- Email: [email protected]
Inositol triphosphate, cyclic AMP, and cyclic GMP in rat brain regions after lithium and seizures
BIOL PSYCHIATRY 1992;31:505-514
505
Inositol Trisphosphate, Cyclic AMP, and Cyclic GMP in Rat Brain Regions After Lithium and Seizures Richard S. Jope, 'Ling Song, and Krystyna Kolasa
The mechanism of action of lithium, the primary treatment for bipolar affective disorder, is unknown but may involve inhibin'on of second messenger production in the brain. Therefore, the concentrations of three second messengers, inositol 1,4,5 trisphosphate (lns 1,4,5P3), cyclic adenosine monophosphate (AMP), and cyclic guanosine monophosphate (GMP), were measured in rat cerebral cortex and hippocampus after acute or chronic lithium administration, as well as after treatment with the cholinergic agonist pilocarpine alone or in combination with lithium at a dose that induces seizures only in ~ lithium pretreated rats. Neither acute nor chronic lithium treatment altered the hippocampal or cortical concentration of Ins 1,4,5P3, cyclic AMP, or cyclic GMP. Pilocarpine administered alone increased ins 1,4,5P3 in both regions, did not alter cyclic AMP, and slightly increased cyclic GMP in the cortex. Coadministration of lithium plus pilocarpinz caused large increases in the concentrations of all three second messengers and the production of each of them was uniquely attenuated: lithium reduced pilocarpine-induced increases of Ins 1,4,5P3 in the cortex at 60 rain; chronic lithium administration redfaced stimulated cyclic AMP production in the hippocampus; and chronic lithium treatment impaired stimulated cyclic GMP production in both regions. In summary, chronic lithium treatment appeared only to reduce lns 1,4.5P3 and cyclic AMP concentrations after a long period of stimulation whereas cyclic GMP production was reduced by chronic lithium administration after both short and long periods of stimulation. Thus cyclic CaMP was most sensitive to lithium and lithium attenuation of second messenger formation may be most importat~t in excessively activated pathways.
Introduction Investigations of the therapeutic mechanism of action of lithium, an important drug in the treatment of bipolar affective disorders, have indicated second messenger systems in the brain as potentially important sites of action. Evidence has been reported that phosphoinositide hydrolysis, cyclic adenosine monophosphate (AMP), and cyclic guanosine monophosphate (GMP) each can be affected by lithium. Therapeutic concentrations of lithium (approximately 1 mM) inhibit inositol monoFrom the Department of Psychiatry and Behavioral Neurobiology, University of Alabama .',t Birmingham, Bilmingham, AL (RSJ, L3, KK) and Delpartment of Pharmacology, Medical School, Lublin, Poland (KK). Address reprint requests to Dr. Richard S. Jope, Department of Psychiatry and Behavioral Neurobiology, Sparks Center 910, UAB Station, Birmin,?ham, AL 35294. Received August 20, 1991; revised October 21, 1991. © 1992 Society of Biological Psychiafry
0006-3223/92/$05.00
506
BIOL PSYCHIATRY 1992;31:505-514
R.S. Jope et al
phosphatase, resulting in increased inositol monophosphate concentrations in brain slices in vitro and increases iri rat brain regions in vivo after lithium treatment (reviewed by Sherman et al 1986). Chronic lithium treatment in vivo results in impaired agoniststimulated phosphoinositide hydrolysis measured in rat brain slices (Casebolt and Jope 1987, 1989; Kendall and Nahorski 1987; Elphick et al 1988). Impairment of phosphoinositide hydrolysis by lithium has been suggested to be due to depletion of phosphoinositides (Berridge et al 1982) or to impaired G-protein function (Avissar et al 1988). However, the measurements of phosphoinositide responses in slices after chronic lithium treatment were limited by methodological requirements (e.g., preparation and mcubation of slices and use of [3H]inositol), which could affect detection of the in vivo actions of lithium. Indeed, a recent study in mice reported that chronic lithium treatment increased, rather than decreased, the mass of endogenous inositol 1,4,5 trisphosphate (Ins 1,4,5P3) in the cerebral cortex (Whitworth et al 1990). This latter approach more directly measures the in vivo effects of lithium on the production of the major inositol phosphate second messenger than do in vitro assays with slices, and it Js certainly of interest that enhancement, rather than impairment, of phosphoinositide metabolism was observed in that study. Therefore, in the present investigation the in situ concentration of Ins 1,4,5P3 was measured in rat brain regions to directly determine the effects of chronic and acute lithium administration on phosphoinositide metabolism. Lithium can also impair the production of cyclic AMP or cyclic GMP. A therapeutic concentration of lithium in vitro directly inhibits adenylate cyclase (Mork and Geisler 1987; Newman and Belmaker 1987), whereas evidence suggests an additional inhibitory siee of action of lithium in vivo, possibly at the level of the G proteins, when cyclic AMP was measured with in vitro assays (Ebstein et al 1980; Avissar et al 1988; Mork and Geisler 1989). However, most of the greatest inhibitory effects of lithium were observed using concentrations greater than I mM and the in vivo effects of lithium on cyclic AMP are generally less impressive than the in vitro effects. However, a recent study reported a 60% decrease of the in vivo cyclic AMP concentration in rat cortex after 3 weeks of lithium treatment (Harvey et al 1990). The same study also reported a 90% increase in the cortical cyclic GMP concentration after lithium treatment (Harvey et al 1990). This contrasts with the reported inhibition by lithium of cyclic GMP production measured in vitro (Kanba et al 1985; 1986; Schubert and Miiller 1990), but the latter studies generally used lithium concentrations greater than 1 mM. These wide-ranging, generally inhibitory effects of lithium on second messenger systems prompted this investigation of the in vivo effects of chronic lithium treatment on the concentrations of three second messengers, Ins 1,4,5P3, cyclic AMP, and cyclic GMP. This approach has the advantage that in vivo alterations of these second messengers can be directly measured using a therapeutically relevant protocol for lithium administration without introducing potential artifacts arising from in vitro preparations of brain tissue. However, it suffers from the inability to manipulate individually each neurotransmitter system coupled with the second messengers. Therefore, widespread effects of lithium can be identified, but limited influences on discrete systems may not be detected. To address the question of effects on stimulated, as well as basal, production of second messengers, the cholinergic agonist pilocarpine was administered to some rats, both with and without lithium pretreatment. In rats pretreated with lithium acutely (3 mmol/kg) or chronically, administration of pilocarpine induces gent,ralized convulsive status epileptitus ~Honchar et al 1983; Jope et al 1986; Morrisett et 81 1987; Ormandy et al 1989). Without lithium pretreatment this dose of pilocarpine (30 mg/kg) causes no seizure
Second Messengers in Brain
BIOL PSYCHIATRY ! 992 ;3 ! :505-514
507
Table 1. Second Messenger Concentrations in Control Rat Brainsa Ins 1,4,5P3 (pmol/mg tissue) Hippocampus Cerebral cortex
19.0 - 1.0 17.1 -- 1.2
(n = 28) (n = 30)
cyclic A M P (pmol/mg protein) 4.7 _+ 0.3 4.0 _ 0.2
(n = 20) (n = 20)
cyclic G M P (pmol/mg protein) 0.57 _ 0.05 0.39 _ 0.04
(n = 20) (n = 20)
°Adult, male rats were killed by focused beam microwave irradiation, regions were removed and weighed, and second messenger concentrations were measured as described in the Methods section. Means -- SEM.
activity, but with lithium there is a very reproducible development of seizures observable by EEG consisting of paroxysmal spikes and spike trains 15-20 rain after pilocarpine followed in a few minutes by the onset of status epilepticus, which continues unabated for several hours (Morrisett et al 1987). After the cholinergic receptor-mediated initiation of seizures, status epilepticus, is generalized and unaffected by atropine administration, indicating that the stimulation is no longer limited to cholinergic receptors (Ormandy et al 1989). Thus, lhis combination of drugs allows for the measurement of the responses of second messenger systems to a strong stimulus after either acute or chronic lithium treatment. Methods Male, Sprague-Dawley rats were maintained on a 12 hr light/dark cycle and weighed 225-275 g at the time of death. For chronic lithium treatment, rats were fed pelleted rat chow containing LiCI (1.696 g/kg diet; Teklad, Madison, WI) for 4 weeks. Food, water, and 0.9% saline (to prevent lithium toxicity) were provided ad libitum. This method of lithium administration produces plasma lithium concentrations of approximately 0.8 mM and the body weights of lithium-treated rats do not differ significantly from controls (Casebolt and Jope 1991). Other rats were given LiCl acutely (3 mmol/kg IP) and/or pilocarpine (30 mg/kg SC; 20 hr after acute LiCk or NaCI for controls). Rats receiving pilocarpine were also given N-methylatropine (5 mg/kg) in the same injection to block peripheral cholinergic stimulation. All rats were killed by a beam of microwave irradiation focused on the head (Gerling Labs, Modesto, CA) to prevent postmortem changes in the concentrations of the second messengers. The concentration of Ins 1,4,5P3 was measured by a radioreceptor binding assay (Challis et al 1988) available commercially (Amersham, Arlington Heights, IL) using the protocol exactly as described by Whitworth et al (1990). The concentration of cyclic AMP was measured using a protein binding assay (Brown et al 1971) as described previously (Johnson and Jope 1987). The concentratioa of cyclic GMP was measured by radioimmunoassay usirg a commercial kit (Amer~ham). Data were analyzed by analysis of variance (ANOVA) and differences from control were considered statistically significant when p < 0.05.
Results The concentrations of Ins 1,4,5P3, cyclic AMP, and cyclic GMP in control rat brain regions, wJ.th which samples from experimental rats were compared, are given in Table I. Some rats were treated with pilocarpine alone at a dose (30 mg/kg), which causes no
508
R.S. Jope et al
BIOL PSYCHIATRY 1992;31:505-514
CORTEX
HIPPOCAMPUS [] CONTROL =~ []
50
I 40 30
PILO20 PILO60
40
[]
S=
CONTROL
[] P LO 20rnin
[
3o
~"
20
20
~
I0 0
1o o
NOUCL
A~te HCI
Chto~c IJCL
NOLi@
~=ute LiCL
Chronic I.i~
Figure 1. Ins 1,4,5P3 concentrations in rat hippocampus and cerebral cortex. Rats were treated with saline (controls, n = 28-30), pilocarpine [PILO; 30 mg/kg SC, 20 min (n = 10) or 60 rain (n = 7) prior], acute LiCI [3 mmol/kg IP, 20 hr prior (n = 9)], acute LiCI plus pilocarpine [20 min (n = 9), or 60 rain (n = 8) prior], cta'onic L!, [4 weeks (n = 11)], or chronic Li plus pilocarpine [20 rain (n = 10) or 60 rain (n = 13)1. Mean -4- SEM. *p < 0.05 (ANOVA). seizure activity (Jope et al 1986) or after acute or chronic lithium treatment. Administration of this dose of pilocarpine to rats treated acutely or chronically with lithium causes generalized convulsive status epilepticus with seizures beginning approximately 20 min after pilocarpine and status epilepticus continuing unabated for several hours (Morrisett et al 1987). Ill the following experiments the effects of each drug alone were measured on the corlcentrations of second messengers and the combination of pilocarpine plus acute or chronic lithium pretreatment was applied :~ observe the responses of second messengers to a large stimulus. In the hippocampus pilocarpine administration induced a 54% increase in the conce,tration of Ins 1,4,5P3 after 20 min followed by a larger increase (110%) at 60 min (Figure 1). Neither acute nor chronic lithium treatment altered Ins 1,4,5P3, but administration of pilocarpine to either group caused a larger increase (116% and 127%, respectively) in ins 1,4,5P.~ at 20 min than did pilocarpine in lithium-free rats, and Ins 1,4,5P3 remained elevated at 60 min in lithium-pretreated rats. In the cortex, the responses to each drug alone were the same as observed in the hippocampus, but there were differences after administration of both drugs. Thus, pilocarpine increased the concentration of Ins 1,4,5P3 by 37% at 20 min and it increased further (by 83%) at 60 min, while neither acute nor chronic lithium altered Ins 1,4,5P3. Administration of pilocarpine to either group of lithium-treated animals caused an increase in the concentration of Ins 1,4,5P3 at 20 rain (by 71% with acute and 101% with chronic lithium), ~ollowed by a decrease to (acute lithium) or below (chronic lithium) control values at 6,0 min. These decreases are in contrast to the increases in the lithium-free rats and to the responses to both drugs in the hippocampus where there was a further increase at 60 min. The co.ncenuation of cyclic AMP in the hippocampus and in the cerebral cortex was unchanged by acute or chronic lithium treatment (Figure 2). Administration of pilocarpine to lithium-free rats did not significantly alter the cyclic AMP concentration in either region, although there was a tendency towards an increase in the hippoeampus. Pilocarpine given to rats after acute lithium treatment induced a rise in cyclic AMP in the cortex prior to seizures (10 min), but in the hippocampus the slight increase was not statistically significam. Cyclic AMP was significantly increased in both regions at the time coinciding with initiation of seizures (20 min) and it remained elevated during status epilepticus (60 min). In rats treated chronically with lithium, pilocarpine induced an increase in the cyclic
Second Messengers in Brain
HIPPOCAMPUS
509
BIOL PSYCHIATRY 1992;31:505-514
CORTEX
iii
Q
T Q.--
_I
"
%7
I1
~
<
I
g/l IJ
o~'s
T U
°'4
,
3
•- 9 4 ,
C
P
60
P
P ACUTE Li
3
p 60
P
I0 20 60 = "
/
C
,~CHRONIC Li "-~
P 60
P I0 4--.---
P P 20 60 ACUTE Li ~-
P
60 CHROCqC .q~.-- L i--~
Figure 2. Cyclic AMP concentrations in rat hippocampus and cerebral cortex. Rats were treated with saline (C; controls, n = 18), pilocarpine (P; 30 mg/kg, SC, 60 min prior, n = 6), acute LiCi (30 mmol/kg, ip, 20 hr prior, n = 3), acute LiCl plus pilocarpine followed by measurements at 10 min (before seizures, n = 5), 20 min (at initiation of seizures, n - 6) or 60 rain (during status epilepticus, n = 6), chronic Li (4 weeks, n = 10), or chronic Li plus pilocarpine (sacrificed after 60 min during seizu~'es, n = 6). Hatched bars indicate rats that were seizing. Mean _+ SEM. *p < 0.05 (ANOVA).
AMP concentration in the cortex, but there was no increase in cyclic AMP in the hippocampus. In the hippocampus, pilocarpine induced only a small rise (17%) in cyclic GMP after 20 min and it returned to the control va]lue by 60 rain (Figure 3). Acute lithium did not significantly reduce cyclic GMP, and with pilocarpine administration the cyclic GMP concentration increased at 20 min (p < 0.05 compared with acute lithium-treated group), and by 60 min cyclic GMP had increased to a very high concentration. Chronic lithium administration only slightly reduced (15%) cyclic GMP, and with pilocarpine there was an increase in cyclic GMP similar to that observed with acute lithium at 20 min, but at 60 min the cyclic GMP concentration had not risen any further.
IilPPOCAMPUS '~ 2
CORTEX
n CONTROL F'/I PILO20
e
I1 PIL0 60
~ 2 o ~.
,
~
[] !~
CONTROL PILO20
~
PILO60
gg
"6
"6
...1
.J
NOI.iCL
,O¢uteIJCl.
Chm~ IJCI
NOI.iCL
/Cute IJCL
Chron¢ bCL
Figure 3. Cyclic (IMP concentrations in rat hippocampus and cerebral cortex. Rats were treated with saline (controls, n = 14-.20), pilocarpine [pilo; 30 mg/kg, SC, 20 min (n = 7) or 60 rain (n = 6) prior], acute LiCl Flus pilocarpine followed by measurements at 20 min (at initiation of seizures; n = 4) or 60 min (during status epilepticus; n = 4), chronic Li (4 weeks; n = 10), or chronic Li plus pilocarpine followed by measurements at 20 min (n = 5) or 60 min (n = 8). Mean -+ SEM, *p < 0.05 (ANOVA).
510
BIOL PSYCH)ATRY 1992;31:505-514
R.S. Jope et al
The pattern of cyclic GMP responses to lithium and pilocarpine in the cortex was generally similar to the hippocampus, with pilocarpine causing an increase at 20 min (30%) that was somewhat, but not entirely, reduced at 60 min. Acute lithium did not significantly change cyclic GMP, but with pilocarpine there was a greater increase than in the lithium-free rats at 20 min and a very large increase at 60 min. Chronic lithium also only slightly reduced cyclic GMP (22%), and with pilocarpine treatment there was a clear attenuation of cyclic GMP accumulation, with only a slight increase in cyclic GMP at 20 min, and at 60 min an increase that was only about half of that seen after acute lithium p!!ls pilocarpine. Discussion This investigation measured the effects of lithium treatment on second messenger concentrations in rat cerebral cortex and hippocampus. Both acute and chronic lithium treatments were used to attempt to distinguish those effects that might be relevant to the therapeutic response to lithium, which requires at least 2 weeks of treatment. Pilocarpine administration was included because it causes seizures after either acute or chronic lithium treatment, which are virtually indistinguishable behaviorally or by EEG recordings (Morriser et al 1987), thus providing a means to induce a generalized stimulation after each of these treatments. Therefore, we were able to test the hypotheses that lithium administration impairs production of each of the second messengers under basal or stimulated conditions after either acute or chronic lithium administration. Although these seizures are initiated by cholinergic stimulation, their maintenance has been shown to be more generalized and no longer dependent on activation of cholinergic receptors (Ormandy et al 1989). The concentration of Ins 1,4,5P3 was unaltered by acute or chronic lithium treatment. Also, neither method of lithium administration in,paired the rapid rise of Ins 1,4,5P~ induced by administration of pilocarpine. These results are compatible with the recent finding in mouse cerebral cortex that chronic lithium treatment did not reduce Ins 1,4,5P3 (Whitwol~h et al 1990), although in mouse there was a significant increase in Ins 1,4,5P.~ after chronic lithium treatment, whereas in the present study in rats there was no significant change from controls. The most widely disseminated hypothetical mechanism of action of lithium purposes that phosphoinositides and second messengers derived from them are depleted due to inositol deficiency resulting from its impaired recycling when lithium inhibits inositol monophosphatase (Berridge et al 1982). However, large decreases in the concentrations of phosphoinositides after lithium treatment have not be consistently observed (Sherman et al 1986; Joseph et al 1987; Honchar et al 1989), brain inositol may not be severely reduced after chronic lithium t~eatment (Sherman et al 19115; Hirvonen and Savolainen 1991), and it remains unknown *o what degree inositol mu~;;tbe depleted before it does become rate limiting in the syn~ilesis of phosphoinositides. The present and a previous study (Whitworth et al 1990) show that chronic lithium treatment neither decreases brain Ins 1~4,5P3 concentrations nor blocks its production upon stimulation unless it is very prolonged. Thus, there is no direct evidence that phosphoinositides were depleted by lithium treatment. Another hypothesis suggests that lithium may impair Gprotein coupling (Avissar et al 1988), but direct confirmation of these results has not been published. Several studies have shown that brain slices from lithium-treated rats hydrolyze pho~,phoinositides upon stimulation to a lesser degree than do controls, but as
Secona Messengers in Brai~
BIOL PSYCHIATRY 1992;31:505-5!4
511
discussed in those reports the mechanism accounting for this effect is not yet clear (Casebolt and Jope 1987, 1989; Kendall and Naborski 1987; Elphick et al 1988). In the present study, the clearest indication of impaired Ins 1,4,5Ps production after lithium treatment was the low concentration in the cortex 60 rain after pilocarpine administration to acute or chronic lithium-treated rats. This could result from either of the mechanisms discussed above with the caveat that they only become effective after a relatively long stimulus since initial stimulatow responses (20 min after pilocarpine with lithium) were similar to those in lithium-naive rats. Another possibility was raised by Honchar et al (1990), as they proposed that the effect of lithium on phosphoinositide hydrolysis is already present after acute lithium administration and that this persists through chronic treatment. Thus, potential stimulus-induced increases in Ins 1,4,5 P3 during seizures may be equally attenuated in the acute and chronic lithium-treated rats. Finally, as noted in the Introduction, inhibitory effects of lithium may be restricted to a small subset of phosphoinositide-coupled systems that cannot be isolated easily. Neither acute nor chronic lithium treatment ~tered the cyclic AMP concentration in rat hippocampus or cerebral cortex. This indicates that lithium treatment does not have a major effect on the basal level of this second messenger. This differs from the reduction reported by Harvey et al (1990) following lithium treatment, perhaps due to the different methods of lithium administration or postmortem changes after decapitation, but the control values cannot be directly compared due to the concentration units reported by those authors (picomoles per incubation tube). In the cortex, seizures induced by pilocarpine administration to lithium-treated rats caused similar increases in the concentration of cyclic AMP after acute and chronic lithium. However, in the hippocampus there was no increase in cyclic AMP during seizures in rats given lithium chronically, whereas seizures caused a substantial increase in cyclic AMP in rats treated acutely with lithium. This suggests that chronic lithium treatment impairs cyclic AMP synthesis during stimulation, in agreement with previous reports that lithium inhibits stimulated cyclic AMP production measured with in vitro assays (Ebstein et al 1~0; Mork and Geisler 1987, 1989; Newman and Belmaker 198"/; Avissar et al 1988), but the effect was limited to chronic lithium treatment and to the hippocampus and it was not a large impairment. Nonetheless, if this impairment is restricted to only a portion of the cyclic AMP-generating cells present in the hippocampus it would be likely ~o have a profound effect on signal transduction. The concentration of cyclic GMP was slightly reduced by acute lithium treatment in both brain regions that were examined, comprising decreases of 13% and 28% in the cortex and hippocampus, respectively. Upon administration of pil~arpinc af:er acute lithium the cyclic GMP concentrations increased to very high values that were much greater than those occurring after pilocarpine ad~'~inistered to lithium-free rats. These increases with lithium plus pilocarpine are likely due to the ~e~zures induced by this treatment, rather than to cholinergic stimulation, as seizures are well-known to increase brain cyclic GMP concentrations (e.g., Ferrendelli 1984). Thus, :~cute lithjnm treatment did not block stimulated cyclic GMP accumulation. Chronic lithium treatment slightly reduced cycii~ GMP itJ the cortex and hippocampus (by 22% and 16%, respectively). The combination of chronic lithium plus pilocarpine revealed the clearest impairment of cyclic GMP accumulation of all of the treatments tested. Upon initiation of seizures (20 min after pilocarpine) cyclic GMP increased similarly to that which occurred with acute lithium plus pilocarpine in the hippocampus, but
512
BIOL PSYCHIATRY 1992;31:505-514
R.S. Jope et al
the cyclic GMP concentration was only half as high in the cortex in the chronic lithiumtreated rats. Furthermore, 60 min after pilocarpine there was a clear attenuation of cyclic GMP accumulation after chronic lithium treatment, as in both regions it was only half of that observed with acute lithium plus pilocarpine, although with both types of lithium treatment rats were undergoing status epilepticus as reported previously (Morrisett et al 1987). These results demonstrate that lithium treatment slightly reduces cyclic GMP concentrations in rat brain regions and that chronic lithium administration clearly attenuates seizure-induced accuraulation of cyclic GMP. These results support previous in vitro findings that lithium impairs cyclic GMP production (Belmaker et al 1980; Kanba et al 1985; 1986; Sc~abett and Miiller 1990). In summary, the responses of all three of the second messenger systems that were examined were attenuated to some degree upon relative long periods of stimulation while basal values were relatively resistent to lithium. Cyclic GMP production was most sensitive to attenuation by lithium treatment. Ins 1,4,5Pa and cyclic AMP production required excessive stimulation to detect attenuation, perhaps indicating that only highly activated pathways utilizing these second messengers will be limited by lithium administration. Whether these effects are due to a single or multiple sites of action of lithium remains to be established, but these findings may support the suggestion (Berridge et al 1982) that the therapeutic effect of lithium is derived in part by attenuation of overly active signal transduction pathways while sparing those operating within a normal range of activity. The authors thank Dot McAdory for manuscript preparation. This work was supported by NIMH grant MH38752 and by the U.S. Army Medical Research and Development Command under contract No. DAMDI7-89-C9037. Opinions, interpretations, and recommendations are those of the authors and are not necessarily endorsed by the U.S. Army. ~n conducting research using animals, the investigators adhered to the Guide for the Care and Use of Laboratory Animals prepared by the Committee on Care and Use of Laboratory Animals of the Institute of Laboratory Animal Resources, National Research Council.
References Avissar S, Schreiber G, Danon A, Belmaker RH (1988): Lithium inhibits adrenergic and cholinergic increases in GTP binding in rat cortex. Nature 440-442. Belmaker RH, Kon M, Ebstein RP, Oasberg H (1980): Partial inhibition by lithium of the epinephrine-stimulated rise in plasma cyclic GMP in humans. Biol Psychiatry 15:3-8. Berridge MJ, Downes CP, Hanley MR (1982): Lithium amplifies agonist-dependent reslxmses in brain and salivary glands. Biochem J 206:587-595. Brown BL, Albano JDM, Elkins R?, 5~i~erzi AM (1971): A simple, sensitive saturation assay memod for the measurement of adeaosine Y-5'-cyclic monophosphate. Biochem J 121:561562. Casebolt TL, Jope RS (1987): Chronic lithium treatment reduces norepinephrine-stimulated inositol phospholipid hydrolysis in rat cortex. Eur J Pharmacol 140:245-246. Casebolt TL, Jope RS (1989): Long term lithium treatment selectively reduces receptor-coupled inositol phospholipid hydrolysis in rat brain. Biol Psychiatry 25:329-340. Casebolt TL, Jope RJ (I991): Effects of chronic lithium treatment on protein kinase C and cyclic AMP de~ndentprotein phosphorylation. Biol Psychiatry 29:233-243. Challis RAH, Batty IA, Nahorski SR (1988)" Mass measurements of inositol (1,4,5) trisphosphate in rat cerebral cortex slices using a radioreceptor assay: Effects of neurotransmitters and depolarization. Biochem Biophys Res Commun 157:684-694.
Second Meszcngers in Brain
BIOL PSYCHIATRY 1992;31:505-514
513
Ebstein RP, Hermoni M, Belmaker RH (19S0): The effect of lithium on noradrenaline-induced cyclic AMP accmnulation in rat brain: Inhibition after chronic treatment and absence of supersensitivity. J Pharmacol Exp Ther 212:161-167. Elphick M, Taghavi Z, Powell T, Godfrey PP (1988): Alteration of inositol phospholipid metabolism in rat cortex by lithium but not carbamazepine. Fur J Pharmacol 156:411-414. Ferrendelli JA (1984): Roles of biogenic amines and cyclic nucleotides in seizure mechanisms. Ann Neurol 16 (suppl):S98-S 103. Harvey B, Carstens M, Taljaard J (1990): Lithium modulation of cortical cyclic nucleotides: Evidence for the Yin-Yang hypothesis. Eur J Pharmacol 175:129-- 136. Hirvonen M-R, Savolainen K (1991): Lithium-induced decrease of brain inositol and increase of brain inositol-l-phosphate is transient. Neurochem Res 16:905-911. Honchar MP, Olney JW, Sherman WR (1983): Systematic cholinergic agents induce seizures and brain damage in lithium-treated rats. Science 220:323-325. Honchar MP, Ackerman KE, Sherman WR (1989): Chronically administered lithium alters neither myo-inositol monophosphatase activity nor phosphoinositide levels in rat brain. J Neurochem 53:590-594. Honchar MP, Vogler GP, Gish BG, Sherman WR (1990). Evidence that phosphoinositide metabolism in rat cerebral cortex stimulated by pilocarpine, physostigmine, and pargyline in vivo is not changed by chronic lithium treatment. J Neurochem 55:1521-1525. Johnson GVW, Jope RS (1987). Aluminum alters cyclic AMP and cyclic GMP levels but not presynaptic cholinergic markers in rat brain in vivo. Brain Res 403:1-6. Jopc KS, Morrisett RA, Snead OC (1986). Characterization of lithium potentiation of pilocarpineinduced status epilepticus in rats. Exp Neurol 91:471-480. Joseph NE, Renshaw PF, Leigh JS Jr (1987). Systemic lithium administration alters rat cerebral ~'ortex phospholipids. Biol Psychiatry 22:540-544. Kan~a S, Pfenning M, Ricbelson E (1985). Lithium ions inhibit function of low but not high affinity muscarinic receptors of murine neuroblastoma cells (clone NIE-1 !.5). Psychopharmacology 86:413-416. Kanba S, Pfenning M, Kanba KS, Richelson E (1986)" Lithium ions have a potent and selective inhibitory effect on cyclic GMP formation stimulated by neurotensin, angiotensin II and bradykinin. Fur J Pharmacol 126: I I 1. 116. Kendall DA, Nahorski SR (1987): Acute and chrov,ic litl~ium treatments influence agonist and depolarization-stimulated inositol phospholipid hydrolysis in rat cerebral cortex. J. Pharmacol Exp Ther 241:1023-1027. Mork A, Geisler A (1987): Mode of action of lithium on the catalytic unit of adenylate cyclase from rat brain. Pharmacol ToMcol 60:241-248. Mork A, Geisler A (1989): Effects of GTP on hormone stimulated adenylate cyclase activity in cerebral cortex, striatum, and hippocampus from rats treated chronically with lithium. Biol Psychiatry 26:279-288. Morfisett RA, Jot:c RS, Snead OC III (1987): Status epilepticus is produced by administration of cholinergic agonists to lithium-treated rats: Comparison with kainic acid. Exp Neurol 98:594605. Newman ME, Belmaker RH (1987): Effect:~ of lithium in vitro and ex vivo on components of the adenylate cyclase system in membranes from the cerebral cortex of the rat. Neurooi~armacology 26:211-217. Ormandy GC, .lopo RS, Snead OC (1989): Anticonvulsant actions of MK-801 on the lithiumpilocarpine model of status epilepticus in rats. Exp Neurol 106:172-180. Schubert T, Mfillcr WE (1990): Lithium but not cholincrgic ligands influence guanylate cyclase activity in intact human lymphocytes. Biochem Pharmacol 39:439-444.. Sherman WR, Munsell LY, Olsh BG, Honchar MP (1985): Effects of systemically administered lithium on phosphoinositide metabolism in rat brain, kidney, and testis. J Neurochem 44:798807.
514
BIOLPSYCHIATRY 1992;31:505-514
R.S. Jope et ~1
Sherman WR, Gish BG, Honchar MP, Munsell LY (1986): Effects of lithium on phosphoinositide metabolism in vivo. Fed Proc 45:2639-2646. Whitworth P, Heal DJ, Kendall DA (1990): The effects of acute and chronic lithium treatment on pilocarpine stimulated phosphoinositide hydrolysis in mouse brain in vivo. Br J Pharmacol 101:39-44.
505
Inositol Trisphosphate, Cyclic AMP, and Cyclic GMP in Rat Brain Regions After Lithium and Seizures Richard S. Jope, 'Ling Song, and Krystyna Kolasa
The mechanism of action of lithium, the primary treatment for bipolar affective disorder, is unknown but may involve inhibin'on of second messenger production in the brain. Therefore, the concentrations of three second messengers, inositol 1,4,5 trisphosphate (lns 1,4,5P3), cyclic adenosine monophosphate (AMP), and cyclic guanosine monophosphate (GMP), were measured in rat cerebral cortex and hippocampus after acute or chronic lithium administration, as well as after treatment with the cholinergic agonist pilocarpine alone or in combination with lithium at a dose that induces seizures only in ~ lithium pretreated rats. Neither acute nor chronic lithium treatment altered the hippocampal or cortical concentration of Ins 1,4,5P3, cyclic AMP, or cyclic GMP. Pilocarpine administered alone increased ins 1,4,5P3 in both regions, did not alter cyclic AMP, and slightly increased cyclic GMP in the cortex. Coadministration of lithium plus pilocarpinz caused large increases in the concentrations of all three second messengers and the production of each of them was uniquely attenuated: lithium reduced pilocarpine-induced increases of Ins 1,4,5P3 in the cortex at 60 rain; chronic lithium administration redfaced stimulated cyclic AMP production in the hippocampus; and chronic lithium treatment impaired stimulated cyclic GMP production in both regions. In summary, chronic lithium treatment appeared only to reduce lns 1,4.5P3 and cyclic AMP concentrations after a long period of stimulation whereas cyclic GMP production was reduced by chronic lithium administration after both short and long periods of stimulation. Thus cyclic CaMP was most sensitive to lithium and lithium attenuation of second messenger formation may be most importat~t in excessively activated pathways.
Introduction Investigations of the therapeutic mechanism of action of lithium, an important drug in the treatment of bipolar affective disorders, have indicated second messenger systems in the brain as potentially important sites of action. Evidence has been reported that phosphoinositide hydrolysis, cyclic adenosine monophosphate (AMP), and cyclic guanosine monophosphate (GMP) each can be affected by lithium. Therapeutic concentrations of lithium (approximately 1 mM) inhibit inositol monoFrom the Department of Psychiatry and Behavioral Neurobiology, University of Alabama .',t Birmingham, Bilmingham, AL (RSJ, L3, KK) and Delpartment of Pharmacology, Medical School, Lublin, Poland (KK). Address reprint requests to Dr. Richard S. Jope, Department of Psychiatry and Behavioral Neurobiology, Sparks Center 910, UAB Station, Birmin,?ham, AL 35294. Received August 20, 1991; revised October 21, 1991. © 1992 Society of Biological Psychiafry
0006-3223/92/$05.00
506
BIOL PSYCHIATRY 1992;31:505-514
R.S. Jope et al
phosphatase, resulting in increased inositol monophosphate concentrations in brain slices in vitro and increases iri rat brain regions in vivo after lithium treatment (reviewed by Sherman et al 1986). Chronic lithium treatment in vivo results in impaired agoniststimulated phosphoinositide hydrolysis measured in rat brain slices (Casebolt and Jope 1987, 1989; Kendall and Nahorski 1987; Elphick et al 1988). Impairment of phosphoinositide hydrolysis by lithium has been suggested to be due to depletion of phosphoinositides (Berridge et al 1982) or to impaired G-protein function (Avissar et al 1988). However, the measurements of phosphoinositide responses in slices after chronic lithium treatment were limited by methodological requirements (e.g., preparation and mcubation of slices and use of [3H]inositol), which could affect detection of the in vivo actions of lithium. Indeed, a recent study in mice reported that chronic lithium treatment increased, rather than decreased, the mass of endogenous inositol 1,4,5 trisphosphate (Ins 1,4,5P3) in the cerebral cortex (Whitworth et al 1990). This latter approach more directly measures the in vivo effects of lithium on the production of the major inositol phosphate second messenger than do in vitro assays with slices, and it Js certainly of interest that enhancement, rather than impairment, of phosphoinositide metabolism was observed in that study. Therefore, in the present investigation the in situ concentration of Ins 1,4,5P3 was measured in rat brain regions to directly determine the effects of chronic and acute lithium administration on phosphoinositide metabolism. Lithium can also impair the production of cyclic AMP or cyclic GMP. A therapeutic concentration of lithium in vitro directly inhibits adenylate cyclase (Mork and Geisler 1987; Newman and Belmaker 1987), whereas evidence suggests an additional inhibitory siee of action of lithium in vivo, possibly at the level of the G proteins, when cyclic AMP was measured with in vitro assays (Ebstein et al 1980; Avissar et al 1988; Mork and Geisler 1989). However, most of the greatest inhibitory effects of lithium were observed using concentrations greater than I mM and the in vivo effects of lithium on cyclic AMP are generally less impressive than the in vitro effects. However, a recent study reported a 60% decrease of the in vivo cyclic AMP concentration in rat cortex after 3 weeks of lithium treatment (Harvey et al 1990). The same study also reported a 90% increase in the cortical cyclic GMP concentration after lithium treatment (Harvey et al 1990). This contrasts with the reported inhibition by lithium of cyclic GMP production measured in vitro (Kanba et al 1985; 1986; Schubert and Miiller 1990), but the latter studies generally used lithium concentrations greater than 1 mM. These wide-ranging, generally inhibitory effects of lithium on second messenger systems prompted this investigation of the in vivo effects of chronic lithium treatment on the concentrations of three second messengers, Ins 1,4,5P3, cyclic AMP, and cyclic GMP. This approach has the advantage that in vivo alterations of these second messengers can be directly measured using a therapeutically relevant protocol for lithium administration without introducing potential artifacts arising from in vitro preparations of brain tissue. However, it suffers from the inability to manipulate individually each neurotransmitter system coupled with the second messengers. Therefore, widespread effects of lithium can be identified, but limited influences on discrete systems may not be detected. To address the question of effects on stimulated, as well as basal, production of second messengers, the cholinergic agonist pilocarpine was administered to some rats, both with and without lithium pretreatment. In rats pretreated with lithium acutely (3 mmol/kg) or chronically, administration of pilocarpine induces gent,ralized convulsive status epileptitus ~Honchar et al 1983; Jope et al 1986; Morrisett et 81 1987; Ormandy et al 1989). Without lithium pretreatment this dose of pilocarpine (30 mg/kg) causes no seizure
Second Messengers in Brain
BIOL PSYCHIATRY ! 992 ;3 ! :505-514
507
Table 1. Second Messenger Concentrations in Control Rat Brainsa Ins 1,4,5P3 (pmol/mg tissue) Hippocampus Cerebral cortex
19.0 - 1.0 17.1 -- 1.2
(n = 28) (n = 30)
cyclic A M P (pmol/mg protein) 4.7 _+ 0.3 4.0 _ 0.2
(n = 20) (n = 20)
cyclic G M P (pmol/mg protein) 0.57 _ 0.05 0.39 _ 0.04
(n = 20) (n = 20)
°Adult, male rats were killed by focused beam microwave irradiation, regions were removed and weighed, and second messenger concentrations were measured as described in the Methods section. Means -- SEM.
activity, but with lithium there is a very reproducible development of seizures observable by EEG consisting of paroxysmal spikes and spike trains 15-20 rain after pilocarpine followed in a few minutes by the onset of status epilepticus, which continues unabated for several hours (Morrisett et al 1987). After the cholinergic receptor-mediated initiation of seizures, status epilepticus, is generalized and unaffected by atropine administration, indicating that the stimulation is no longer limited to cholinergic receptors (Ormandy et al 1989). Thus, lhis combination of drugs allows for the measurement of the responses of second messenger systems to a strong stimulus after either acute or chronic lithium treatment. Methods Male, Sprague-Dawley rats were maintained on a 12 hr light/dark cycle and weighed 225-275 g at the time of death. For chronic lithium treatment, rats were fed pelleted rat chow containing LiCI (1.696 g/kg diet; Teklad, Madison, WI) for 4 weeks. Food, water, and 0.9% saline (to prevent lithium toxicity) were provided ad libitum. This method of lithium administration produces plasma lithium concentrations of approximately 0.8 mM and the body weights of lithium-treated rats do not differ significantly from controls (Casebolt and Jope 1991). Other rats were given LiCl acutely (3 mmol/kg IP) and/or pilocarpine (30 mg/kg SC; 20 hr after acute LiCk or NaCI for controls). Rats receiving pilocarpine were also given N-methylatropine (5 mg/kg) in the same injection to block peripheral cholinergic stimulation. All rats were killed by a beam of microwave irradiation focused on the head (Gerling Labs, Modesto, CA) to prevent postmortem changes in the concentrations of the second messengers. The concentration of Ins 1,4,5P3 was measured by a radioreceptor binding assay (Challis et al 1988) available commercially (Amersham, Arlington Heights, IL) using the protocol exactly as described by Whitworth et al (1990). The concentration of cyclic AMP was measured using a protein binding assay (Brown et al 1971) as described previously (Johnson and Jope 1987). The concentratioa of cyclic GMP was measured by radioimmunoassay usirg a commercial kit (Amer~ham). Data were analyzed by analysis of variance (ANOVA) and differences from control were considered statistically significant when p < 0.05.
Results The concentrations of Ins 1,4,5P3, cyclic AMP, and cyclic GMP in control rat brain regions, wJ.th which samples from experimental rats were compared, are given in Table I. Some rats were treated with pilocarpine alone at a dose (30 mg/kg), which causes no
508
R.S. Jope et al
BIOL PSYCHIATRY 1992;31:505-514
CORTEX
HIPPOCAMPUS [] CONTROL =~ []
50
I 40 30
PILO20 PILO60
40
[]
S=
CONTROL
[] P LO 20rnin
[
3o
~"
20
20
~
I0 0
1o o
NOUCL
A~te HCI
Chto~c IJCL
NOLi@
~=ute LiCL
Chronic I.i~
Figure 1. Ins 1,4,5P3 concentrations in rat hippocampus and cerebral cortex. Rats were treated with saline (controls, n = 28-30), pilocarpine [PILO; 30 mg/kg SC, 20 min (n = 10) or 60 rain (n = 7) prior], acute LiCI [3 mmol/kg IP, 20 hr prior (n = 9)], acute LiCI plus pilocarpine [20 min (n = 9), or 60 rain (n = 8) prior], cta'onic L!, [4 weeks (n = 11)], or chronic Li plus pilocarpine [20 rain (n = 10) or 60 rain (n = 13)1. Mean -4- SEM. *p < 0.05 (ANOVA). seizure activity (Jope et al 1986) or after acute or chronic lithium treatment. Administration of this dose of pilocarpine to rats treated acutely or chronically with lithium causes generalized convulsive status epilepticus with seizures beginning approximately 20 min after pilocarpine and status epilepticus continuing unabated for several hours (Morrisett et al 1987). Ill the following experiments the effects of each drug alone were measured on the corlcentrations of second messengers and the combination of pilocarpine plus acute or chronic lithium pretreatment was applied :~ observe the responses of second messengers to a large stimulus. In the hippocampus pilocarpine administration induced a 54% increase in the conce,tration of Ins 1,4,5P3 after 20 min followed by a larger increase (110%) at 60 min (Figure 1). Neither acute nor chronic lithium treatment altered Ins 1,4,5P3, but administration of pilocarpine to either group caused a larger increase (116% and 127%, respectively) in ins 1,4,5P.~ at 20 min than did pilocarpine in lithium-free rats, and Ins 1,4,5P3 remained elevated at 60 min in lithium-pretreated rats. In the cortex, the responses to each drug alone were the same as observed in the hippocampus, but there were differences after administration of both drugs. Thus, pilocarpine increased the concentration of Ins 1,4,5P3 by 37% at 20 min and it increased further (by 83%) at 60 min, while neither acute nor chronic lithium altered Ins 1,4,5P3. Administration of pilocarpine to either group of lithium-treated animals caused an increase in the concentration of Ins 1,4,5P3 at 20 rain (by 71% with acute and 101% with chronic lithium), ~ollowed by a decrease to (acute lithium) or below (chronic lithium) control values at 6,0 min. These decreases are in contrast to the increases in the lithium-free rats and to the responses to both drugs in the hippocampus where there was a further increase at 60 min. The co.ncenuation of cyclic AMP in the hippocampus and in the cerebral cortex was unchanged by acute or chronic lithium treatment (Figure 2). Administration of pilocarpine to lithium-free rats did not significantly alter the cyclic AMP concentration in either region, although there was a tendency towards an increase in the hippoeampus. Pilocarpine given to rats after acute lithium treatment induced a rise in cyclic AMP in the cortex prior to seizures (10 min), but in the hippocampus the slight increase was not statistically significam. Cyclic AMP was significantly increased in both regions at the time coinciding with initiation of seizures (20 min) and it remained elevated during status epilepticus (60 min). In rats treated chronically with lithium, pilocarpine induced an increase in the cyclic
Second Messengers in Brain
HIPPOCAMPUS
509
BIOL PSYCHIATRY 1992;31:505-514
CORTEX
iii
Q
T Q.--
_I
"
%7
I1
~
<
I
g/l IJ
o~'s
T U
°'4
,
3
•- 9 4 ,
C
P
60
P
P ACUTE Li
3
p 60
P
I0 20 60 = "
/
C
,~CHRONIC Li "-~
P 60
P I0 4--.---
P P 20 60 ACUTE Li ~-
P
60 CHROCqC .q~.-- L i--~
Figure 2. Cyclic AMP concentrations in rat hippocampus and cerebral cortex. Rats were treated with saline (C; controls, n = 18), pilocarpine (P; 30 mg/kg, SC, 60 min prior, n = 6), acute LiCi (30 mmol/kg, ip, 20 hr prior, n = 3), acute LiCl plus pilocarpine followed by measurements at 10 min (before seizures, n = 5), 20 min (at initiation of seizures, n - 6) or 60 rain (during status epilepticus, n = 6), chronic Li (4 weeks, n = 10), or chronic Li plus pilocarpine (sacrificed after 60 min during seizu~'es, n = 6). Hatched bars indicate rats that were seizing. Mean _+ SEM. *p < 0.05 (ANOVA).
AMP concentration in the cortex, but there was no increase in cyclic AMP in the hippocampus. In the hippocampus, pilocarpine induced only a small rise (17%) in cyclic GMP after 20 min and it returned to the control va]lue by 60 rain (Figure 3). Acute lithium did not significantly reduce cyclic GMP, and with pilocarpine administration the cyclic GMP concentration increased at 20 min (p < 0.05 compared with acute lithium-treated group), and by 60 min cyclic GMP had increased to a very high concentration. Chronic lithium administration only slightly reduced (15%) cyclic GMP, and with pilocarpine there was an increase in cyclic GMP similar to that observed with acute lithium at 20 min, but at 60 min the cyclic GMP concentration had not risen any further.
IilPPOCAMPUS '~ 2
CORTEX
n CONTROL F'/I PILO20
e
I1 PIL0 60
~ 2 o ~.
,
~
[] !~
CONTROL PILO20
~
PILO60
gg
"6
"6
...1
.J
NOI.iCL
,O¢uteIJCl.
Chm~ IJCI
NOI.iCL
/Cute IJCL
Chron¢ bCL
Figure 3. Cyclic (IMP concentrations in rat hippocampus and cerebral cortex. Rats were treated with saline (controls, n = 14-.20), pilocarpine [pilo; 30 mg/kg, SC, 20 min (n = 7) or 60 rain (n = 6) prior], acute LiCl Flus pilocarpine followed by measurements at 20 min (at initiation of seizures; n = 4) or 60 min (during status epilepticus; n = 4), chronic Li (4 weeks; n = 10), or chronic Li plus pilocarpine followed by measurements at 20 min (n = 5) or 60 min (n = 8). Mean -+ SEM, *p < 0.05 (ANOVA).
510
BIOL PSYCH)ATRY 1992;31:505-514
R.S. Jope et al
The pattern of cyclic GMP responses to lithium and pilocarpine in the cortex was generally similar to the hippocampus, with pilocarpine causing an increase at 20 min (30%) that was somewhat, but not entirely, reduced at 60 min. Acute lithium did not significantly change cyclic GMP, but with pilocarpine there was a greater increase than in the lithium-free rats at 20 min and a very large increase at 60 min. Chronic lithium also only slightly reduced cyclic GMP (22%), and with pilocarpine treatment there was a clear attenuation of cyclic GMP accumulation, with only a slight increase in cyclic GMP at 20 min, and at 60 min an increase that was only about half of that seen after acute lithium p!!ls pilocarpine. Discussion This investigation measured the effects of lithium treatment on second messenger concentrations in rat cerebral cortex and hippocampus. Both acute and chronic lithium treatments were used to attempt to distinguish those effects that might be relevant to the therapeutic response to lithium, which requires at least 2 weeks of treatment. Pilocarpine administration was included because it causes seizures after either acute or chronic lithium treatment, which are virtually indistinguishable behaviorally or by EEG recordings (Morriser et al 1987), thus providing a means to induce a generalized stimulation after each of these treatments. Therefore, we were able to test the hypotheses that lithium administration impairs production of each of the second messengers under basal or stimulated conditions after either acute or chronic lithium administration. Although these seizures are initiated by cholinergic stimulation, their maintenance has been shown to be more generalized and no longer dependent on activation of cholinergic receptors (Ormandy et al 1989). The concentration of Ins 1,4,5P3 was unaltered by acute or chronic lithium treatment. Also, neither method of lithium administration in,paired the rapid rise of Ins 1,4,5P~ induced by administration of pilocarpine. These results are compatible with the recent finding in mouse cerebral cortex that chronic lithium treatment did not reduce Ins 1,4,5P3 (Whitwol~h et al 1990), although in mouse there was a significant increase in Ins 1,4,5P.~ after chronic lithium treatment, whereas in the present study in rats there was no significant change from controls. The most widely disseminated hypothetical mechanism of action of lithium purposes that phosphoinositides and second messengers derived from them are depleted due to inositol deficiency resulting from its impaired recycling when lithium inhibits inositol monophosphatase (Berridge et al 1982). However, large decreases in the concentrations of phosphoinositides after lithium treatment have not be consistently observed (Sherman et al 1986; Joseph et al 1987; Honchar et al 1989), brain inositol may not be severely reduced after chronic lithium t~eatment (Sherman et al 19115; Hirvonen and Savolainen 1991), and it remains unknown *o what degree inositol mu~;;tbe depleted before it does become rate limiting in the syn~ilesis of phosphoinositides. The present and a previous study (Whitworth et al 1990) show that chronic lithium treatment neither decreases brain Ins 1~4,5P3 concentrations nor blocks its production upon stimulation unless it is very prolonged. Thus, there is no direct evidence that phosphoinositides were depleted by lithium treatment. Another hypothesis suggests that lithium may impair Gprotein coupling (Avissar et al 1988), but direct confirmation of these results has not been published. Several studies have shown that brain slices from lithium-treated rats hydrolyze pho~,phoinositides upon stimulation to a lesser degree than do controls, but as
Secona Messengers in Brai~
BIOL PSYCHIATRY 1992;31:505-5!4
511
discussed in those reports the mechanism accounting for this effect is not yet clear (Casebolt and Jope 1987, 1989; Kendall and Naborski 1987; Elphick et al 1988). In the present study, the clearest indication of impaired Ins 1,4,5Ps production after lithium treatment was the low concentration in the cortex 60 rain after pilocarpine administration to acute or chronic lithium-treated rats. This could result from either of the mechanisms discussed above with the caveat that they only become effective after a relatively long stimulus since initial stimulatow responses (20 min after pilocarpine with lithium) were similar to those in lithium-naive rats. Another possibility was raised by Honchar et al (1990), as they proposed that the effect of lithium on phosphoinositide hydrolysis is already present after acute lithium administration and that this persists through chronic treatment. Thus, potential stimulus-induced increases in Ins 1,4,5 P3 during seizures may be equally attenuated in the acute and chronic lithium-treated rats. Finally, as noted in the Introduction, inhibitory effects of lithium may be restricted to a small subset of phosphoinositide-coupled systems that cannot be isolated easily. Neither acute nor chronic lithium treatment ~tered the cyclic AMP concentration in rat hippocampus or cerebral cortex. This indicates that lithium treatment does not have a major effect on the basal level of this second messenger. This differs from the reduction reported by Harvey et al (1990) following lithium treatment, perhaps due to the different methods of lithium administration or postmortem changes after decapitation, but the control values cannot be directly compared due to the concentration units reported by those authors (picomoles per incubation tube). In the cortex, seizures induced by pilocarpine administration to lithium-treated rats caused similar increases in the concentration of cyclic AMP after acute and chronic lithium. However, in the hippocampus there was no increase in cyclic AMP during seizures in rats given lithium chronically, whereas seizures caused a substantial increase in cyclic AMP in rats treated acutely with lithium. This suggests that chronic lithium treatment impairs cyclic AMP synthesis during stimulation, in agreement with previous reports that lithium inhibits stimulated cyclic AMP production measured with in vitro assays (Ebstein et al 1~0; Mork and Geisler 1987, 1989; Newman and Belmaker 198"/; Avissar et al 1988), but the effect was limited to chronic lithium treatment and to the hippocampus and it was not a large impairment. Nonetheless, if this impairment is restricted to only a portion of the cyclic AMP-generating cells present in the hippocampus it would be likely ~o have a profound effect on signal transduction. The concentration of cyclic GMP was slightly reduced by acute lithium treatment in both brain regions that were examined, comprising decreases of 13% and 28% in the cortex and hippocampus, respectively. Upon administration of pil~arpinc af:er acute lithium the cyclic GMP concentrations increased to very high values that were much greater than those occurring after pilocarpine ad~'~inistered to lithium-free rats. These increases with lithium plus pilocarpine are likely due to the ~e~zures induced by this treatment, rather than to cholinergic stimulation, as seizures are well-known to increase brain cyclic GMP concentrations (e.g., Ferrendelli 1984). Thus, :~cute lithjnm treatment did not block stimulated cyclic GMP accumulation. Chronic lithium treatment slightly reduced cycii~ GMP itJ the cortex and hippocampus (by 22% and 16%, respectively). The combination of chronic lithium plus pilocarpine revealed the clearest impairment of cyclic GMP accumulation of all of the treatments tested. Upon initiation of seizures (20 min after pilocarpine) cyclic GMP increased similarly to that which occurred with acute lithium plus pilocarpine in the hippocampus, but
512
BIOL PSYCHIATRY 1992;31:505-514
R.S. Jope et al
the cyclic GMP concentration was only half as high in the cortex in the chronic lithiumtreated rats. Furthermore, 60 min after pilocarpine there was a clear attenuation of cyclic GMP accumulation after chronic lithium treatment, as in both regions it was only half of that observed with acute lithium plus pilocarpine, although with both types of lithium treatment rats were undergoing status epilepticus as reported previously (Morrisett et al 1987). These results demonstrate that lithium treatment slightly reduces cyclic GMP concentrations in rat brain regions and that chronic lithium administration clearly attenuates seizure-induced accuraulation of cyclic GMP. These results support previous in vitro findings that lithium impairs cyclic GMP production (Belmaker et al 1980; Kanba et al 1985; 1986; Sc~abett and Miiller 1990). In summary, the responses of all three of the second messenger systems that were examined were attenuated to some degree upon relative long periods of stimulation while basal values were relatively resistent to lithium. Cyclic GMP production was most sensitive to attenuation by lithium treatment. Ins 1,4,5Pa and cyclic AMP production required excessive stimulation to detect attenuation, perhaps indicating that only highly activated pathways utilizing these second messengers will be limited by lithium administration. Whether these effects are due to a single or multiple sites of action of lithium remains to be established, but these findings may support the suggestion (Berridge et al 1982) that the therapeutic effect of lithium is derived in part by attenuation of overly active signal transduction pathways while sparing those operating within a normal range of activity. The authors thank Dot McAdory for manuscript preparation. This work was supported by NIMH grant MH38752 and by the U.S. Army Medical Research and Development Command under contract No. DAMDI7-89-C9037. Opinions, interpretations, and recommendations are those of the authors and are not necessarily endorsed by the U.S. Army. ~n conducting research using animals, the investigators adhered to the Guide for the Care and Use of Laboratory Animals prepared by the Committee on Care and Use of Laboratory Animals of the Institute of Laboratory Animal Resources, National Research Council.
References Avissar S, Schreiber G, Danon A, Belmaker RH (1988): Lithium inhibits adrenergic and cholinergic increases in GTP binding in rat cortex. Nature 440-442. Belmaker RH, Kon M, Ebstein RP, Oasberg H (1980): Partial inhibition by lithium of the epinephrine-stimulated rise in plasma cyclic GMP in humans. Biol Psychiatry 15:3-8. Berridge MJ, Downes CP, Hanley MR (1982): Lithium amplifies agonist-dependent reslxmses in brain and salivary glands. Biochem J 206:587-595. Brown BL, Albano JDM, Elkins R?, 5~i~erzi AM (1971): A simple, sensitive saturation assay memod for the measurement of adeaosine Y-5'-cyclic monophosphate. Biochem J 121:561562. Casebolt TL, Jope RS (1987): Chronic lithium treatment reduces norepinephrine-stimulated inositol phospholipid hydrolysis in rat cortex. Eur J Pharmacol 140:245-246. Casebolt TL, Jope RS (1989): Long term lithium treatment selectively reduces receptor-coupled inositol phospholipid hydrolysis in rat brain. Biol Psychiatry 25:329-340. Casebolt TL, Jope RJ (I991): Effects of chronic lithium treatment on protein kinase C and cyclic AMP de~ndentprotein phosphorylation. Biol Psychiatry 29:233-243. Challis RAH, Batty IA, Nahorski SR (1988)" Mass measurements of inositol (1,4,5) trisphosphate in rat cerebral cortex slices using a radioreceptor assay: Effects of neurotransmitters and depolarization. Biochem Biophys Res Commun 157:684-694.
Second Meszcngers in Brain
BIOL PSYCHIATRY 1992;31:505-514
513
Ebstein RP, Hermoni M, Belmaker RH (19S0): The effect of lithium on noradrenaline-induced cyclic AMP accmnulation in rat brain: Inhibition after chronic treatment and absence of supersensitivity. J Pharmacol Exp Ther 212:161-167. Elphick M, Taghavi Z, Powell T, Godfrey PP (1988): Alteration of inositol phospholipid metabolism in rat cortex by lithium but not carbamazepine. Fur J Pharmacol 156:411-414. Ferrendelli JA (1984): Roles of biogenic amines and cyclic nucleotides in seizure mechanisms. Ann Neurol 16 (suppl):S98-S 103. Harvey B, Carstens M, Taljaard J (1990): Lithium modulation of cortical cyclic nucleotides: Evidence for the Yin-Yang hypothesis. Eur J Pharmacol 175:129-- 136. Hirvonen M-R, Savolainen K (1991): Lithium-induced decrease of brain inositol and increase of brain inositol-l-phosphate is transient. Neurochem Res 16:905-911. Honchar MP, Olney JW, Sherman WR (1983): Systematic cholinergic agents induce seizures and brain damage in lithium-treated rats. Science 220:323-325. Honchar MP, Ackerman KE, Sherman WR (1989): Chronically administered lithium alters neither myo-inositol monophosphatase activity nor phosphoinositide levels in rat brain. J Neurochem 53:590-594. Honchar MP, Vogler GP, Gish BG, Sherman WR (1990). Evidence that phosphoinositide metabolism in rat cerebral cortex stimulated by pilocarpine, physostigmine, and pargyline in vivo is not changed by chronic lithium treatment. J Neurochem 55:1521-1525. Johnson GVW, Jope RS (1987). Aluminum alters cyclic AMP and cyclic GMP levels but not presynaptic cholinergic markers in rat brain in vivo. Brain Res 403:1-6. Jopc KS, Morrisett RA, Snead OC (1986). Characterization of lithium potentiation of pilocarpineinduced status epilepticus in rats. Exp Neurol 91:471-480. Joseph NE, Renshaw PF, Leigh JS Jr (1987). Systemic lithium administration alters rat cerebral ~'ortex phospholipids. Biol Psychiatry 22:540-544. Kan~a S, Pfenning M, Ricbelson E (1985). Lithium ions inhibit function of low but not high affinity muscarinic receptors of murine neuroblastoma cells (clone NIE-1 !.5). Psychopharmacology 86:413-416. Kanba S, Pfenning M, Kanba KS, Richelson E (1986)" Lithium ions have a potent and selective inhibitory effect on cyclic GMP formation stimulated by neurotensin, angiotensin II and bradykinin. Fur J Pharmacol 126: I I 1. 116. Kendall DA, Nahorski SR (1987): Acute and chrov,ic litl~ium treatments influence agonist and depolarization-stimulated inositol phospholipid hydrolysis in rat cerebral cortex. J. Pharmacol Exp Ther 241:1023-1027. Mork A, Geisler A (1987): Mode of action of lithium on the catalytic unit of adenylate cyclase from rat brain. Pharmacol ToMcol 60:241-248. Mork A, Geisler A (1989): Effects of GTP on hormone stimulated adenylate cyclase activity in cerebral cortex, striatum, and hippocampus from rats treated chronically with lithium. Biol Psychiatry 26:279-288. Morfisett RA, Jot:c RS, Snead OC III (1987): Status epilepticus is produced by administration of cholinergic agonists to lithium-treated rats: Comparison with kainic acid. Exp Neurol 98:594605. Newman ME, Belmaker RH (1987): Effect:~ of lithium in vitro and ex vivo on components of the adenylate cyclase system in membranes from the cerebral cortex of the rat. Neurooi~armacology 26:211-217. Ormandy GC, .lopo RS, Snead OC (1989): Anticonvulsant actions of MK-801 on the lithiumpilocarpine model of status epilepticus in rats. Exp Neurol 106:172-180. Schubert T, Mfillcr WE (1990): Lithium but not cholincrgic ligands influence guanylate cyclase activity in intact human lymphocytes. Biochem Pharmacol 39:439-444.. Sherman WR, Munsell LY, Olsh BG, Honchar MP (1985): Effects of systemically administered lithium on phosphoinositide metabolism in rat brain, kidney, and testis. J Neurochem 44:798807.
514
BIOLPSYCHIATRY 1992;31:505-514
R.S. Jope et ~1
Sherman WR, Gish BG, Honchar MP, Munsell LY (1986): Effects of lithium on phosphoinositide metabolism in vivo. Fed Proc 45:2639-2646. Whitworth P, Heal DJ, Kendall DA (1990): The effects of acute and chronic lithium treatment on pilocarpine stimulated phosphoinositide hydrolysis in mouse brain in vivo. Br J Pharmacol 101:39-44.