotaxically implanted with chronically indwelling bipolar electrodes (MS-303/2, Plastics One Inc., Roanoke, Va.) aimed for the dorsal hippocampus. The stereotaxic coordinates were 4.0 mm posterior bregma, 2.0 mm lateral from midline, and 4.0 mm ventral to the superior surface of the level skull. The electrode assembly was chronically attached to the skull with three stainless steel jeweler's screws and acrylic cement. Animals were allowed 1 week of postsurgical recovery before further treatments were begun. An additional group of 5 male Long-Evans rats (275 to 300 gm) were surgically implanted with chemitrodes into the dorsal hippocampus as described above. The chemitrode, which permitted electric stimulation and intracranial drug injection at the same brain site, consisted of a bipolar electrode (MS303/3, Plastics One Inc.) cemented to a 23-gauge stainless steel cannula. A 30-gauge obturator plugged the cannula until the time of injection. The ends of the electrode extended 0.5 mm past the end of the cannula. The coordinates for implantation were the same as those described above. A I-week postsurgical recovery period was again used. A final group of 10 male Long-Evans rats (200 to 260 gm) were used for the in vitro hippocampal brain slice experiments as described below. Experimental design and protocol Experiment 1. In this experiment the effects of intraperitoneal injections of magnesium sulfate at 270 and 360 mg/kg on hippocampal epileptiform afterdischarges were determined. Rats implanted with bipolar electrodes in the dorsal hippocampus were placed into a Plexiglas acrylic plastic recording ch:lmber and leads from a polygraph (model 7, Grass Instruments Co., Quincy, Mass.) were attached to the implanted electrodes. Baseline electroencephalographic activity was recorded from the hippocampus via the implanted electrodes. One-second trains (100 Hz, I-millisecond pulse duration, biphasic, square waves) of low-intensity electric stimulation were delivered
April 1992 Am J Obstet Gynecol
through the implanted electrodes beginning at a current intensity of 10 /LA. Stimulation intensity was increased in steps of 10 /LA until a brief, 5- lO-second epileptiform afterdischarge was recorded from the hippocampus immediately after stimulation. The level of stimulation that evoked an afterdischarge was defined as the afterdischarge threshold; this threshold was used for all drug studies. Electric stimulation was delivered by a stimulator (model S88, Grass Instruments) equipped with two constant-current units (Photoelectric Stimulus Isolation Unit, constant current output, Grass Instruments). An oscilloscope (502A, Tektronix, Inc., Portland, Ore.) was used to monitor the amplitude of the electrical stimulation. Twenty-four hours after afterdischarge threshold determination, the animals were returned to the recording chamber. Three to 5 minutes of baseline electroencephalogram was first recorded, with recording continuing throughout the session. Animals were then injected intra peritoneally with either 270 or 360 mg/kg magnesium sulfate or a volume of saline solution equal to that used for the 360 mg/kg dose on one of three test days. The magnesium sulfate doses were chosen on the basis of an earlier study that reported the anticonvulsant effects of magnesium sulfate in rats. Ii The order of drug administration was randomized across days. Ten minutes after injection animals were stimulated through the implanted electrodes at the afterdischarge threshold current. Epileptiform afterdischarge duration, frequency and electroencephalographic amplitude before and after the seizure were recorded. Finally the duration of postictal electroencephalographic depression was determined. All drug injections were separated by a minimum of 48 hours. Magnesium sulfate was made up in sterile saline solution at a concentration of 90 mg/ml. Serum magnesium levels were measured immediately before and at timed intervals after intraperitoneal injection of magnesium sulfate in 10 animals. Each rat was surgically anesthetized and implanted with a chronic, indwelling jugular catheter. The catheters were constructed of polyethylene tubing, filled with heparinized saline solution to reduce obstruction of the cannula, and sealed with a metal pin. Catheters were flushed twice daily with heparinized saline solution, and blood samples were taken 3 days after surgery. In brief, rats were i~ected intraperitoneally with either 180, 270, or 360 mg/kg of magnesium sulfate. At 0, 5, 15, 30, or 60 minutes after injection the pin was removed from the catheter and an 18-gauge needle and syringe inserted into the end of the catheter. Then 0.5 ml samples of blood were collected in 1 ml Microtainer vials (Becton-Dickinson, Rutherford, N.J.). Magnesium levels in 0.7 /LI serum samples were analyzed colorimetrically (Eastman Kodak Co., Rochester, N.Y.).
Volume 166 Number 4
Experiment 2. Direct intracranial injections of magnesium sulfate were made into the dorsal hippocampus through the chronically implanted chemitrodes to determine the effects of magnesium sulfate on seizure activity directly at the seizure focus. In brief, the afterdischarge threshold was determined for these animals as described for experiment I. Again, an electroencephalographic recording was taken continuously throughout this experiment. Animals were then injected with I fl.l of saline solution or I fl.l of saline solution containing an empirically chosen dose of 50 fl.g of magnesium sulfate. 15 Five minutes after chemitrode injection animals were stimulated at the afterdischarge threshold, and the afterdischarge duration and spike frequency were determined. Chemitrode injections were made with a 10 fl.l microsyringe (Hamilton Co., Reno, Nev.) attached to a 30-gauge i~ection needle cut to fit the implanted chemitrode cannula. Injections were made over a 60-second period; the injection needle was left in place an additional 60 seconds to ensure drug delivery to the hippocampus as previously described. IS If drug injection blocked the afterdischarge, current levels were raised in steps of 10 fl.A until an afterdischarge was elicited. Experiment 3. Ten rats (200 to 260 gm) were decapitated and their brains quickly removed. The brain was placed on its dorsal surface; the hippocampus was rapidly dissected on ice at 4° C as previously described. '6 The hippocampus was placed onto a tissue slicer (Stoelting Co., Wood Dale, Ill.), and 400 fl.m slices were cut at a 70 degree angle perpendicular to its septotemporal axis. Slices were immediately placed into cold, oxygenated (95% O 2 /5% CO 2 ) artificial cerebrospinal fluid consisting of 2.4 mmollL calcium chloride, 1.3 mmollL magnesium sulfate, 5 mmollL potassium chloride, 120 mmollL sodium chloride, 1.24 mmollL sodium biphosphate, 26 mmollL sodium bicarbonate, and 10 mmollL dextrose. The magnesium-free medium was prepared by omitting magnesium sulfate. All slices were transferred to artificial cerebrospinal fluid within 7 minutes from the time of death to ensure viability of the tissue. Slices were gently transferred to the recording chamber. The recording chamber was constructed of transparent Plexiglas acrylic plastic. The atmosphere within the chamber was humidified, oxygenated, and heated to 34° C. A brain slice well on which nylon netting was stretched was placed into the chamber and contained 2 ml of artificial cerebrospinal fluid. Hippocampal slices were placed on the netting. Under these condition stable electrophysiologic activity can be maintained in brain slices for 8 to 12 hours. Slices were allowed to equilibrate to the chamber for I hour before electrophysiologic recordings were begun. Extracellular recordings were made with glass micropipettes filled with 2 moll L sodium chloride and with
Anticonvulsant effects of magnesium sulfate
1129
impedances between 1 to 4 megohms. The tip of the recording electrode was placed into the pyramidal c;ell layer of the CA3 subregion of the hippocampus at a depth of approximately 70 to 100 fl.m with a hydraulic micropositioner (model 650, David Kopf Instruments, Wood Dale, Ill.). A concentric bipolar, tungsten microstimulating electrode was used to antidromically stimulate the Schaffer collateral axons in the stratum radiatum of CA3. Extracellular field potentials recorded from pyramidal cells in CA3 were amplified lOOO-fold (model DAM-50, World Precision Instruments, New Haven, Conn.), visualized on a digital storage oscilloscope (model 4094C, Nicolet, Madison, Wisc.) and stored on disk for later analysis. Electric stimulation (single O.l-millisecond monophasic squarewave pulses,S to 30 V) was delivered by a stimulator (model S-48, Grass Intruments) and PSIU-6 stimulus isolation unit. Stimulation of the Schaffer collateral fibers was delivered every 60 seconds, and the intensity was adjusted until a stable extracellular field potential was obtained. After 15 to 20 minutes of superfusion with normal artificial cerebrospinal fluid (i.e., magnesium-containing), magnesium-free artificial cerebrospinal fluid was introduced. Evoked field potentials continued to be recorded every 60 seconds. After 20 to 25 minutes in a magnesium-free artificial cerebrospinal fluid magnesium was added back to the superfusion solution at a concentration of 90 and then 180 fl.moll L. The occurrence of epileptiform bursting under magnesium-free conditions was recorded, as were the effects of the readdition of magnesium sulfate. Histology. At the end of experiments I and 2 animals were injected with an overdose of sodium pentobarbital (120 mg/kg intra peritoneally) and perfused transcardially with 0.9% saline solution followed by 10% formalin. The brains were then frozen-sectioned at 20 fl.m, mounted on microscope slides, and stained with cresyl violet. The precise locations of the implanted electrodes and chemitrodes were then verified microscopically under 40 x magnification. Data analysis. Data were analyzed by a one-way analysis of variance. Post hoc testing was performed when appropriate with the Newman-Keuls test. Values are expressed as mean ± SEM. The significance level was set at p < 0.05.
Results Experiment 1. Intraperitoneal injections of magnesium sulfate produced moderate levels of ataxia, muscular weakness, and lethargy at 270 mg/kg, with severe effects seen at the 360 mg/kg dose. The onset of muscular weakness was within 2 minutes after injection, and animals typically showed complete recovery within 40 to 60 minutes. The effects of intraperitoneal injections of magne-
1130 Cotton, Janusz, and Berman
April 1992 Am J Obslet Gynecol
14
~
-
Ol
12
E
.!!1 Q) > Q)
10
E
8
'in Q) c
6
....J
:::J
Ol
«l
~
E
:::J
4
~
Q)
CJ)
2
.•.
--{}--
-._ _.
180 270 360
O~----~-----'------~-----r----~----~ o 20 40 60
Time After Injection (min) Fig. 1. Serum magnesium levels over a period of I hour after intraperitoneal injection of 180, 270, or 360 mg / kg of magnesium sulfate.
Table I. Summary of effects of intraperitoneal injections of magnesium sulfate on hippocampal electroencephalographic seizure threshold, seizure duration, and hippocampal electroencephalographic amplitude Group
Saline solution (n = 11 ; mean ± SEM) Magnesium sulfate 270 mg/kg (II = II ; mean ± SEM) Magnesium sulfate 360 mg / kg (II = II ; mean ± SEM)
*P
