Effects of Reserpine, Propranolol, and Aminophylline on Seizure Activity and CNS Cyclic Nucleotides Robert A. Gross, AB, and James A. Ferrendelli, MD
Convulsant doses of pentylenetetrazol ( 100 mg/kg) increase levels of both cyclic adenosine monophosphate (AMP) and cyclic guanosine monophosphate (GMP) in mouse cerebral cortex and hippocampus. In animals pretreated with reserpine, propranolol, or aminophylline, pentylenetetrazol seizures were more severe, cyclic AMP elevations were attenuated or blocked, and cyclic GMP increases were unaffected or augmented. These data indicate that norepinephrine, adenosine, and perhaps other biogenic amines have a regulatory effect on cyclic AMP, but not cyclic GMP, levels in epileptic brain. An increased level of cyclic AMP in brain tissue may have an antiepileptic effect leading to seizure attenuation or termination. By contrast, elevated levels of cyclic GMP may have an epileptogenic effect in initiating or maintaining seizure activity. G r o s s R A , Ferrendelli JA: Effects of reserpine, propranolol, and aminophylline on seizure activity and CNS cyclic nucleotides. Ann Neurol 6:296-301, 1979
Seizure discharges, either spontaneous or produced experimentally with electrical stimulation o r convulsant drugs, represent a state of excessive cellular depolarization with marked alterations of metabolism and function in brain. Recent studies have demonstrated that the content of both cyclic adenosine monophosphate (AMP) and cyclic guanosine monophosphate (GMP) is increased in the central nervous system in vivo by experimentally produced seizures [ l , 8, 9, 301. The relationships between seizure discharges and cyclic nucleotides in mammalian brain are still poorly understood, however. T h e findings that derivatives of cyclic G M P produce epileptiform discharges in explants of brain tissue [ 111 and that cyclic AMP has an inhibitory effect on CNS neurons 13, 241 suggest that both of these cyclic nucleotides may have important roles in the parhophysiology of epilepsy. A study was undertaken to better explain the mechanisms responsible for regulation of cyclic nucleotide levels i n epileptic brain and to determine whether alterations of CNS cyclic nucleotide levels modify seizure activity. We investigated the effects of reserpine, propranolol, aminophylline, and other drugs on pentylenetetrazol-induced seizure behavior and on changes in cyclic A M P and cyclic G M P levels in the hippocampus and cerebral cortex of mice. T h e data presented here show that brain cyclic AMP and
Materials and Methods Adult female Swiss Webster mice, given food and warer ad libitum, were used in all experiments. Animals were treated with 100 mg of pentylenetetrazol (Sigma Chemical Company; o r Metrazol, Knoll Pharmaceutical Company) per kilogram of body weight alone o r after prerreatment with aminophylline (Sigma), propranolol (lndrral, Ayerst Laboratories), reserpine (Serpasil, Ciba Pharinaceutical Company), WB-4101 (W. B. Pharmaceuticals), fluphenazine (Prolixin, E. R. Squibb & Sons), o r merhysergide (Sansert, Sandoz Pharmaceuticals). Drug closes and treatment schcdules are described in the Resulrs section. All drugs were dissolved or diluted in water and injected intraperironeally in a volume of 0.01 ml per gram of body weight. Control animals receiving no drug treatment were injected with an equivalent volume of water o r w r e given a sham injection (insertion of the needle only). Following all treatments, behavior of the animals was carefully observed and recorded. At designated times, animals were killed by rapid immersion in liquid nitrogen and srored at - 7 6 T . Faster fixation methods (e.g., microwave irradiation) were nor used, as they precluded the observation of behavior which was considered t o be of importance in this study. Dissection of the frozen mice was carried out at - 20°C. Two cores of tissue weighing 3 to 4 mg each were removed from each brain region studied; the samples were then ex-
From the Departmenr o f Pharmacology and the Department of Neurology anii Neurological Surgery (Neurology), Washington University School o f Medicine, St. Louis, MO.
Address reprint requests to D r Ferrendelli, Department of Neurology, Washington University School of Medicine, 660 S Euclid Ave, St. Louis, MO 631 10.
cyclic G M P levels are differentially regulated during seizures and suggest that the tissue concentrations of the nucleotides may influence seizure activity.
Accepted for publication Apr 3, 1979
296 0364-5 134i~)/llO02')~-06$01.25 @ 1979 by Robert A. Gross
tracted with acid methanol and perchloric acid (PCA). First, the samples were homogenized at -2O"C, with groundglass pestles, in approximately 2 to 3 volumes of 0.1 N hydrochloric acid in 100% methanol. The samples were further homogenized at 0°C after addition of 9 to 11 volumes of 0.5 N PCA/1 mM EDTA and centrifuged at 14,500 g for 20 minutes at 4°C. An aliquot of clear supernatant fluid was then removed; perchlorate was precipitated and the solution neutralized by addition of 2.5 M potassium bicarbonate. The samples were then centrifuged at 4°C for 15 minutes at 14,500 g to remove the potassium perchlorate. To a 50 p l aliquot of neutralized extract, 1 pl of triethylamine (Eastman Kodak Company) and 0.5 pl of acetic anhydride (Fisher Scientific Company) were added to acetylate the cyclic nucleotides [13]. The samples were then dried under vacuum and the residue dissolved in 50 mM sodium acetate buffer, pH 6, and assayed for cyclic nucleotides by radioimmunoassay [27]. Perchloric acid precipitates were solubilized in 1 N sodium hydroxide and assayed for protein by the method of Lowry et a1 [19]. Statistical analyses of data were computed using Student's two-tailed t test.
Results As previously reported [7, 81, pentylenetetrazol, 100 mg/kg, caused highly reproducible and characteristic seizure behavior in mice, consisting of a generalized myoclonic twitch about 1 minute after intraperitoneal injection, immediately followed by a clonic seizure. Five to 6 minutes later, and following several additional myoclonic twitches, clonic seizures, or both, almost all animals had a clonic-tonic convulsion and died (Fig 1A). Cyclic AMP levels were elevated 2- to 4-fold in the cerebral cortex and hippocampus of animals frozen during the first clonic seizure, and they increased by additional small amounts with continued seizure activity (Fig 1B). Cyclic GMP levels increased slightly, 1%- to 2-fold or less, during the first clonic seizure; but with continued seizure activity they rose further, to concentrations 2 to 5 times the level in untreated animals. Pretreatment of animals with reserpine (0.1 to 10 mg/kg) 3 hours prior to pentylenetetrazol treatment produced sedation, ptosis, and diarrhea at doses of 1 mg/kg or higher. Basal levels of cyclic AMP in the hippocampus were slightly elevated by reserpine, 0.3 mg/kg (Fig 1B). Higher doses caused a progressive decrease in cyclic AMP levels in both hippocampus and cerebral cortex, and 10 mg/kg reduced its levels by 25 to 50%. In reserpine-pretreated animals, pentylenetetrazol-induced tonic seizures occurred more rapidly. This effect was dose dependent, and in animals pretreated with 10 mg/kg of reserpine, tonic seizures began almost immediately after the first clonic seizure (Fig 1A). Except for the more rapid onset of myoclonic twitching in animals treated with 10 mg/kg of reserpine, this drug had no effect on the time of onset or character of the first myoclonic
FIRST
MYOCLONC
FIRST CLONIC
FIRST TONIC
S E / Z /
,
I
0
1
2
I
I
,
I
I
3
4
5
6
7
,
MINUTES
A
CYCLIC AMP (pmoleshg PROTI 20
CONTROL
CONTROL
RESERPINE ( r n g / k g )
B F i g I. (A) The effect of reserpine on pentylenetetrazol-induced seizures. Mice iuere treated with the indicated doses of reserpine (mglkd 3 hours prior t o injection of I00 mglkg of pentylenetetrazol (Metrazol), and the times of appearance of three types of readily identifiable seizures were recorded. Control animals received no reserpine. Closed symbols indicate values that were significantly different ( p < 0.05)from controls. Each point represents the mean of 5 to 18 animals. (B) Cyclic AMP levels in the hippocampus and cerebral cortex of mice treated with reserpine only (controls)or treated with both reserpine and pentylenetetrazol (Metrazol). Reserpine was injected, at the indicated doses, 3 hours prior t o injection of 100 mglkg of pentylenetetrazol. Pentylenetetrazol-treatedanimals were frozen in liquid nitrogen at the end of the first clonic seizure; control animals were frozen 3 hours after reserpine injection. Symbols and vertical bars are the means ? SEM, respectively, of 5 t o 8 animals. Asterisks indicate values significantly different (p < 0.03)from reserpine control and pentylenetetrazol-treated animals receiving no reserpine.
twitch or clonic seizure. Reserpine blocked elevations of cyclic AMP levels in both cerebral cortex and hippocampus of pentylenetetrazol-treated animals frozen during the first clonic seizure. This effect was also dose dependent, and 10 mg/kg completely blocked seizure-induced elevations of cyclic AMP in one experiment and reduced the elevation 90% in another. We next tested the possibility that reserpine would inhibit pentylenetetrazol-induced elevations of cyclic AMP levels by depleting biogenic amine neurotransmitters in brain, thereby preventing their actions at synaptic receptor sites. Accordingly, we examined
Gross and Ferrendelli: Seizures and Cyclic Nucleotides
297
-
8
rn
[ HIPPOCAMPUS
CEREBRALCORTEX]
t
40-
p*.
-aE, 20-
' I
3
5
7 1 MINUTES
3
I
1
5
7
F i g 2. The rff;.(-tof propratiolol and anritiophylline o n pentjlr.nrtrtruzof-itidul.cdsrizure behazior. Each drug was in.jel-trd(30 niirrutrs prior to injection 14100 niglkg ofprntjlrwrtrtrazol. Thr nunibim rrfrr t o the d m (niglkd of drug given. Closed symbols indicate a signdficant difference (p < 0.05)from thr srizurr tittic imrsr of aninials reming o n l y pentylenetetrazol (control).Each point is the mean of 5 to 18 arrinrals.
0%'0.31
Vol 6
No 4
October 1979
CONTROL
'fi*;i;b' -%+
8 3'0' $03 PROPRANOLOL ( m g / k g )
- 5 0 HIPPOCAMPUS
CEREBRAL CORTE)
-
a
a 40-
\
ar Ln
0 30-
5 2
1
20-
a V
;10-
' 'k'lb
298 Annals of Neurology
T
A
0
the effects of several neurotransmitter receptor blockers. The following agents were injected 30 minutes prior to pentylenetetrazol treatment: 50 mglkg of methysergide, a serotonin receptor antagonist; 3 pglkg of WB-4101, a potent centrally acting aadrenergic receptor blocker; 20 mglkg of propranolol, a centrally acting p-adrenergic receptor antagonist; 1 mglkg of fluphenazine, a dopamine receptor blocker; o r 50 mglkg of aminophylline, an adenosine receptor antagonist. Only propranolol and arninophylline reduced pentylenetetrazol-induced elevations of cyclic AMP in hippocampus; therefore, the two drugs were studied further. Propranolol pretreatment (0.3 to 10 mglkg) had no observable effect on behavior, nor did it influence the onset of the first myoclonic twitch or first clonic seizure. However, propranolol-pretreated animals had more frequent clonic seizures and the drug appeared t o shorten the time to the tonic seizure, although this was not a clear dose-dependent effect (Fig 2). In addition, at higher doses (>3 mglkg), many propranolol-treated animals survived the initial tonic seizure but subsequently died after several additional clonic o r tonic seizures. Propranolol reduced both basal levels and pentylenetetrazol-induced accumulation of cyclic AMP in cerebral cortex and hippocampus in a dose-related manner (Fig 3A). In the illustrated experiment, the drug was less effective in cerebral cortex, but other experiments showed an equivalent effect on both regions of brain. In either cerebral region, propranolol reduced seizureinduced elevations of cyclic AMP by only 50%, and this appeared to be a maximal effect since 20 to 30
""io
METRAZOL
CONTROL
'2'5' '
Convulsant doses of pentylenetetrazol ( 100 mg/kg) increase levels of both cyclic adenosine monophosphate (AMP) and cyclic guanosine monophosphate (GMP) in mouse cerebral cortex and hippocampus. In animals pretreated with reserpine, propranolol, or aminophylline, pentylenetetrazol seizures were more severe, cyclic AMP elevations were attenuated or blocked, and cyclic GMP increases were unaffected or augmented. These data indicate that norepinephrine, adenosine, and perhaps other biogenic amines have a regulatory effect on cyclic AMP, but not cyclic GMP, levels in epileptic brain. An increased level of cyclic AMP in brain tissue may have an antiepileptic effect leading to seizure attenuation or termination. By contrast, elevated levels of cyclic GMP may have an epileptogenic effect in initiating or maintaining seizure activity. G r o s s R A , Ferrendelli JA: Effects of reserpine, propranolol, and aminophylline on seizure activity and CNS cyclic nucleotides. Ann Neurol 6:296-301, 1979
Seizure discharges, either spontaneous or produced experimentally with electrical stimulation o r convulsant drugs, represent a state of excessive cellular depolarization with marked alterations of metabolism and function in brain. Recent studies have demonstrated that the content of both cyclic adenosine monophosphate (AMP) and cyclic guanosine monophosphate (GMP) is increased in the central nervous system in vivo by experimentally produced seizures [ l , 8, 9, 301. The relationships between seizure discharges and cyclic nucleotides in mammalian brain are still poorly understood, however. T h e findings that derivatives of cyclic G M P produce epileptiform discharges in explants of brain tissue [ 111 and that cyclic AMP has an inhibitory effect on CNS neurons 13, 241 suggest that both of these cyclic nucleotides may have important roles in the parhophysiology of epilepsy. A study was undertaken to better explain the mechanisms responsible for regulation of cyclic nucleotide levels i n epileptic brain and to determine whether alterations of CNS cyclic nucleotide levels modify seizure activity. We investigated the effects of reserpine, propranolol, aminophylline, and other drugs on pentylenetetrazol-induced seizure behavior and on changes in cyclic A M P and cyclic G M P levels in the hippocampus and cerebral cortex of mice. T h e data presented here show that brain cyclic AMP and
Materials and Methods Adult female Swiss Webster mice, given food and warer ad libitum, were used in all experiments. Animals were treated with 100 mg of pentylenetetrazol (Sigma Chemical Company; o r Metrazol, Knoll Pharmaceutical Company) per kilogram of body weight alone o r after prerreatment with aminophylline (Sigma), propranolol (lndrral, Ayerst Laboratories), reserpine (Serpasil, Ciba Pharinaceutical Company), WB-4101 (W. B. Pharmaceuticals), fluphenazine (Prolixin, E. R. Squibb & Sons), o r merhysergide (Sansert, Sandoz Pharmaceuticals). Drug closes and treatment schcdules are described in the Resulrs section. All drugs were dissolved or diluted in water and injected intraperironeally in a volume of 0.01 ml per gram of body weight. Control animals receiving no drug treatment were injected with an equivalent volume of water o r w r e given a sham injection (insertion of the needle only). Following all treatments, behavior of the animals was carefully observed and recorded. At designated times, animals were killed by rapid immersion in liquid nitrogen and srored at - 7 6 T . Faster fixation methods (e.g., microwave irradiation) were nor used, as they precluded the observation of behavior which was considered t o be of importance in this study. Dissection of the frozen mice was carried out at - 20°C. Two cores of tissue weighing 3 to 4 mg each were removed from each brain region studied; the samples were then ex-
From the Departmenr o f Pharmacology and the Department of Neurology anii Neurological Surgery (Neurology), Washington University School o f Medicine, St. Louis, MO.
Address reprint requests to D r Ferrendelli, Department of Neurology, Washington University School of Medicine, 660 S Euclid Ave, St. Louis, MO 631 10.
cyclic G M P levels are differentially regulated during seizures and suggest that the tissue concentrations of the nucleotides may influence seizure activity.
Accepted for publication Apr 3, 1979
296 0364-5 134i~)/llO02')~-06$01.25 @ 1979 by Robert A. Gross
tracted with acid methanol and perchloric acid (PCA). First, the samples were homogenized at -2O"C, with groundglass pestles, in approximately 2 to 3 volumes of 0.1 N hydrochloric acid in 100% methanol. The samples were further homogenized at 0°C after addition of 9 to 11 volumes of 0.5 N PCA/1 mM EDTA and centrifuged at 14,500 g for 20 minutes at 4°C. An aliquot of clear supernatant fluid was then removed; perchlorate was precipitated and the solution neutralized by addition of 2.5 M potassium bicarbonate. The samples were then centrifuged at 4°C for 15 minutes at 14,500 g to remove the potassium perchlorate. To a 50 p l aliquot of neutralized extract, 1 pl of triethylamine (Eastman Kodak Company) and 0.5 pl of acetic anhydride (Fisher Scientific Company) were added to acetylate the cyclic nucleotides [13]. The samples were then dried under vacuum and the residue dissolved in 50 mM sodium acetate buffer, pH 6, and assayed for cyclic nucleotides by radioimmunoassay [27]. Perchloric acid precipitates were solubilized in 1 N sodium hydroxide and assayed for protein by the method of Lowry et a1 [19]. Statistical analyses of data were computed using Student's two-tailed t test.
Results As previously reported [7, 81, pentylenetetrazol, 100 mg/kg, caused highly reproducible and characteristic seizure behavior in mice, consisting of a generalized myoclonic twitch about 1 minute after intraperitoneal injection, immediately followed by a clonic seizure. Five to 6 minutes later, and following several additional myoclonic twitches, clonic seizures, or both, almost all animals had a clonic-tonic convulsion and died (Fig 1A). Cyclic AMP levels were elevated 2- to 4-fold in the cerebral cortex and hippocampus of animals frozen during the first clonic seizure, and they increased by additional small amounts with continued seizure activity (Fig 1B). Cyclic GMP levels increased slightly, 1%- to 2-fold or less, during the first clonic seizure; but with continued seizure activity they rose further, to concentrations 2 to 5 times the level in untreated animals. Pretreatment of animals with reserpine (0.1 to 10 mg/kg) 3 hours prior to pentylenetetrazol treatment produced sedation, ptosis, and diarrhea at doses of 1 mg/kg or higher. Basal levels of cyclic AMP in the hippocampus were slightly elevated by reserpine, 0.3 mg/kg (Fig 1B). Higher doses caused a progressive decrease in cyclic AMP levels in both hippocampus and cerebral cortex, and 10 mg/kg reduced its levels by 25 to 50%. In reserpine-pretreated animals, pentylenetetrazol-induced tonic seizures occurred more rapidly. This effect was dose dependent, and in animals pretreated with 10 mg/kg of reserpine, tonic seizures began almost immediately after the first clonic seizure (Fig 1A). Except for the more rapid onset of myoclonic twitching in animals treated with 10 mg/kg of reserpine, this drug had no effect on the time of onset or character of the first myoclonic
FIRST
MYOCLONC
FIRST CLONIC
FIRST TONIC
S E / Z /
,
I
0
1
2
I
I
,
I
I
3
4
5
6
7
,
MINUTES
A
CYCLIC AMP (pmoleshg PROTI 20
CONTROL
CONTROL
RESERPINE ( r n g / k g )
B F i g I. (A) The effect of reserpine on pentylenetetrazol-induced seizures. Mice iuere treated with the indicated doses of reserpine (mglkd 3 hours prior t o injection of I00 mglkg of pentylenetetrazol (Metrazol), and the times of appearance of three types of readily identifiable seizures were recorded. Control animals received no reserpine. Closed symbols indicate values that were significantly different ( p < 0.05)from controls. Each point represents the mean of 5 to 18 animals. (B) Cyclic AMP levels in the hippocampus and cerebral cortex of mice treated with reserpine only (controls)or treated with both reserpine and pentylenetetrazol (Metrazol). Reserpine was injected, at the indicated doses, 3 hours prior t o injection of 100 mglkg of pentylenetetrazol. Pentylenetetrazol-treatedanimals were frozen in liquid nitrogen at the end of the first clonic seizure; control animals were frozen 3 hours after reserpine injection. Symbols and vertical bars are the means ? SEM, respectively, of 5 t o 8 animals. Asterisks indicate values significantly different (p < 0.03)from reserpine control and pentylenetetrazol-treated animals receiving no reserpine.
twitch or clonic seizure. Reserpine blocked elevations of cyclic AMP levels in both cerebral cortex and hippocampus of pentylenetetrazol-treated animals frozen during the first clonic seizure. This effect was also dose dependent, and 10 mg/kg completely blocked seizure-induced elevations of cyclic AMP in one experiment and reduced the elevation 90% in another. We next tested the possibility that reserpine would inhibit pentylenetetrazol-induced elevations of cyclic AMP levels by depleting biogenic amine neurotransmitters in brain, thereby preventing their actions at synaptic receptor sites. Accordingly, we examined
Gross and Ferrendelli: Seizures and Cyclic Nucleotides
297
-
8
rn
[ HIPPOCAMPUS
CEREBRALCORTEX]
t
40-
p*.
-aE, 20-
' I
3
5
7 1 MINUTES
3
I
1
5
7
F i g 2. The rff;.(-tof propratiolol and anritiophylline o n pentjlr.nrtrtruzof-itidul.cdsrizure behazior. Each drug was in.jel-trd(30 niirrutrs prior to injection 14100 niglkg ofprntjlrwrtrtrazol. Thr nunibim rrfrr t o the d m (niglkd of drug given. Closed symbols indicate a signdficant difference (p < 0.05)from thr srizurr tittic imrsr of aninials reming o n l y pentylenetetrazol (control).Each point is the mean of 5 to 18 arrinrals.
0%'0.31
Vol 6
No 4
October 1979
CONTROL
'fi*;i;b' -%+
8 3'0' $03 PROPRANOLOL ( m g / k g )
- 5 0 HIPPOCAMPUS
CEREBRAL CORTE)
-
a
a 40-
\
ar Ln
0 30-
5 2
1
20-
a V
;10-
' 'k'lb
298 Annals of Neurology
T
A
0
the effects of several neurotransmitter receptor blockers. The following agents were injected 30 minutes prior to pentylenetetrazol treatment: 50 mglkg of methysergide, a serotonin receptor antagonist; 3 pglkg of WB-4101, a potent centrally acting aadrenergic receptor blocker; 20 mglkg of propranolol, a centrally acting p-adrenergic receptor antagonist; 1 mglkg of fluphenazine, a dopamine receptor blocker; o r 50 mglkg of aminophylline, an adenosine receptor antagonist. Only propranolol and arninophylline reduced pentylenetetrazol-induced elevations of cyclic AMP in hippocampus; therefore, the two drugs were studied further. Propranolol pretreatment (0.3 to 10 mglkg) had no observable effect on behavior, nor did it influence the onset of the first myoclonic twitch or first clonic seizure. However, propranolol-pretreated animals had more frequent clonic seizures and the drug appeared t o shorten the time to the tonic seizure, although this was not a clear dose-dependent effect (Fig 2). In addition, at higher doses (>3 mglkg), many propranolol-treated animals survived the initial tonic seizure but subsequently died after several additional clonic o r tonic seizures. Propranolol reduced both basal levels and pentylenetetrazol-induced accumulation of cyclic AMP in cerebral cortex and hippocampus in a dose-related manner (Fig 3A). In the illustrated experiment, the drug was less effective in cerebral cortex, but other experiments showed an equivalent effect on both regions of brain. In either cerebral region, propranolol reduced seizureinduced elevations of cyclic AMP by only 50%, and this appeared to be a maximal effect since 20 to 30
""io
METRAZOL
CONTROL
'2'5' '
