174
BIOL PSYCHIATRY 1990;27:174-182
Acute Effects of Lithium on Synaptic Transmission in Rat Hippocampus Studied in Vitro Yoshiaki Higashitani, Yoshihisa Kudo, Akihiko Ogura, and Hiroshi Kato
Acute effects of lithium on synaptic transmissions in the CAI neurones of rat hippocampus were examined. Perfusion of 2-10 mM lithium chloride (LiCl) produced a dose-dependent increase in the amplitude of field EPSPs, whereas change in the population spikes was variable. The increasing ratio of second field EPSP, which was examined by pairedpulse stimulation, was reduced about 10% by 5 mM LiCl. Intracellularly recorded EPSPs and IPSPs were facilitated by 10 mM LiCl, and the soma membrane was depolarized about 3.2 mV. IntraceUular calcium concentration was measured in single hippocampal neurones using fura-2. Although calcium concentration at rest was approximately 30 nM and was increased to an average of 220 nM by 10 -s M glutamate, 10 mM LiCl had no inflt, ence on it. The effects of Li on calcium-dependent processe~ were not manifested in this study. Variable changes of population spikes may be dependent on the balance between the excitatory and inhibitory postsynaptic potentials during lithium application.
Introduction Lithium has been used extensively in treating many psychiatric disorders, especially manic-depressive illness (Murray 1985), which is thought to be caused by the dysfunction of certain transndtter systems in the brain. Lithium is not only effective for manic episodes, but i~ is also used for some patients with depression. It is interesting that lithium itself has both antimanic and antidepressive properties. Although there are many reports dealing with the therapeutic effects of lithium (Emlich et al. 1982; Meltzer 1986), the mechanism for its clinical action is not understood. A number of investigators have suggested that lithium interfere~ with the Na-K electrogenic pump (Ploeger 1974; Ullrich et al. 1980). On the other hand it has been suggested that lithium modifies calcium-dependent processes and influences specific transmitter systems (Hallcher and Sherman 1980; Dubovsky and Frank 1983; Berridge 1984; Aldenhoff and Lux 1985). In this experiment the acute effects of Li on excitatory and inhibitory synaptic transmissions were examined, using hippocampal slice prepmations. We tried to find out whether these acute effects could be construed as a result of inhibition of the Na-K pump. From the Departments of Neuro-Psychiatry (Y. H.) end Physiology (H. K. ), Yamagata U,iversity School of Medicine, Yamagata; and Department of Neuroscience, Mitsubishi-Kasei Institute of Life Scic~::es, Machioa-shi, Tokyo (A.O., Y.K.) Japan. Address reprint requests to Yoshiaki Higashitani, Department of Neuro-Psychiatry, Yamagata Unive~'~it~,,School of Medicine, Yamagata 990-23, Japan. Received January 26, 1988; revised April 4, 1989. © 1990 Society of Biological Psychiatry
0006-3223/90/$03.50
Effects of Lithium on Synaptic Transmission
BIOL PSYCHIATRY 1990;27:174-182
175
The effects of lithium on intracellular calcium concentrations were investigated in single cultured hippocampal neurones by the use of the intracellular calcium indicator fura-2 (Grynkiewicz et al. 1985; Tsien et al. 1985). Fura-2 is loaded into intact cells by incubating them with a membrane permeant ester derivative. Cytosolic esterases split off the ester groups and leave the membrane-impermeant fura-2 in the cytosol. Increases in fura-2 fluorescence signal increased intracellular calcium.
Methods
Slice Technique Rat hippocampal slices (400 Ixm) were prepared from male Wister rats (150-300 g) with a rotary slicer with a round blade (Kaneda et al. 1986). Slices were maintained in an incubation chamber containing an artificial medium consisting of the folloiying (in mM): NaCI 124, KCI 5.0, NaHePO4 1.25, CaC12 2.0, MgSO4 2.0, NaHCO3 22.0, and glucose 10.0. During the intracellular recording, 2.5 mM calcium was used. The medium was kept at a temperature of 34 _+ I°C and at a pH of 7.4. After incubation for more than 1 hr, a slice was submerged in the medium for experiments. Extracellular potentials were obtained with a glass micropipette filled with the medium (10 MI'I). For intraceUular recordings, a pipette filled with 4 M K-acetate (30-60 M[I) was used. Arrangement of stimulating and recording electrodes is shown in Figure 1. To examine the amplitude change in potentials, the intensity of the stimulus was adjusted to get about 50% of the maximal response. Recordings were begun after responses were stabilized for at least 5 min, and the slice was exposed to a test solution for 10 rain. The test solution was prepared by exchanging an equimolar amount of NaCI in the medium with LiCI.
Fura-2 Technique Fura-2, a fluorescent intracellular calcium indicator, was applied to single-cultured hippocampal neurones. The hippocampus was isolated from a 16-20-day rat embryo following the Banker and Cowan (1977) method. After maintenance for 3-10 days, cells were loaded with fura-2 in a basal salt solution (BSS) consisting of the following (in mM): NaCI 130, glucose 5.5:. KCI 5.4, CaCl21.8, tetrodotoxin0.001, HEPES-NaOH 20.0, and 1.5 IxM fUra2 acetoxymethylester. After 00 rain incubation the cells were mounted on an inverted microscope. The visual field of the microscope was displayed on a cathode ray tube (CRT). The criteria for neurone identification were a large distinct cell body, presence of long neurites, and responsiveness to glutamate and 50 mM KCI stimulation. The intracellulm calcium concentration was estimated from the dual beam excitation ratios of emitted fluorescence (fluorescence intensity of 505 nm wavelength under an excitation light of 340 nm wavelength divided by that under wavelength of 360 nm) (Grynkiewicz et al 1985; Tsien et al 1985). The change of fluorescence was measured before, during, and after the well was perfused with 10 mM LiCI, 50 mM KCI, or IO-5 M glutamate. Results
Slice Technique Field potentials. The amplitude of field EPSPs increased gradually and maintained a high level during the 10-rain application of LiCI. There was a complete recovery, as
176
Y. Higashitani et ai.
BIOL PSYCHIATRY 1990;27:174-182
All
B.
field EPSP
._11my
st.
10ms
/7 population spike
J4mV
4ms intraceUular recording
I
I
1
2
I 2mV 50ms
Figure 1. (A) Diagram of a hippocampal slice with stimulating (st.) and recording (rec.) electrodes. alv., alveus; st. pyr., stratum pyramidale (pyramidal cell layer); st. rad., stratum radiatum. (B) Synaptically evoked potentials. The upper trace is a field EPSP recorded in the st. radiatum. The middle trace is a field population spike recorded in the st. pyramidale. The lower trace is a series of potentials recorded intraceUularly in the st. pyramidale. (*), EPSP; (1), GABA-PSP; (2), Lhp (late hyperpolarizing potential). The amplitude of field EPSPs was measured from the baseline to the point of the maximal negative deflection. The amplitude of the population spikes was measured from the tangent, joining spike onset and offset, to the spike peak. At intracellular recording, the amplitude of EPSPs was measured from the baseline to its peak and that of IPSPs was measured at a fixed latency, about 50 msec and 150 msec after the stimulus.
shown in Figure 2. The increase in amplitude was observed in all 26 slices examined during the 2-10 mM LiCl peffusion (see Figure 3A). Changes in the amplitude of population spikes, on the other hand, were variable without any specific relationship to LiC! concentration (see Figure 3B). The amplitude of the antidromic spikes was constant with less than 5 mM LiCI, but decreased by about 5% with 10 mM LiCI. Presynaptic fiber volley responses showed no significant changes during LiCl perfusion (2-10 mM). These data indicate that field EPSPs increase, but population spikes do not necessarily increase, depending on the concentration of LiCl.
lntraceUular recordings. Intracellular recordings were examined in 36 neurones from 12 rats. The mean membrane potential, spike-amplitude, and membrane resistance were 68 mV, 71 mV, and 49 MfL respectively. Stimulation of afferents to the CA1 neurones elicited a series of potentials (Newberry and Nicoll 1984). These include glutamatemediated EPSP, GABA-IPSP, and Lhp (Alger .,rod Nicoll I982). The membrane potential was depolarized by an average of 3.2 mV during perfusion with 10 mM LiCI (n = 24, p < 0.01; paired-t test), when the mean membrane potential,
Effects of Lithium on Synaptic Transmission
.~3E
BIOL PSYCHIATRY 1990;27:174-182
177
I LiCi 5mM I
v
n
m
2-
O.. hl
•
OoOOO00s
oO000000 • OOoI OOQ
O
..o
.~_ 1-
_
6
io
Figure 2. Example of the effects of 5 mM LiCi on field EPSPs. A stimulating pulse was delivered at a 5-see interval through bipolar, insulated tungsten wires. The amplitude of field EPSPs was increased by 37% and returned to control level within 10 min after removal of LiCI.
to
time ( rain )
spike-amplitude, and membrane resistance were 66 mV, 71 mV, and 50 MI2. But the membrane resistance measured by intracellular current injection showed no significant changes. The amplitude of EPSPs was increased by 31.1 _+ 17.7% (mean _+ SD, n = 8, p < 0.05; paired-t test) when measured from the depolarized membrane potential during perfusion with 10 mM LiCl. As shown in Figure 4, both GABE-IPSP (n = 4) and Lhp (n = 3) were increased in amplitude. As after-hyperpolarizafions (AHPs) are considered to be related to calcium-activated potassium conductance (Hotson and Prince 1980), the effects of lithium on AHPs were examined in three neurones. There was no change in their amplitude or time course, suggesting that calcium concentration may not be altered by lithium in postsynaptic neuroncs.
Paired-pulse facihtation. To examine whether LiCI acts presynaptically, the effects of lithium on paired-pulse facilitation of field EPSPs was measured before and during 5 A. f i e l d
EPSP
80
"(3
D
-40-
O.
-80-
E
B. pop. spike
O ~b OI ¢lu
8040-
•,[
. . . . '...I-. . . . . . .
0-
,I
:-
-40 -80-
2,
5l
LiCI
Ii0
(ram)
Figure 3. Effects of LiCI on field EPSPs (A) and population spikes (B). The amplitude of field EPSPs was increased by 2, 5, and l0 mM LiCl by 14. l 4- 7.7% (mean 4- SD, n = 7), 31.4 4- 14.3% (n = 12), and 49.8 4- 12.9% (n = 7), respectively ( * p < 0 . 0 1 ; **p < 0.005; paired-t test). The amplitude of the population spikes was changed by 3.1 _+ 32.2% (n = 6), 8.0 4. 35.0% (n = 12), and 10.1 4- 27.8% (n - 6), respectively (no significant differences).
178
Y. Higashitani et al.
BIOL PSYCHIATRY 1990;27:174-- 182
11/1
A. GABA-IPSP
1
membrane
potential , (my)
A
E 12.
0
Figure 4. Effects of 10 mM LiCI on two types of IPSPs. The amplitude of IPSPs was measured at a latency of 40 msec (A, GABA-IPSP) and 150 msec (B, Lhp) from the baseline membrane potential. A l0 mM LiCl perfusion caused the amplitude of IPSPs to increase at comparable membrane potentials.
: : control o--o LiCI lOrnM
10"
B. Lhp 1D :~ °D
5.
e~
E
0
:ho
-7O I
I
,L
-5
-10
mM LiCI perfusion. The resalts are shown in Figure 5, where facilitation ratios are plotted agemst first EPSP amplitudes. The facilitation ratio was reduced by about 10% during LiCI perfusion, which suggests that presynaptic mechanisms are involved in the effects of lithium.
Effects of extraceUular calcium concentration. Intracellular calcium concentration was not affected during LiCI peffusion as indicated by the AHP and fura-2 experiments which a
b
40msec =--e control o..o LiCI 5rnM o
150j . . . .
0-, .= . . " " 0
.D
~"
100a
a" 1st
EPSP
(mY)
Figure 5. Effects of LiCI on pairedpulse facilitation. Paired stimulation at 40 msec intervals induced an increase in amplitude of the second field EPSP, as shown in the upper trace. In the lowe~ graph, the facilitation ratio, (sec¢~nd EPSP amplitude/first EPSP amplitude) *100 (%), was plotted against the amplitude of the first EP3P. A 5 mM LiC1 perfusion reduced the facilitation ratio frGm 142.5 e 2.4% to 130.0 -4-_ 10.0% (mean _+ sv, n = 6, p < 0.05; paired-~, tes0, measured at 50% of maxim~d amplitude for the first EPSP.
Effects of Lithium on Synaptic Transmission
PSYCHIATRY 1990;27:174-182
BIOL
179
=I=
60
I
I LiCI 5rnM
0. Q. w
~0'
q,.
o o
20"
!
e-
Figure 6. Effects of extracellular calcium concentration on the action of LiCI. During 5 mM LiCI peffusion, the amplitude of field EPSPs was increased by 10.9 4- 7.1% (mean _ SD, n = 7) with 1.5 mM calcium, and that was increased by 33.2 _ 17.5% (n = 7) with 2.5 mM calcium (*p < 0.01; student's-t test).
lU
O.
!
I
2.5
1.5
[ Ca2+] (mM)
are discussed in the following section. The effects of LiC1, however, may be influenced by the extracellular calcium concentration through a presynaptic mechanism. This was examined using mediums containing 1.5 mM or 2.5 mM calcium. As shown in Figure 6, the facilitating effects of LiC1 on field EPSPs were dependent on extracellular calcium concentration. The higher the extracellular calcium concentration was the more effect LiCI exerted.
Fura-2 Technique After loading fura-2 in the cultured hippucampal neurones, the resting calcium concentration was measured; in the 18 neurones examined, it was approximately 30 nM. As shown in Figure 7A, 10 mM LiCl did not change the resting fluorescence whereas 50 mM KCI increased it in the same neurone. When IO-SM glutamate was applied, fluorescence was increased to 220 _ 74% (mean _+ so, n = 18), which corresponded to about 220 nM in calcium concentration (Figure 7B). As shown in the cell in Figure 7C, 10 mM LiCI had no effect on glutamate-induced fluorescence.
Discussion Haas (1982) has already reported that lithium ions had enhancing effects on field EPSPs and population spikes in rat hippocampal slices. However, our results indicated that the amplitude of field EPSPs was increased by lithium, but that of the population spikes was va.fiable. Two types of IPSPs, GABA-IPSP and Lhp, were observed to be facilitated. Paired-pulse facilitation is thought to result from the presynaptie mechanism (Katz and Miledi 1968; Creager et al. 1980), providing an indicator of the pre~ynaptic function for transmitter release. The changes in facilitation ratios caused by lithium indicate that lithium action may be involved in the presynaptic terminals. Thus, lithium might act on both excitatory and inhibitory presynaptic terminals independent of their functions. These results, that both excitatory and inhibitory transmissions were facilitated, agree with this hypothesis. A number of investigators have sugges ed that lithium interferes with the Na-K pump.
180
Y. Higashitani et al.
BIOL PSYCHIA,':'RY 1990;27:1"/4-182
A.
4..I -PI
i1o
B. 3
I
L iCl,., lOmM
'
I
~KCI
io
'
2o
ilG IG
(/)
BG
IG
,l.J .r-I
U
10
20
'
U
C.
3~
8G
~_ LiCI
lOmM
I
IG
0,-! 21.1.-i
°o
1o
.....
20
time
Figure 7. Effects of LiCI on intracellular calcium concentration. The ordinate shows the fluorescence intensity under an excitation light of 340 nm. (A) The fluorescence at rest was not changed by 10 mM LiCl, but was increased by depolarization with 50 mM KCI. The resting calcium concentration of this neurone was 55 nM. (B) Glutamate, 10-SM, applied for 30 sec (G) induced a remarkable increase in fluorescence. The gradual decline in maximal fluorescence after successive glutamate applications probably reflected desensitization. The glutamate-induced increase in fluorescence was not influenced by l0 mM LiCI.
4"o
( m~n )
Membrane depolarization can be regarded as a result of lithium-induced inhibition of the Na-K electrogenic pump, and this probably occurred more strongly in the presynaptic terminals than in the soma of neurones, due to the volume-surface ratio. Inhibition of the Na-K pump is known to facilitate transmitter release (Brosemer 1985), and does not conflict with the findings in this electrophysiological experiment. The facilitating effect of lithium without extracellular calcium was reposed in the neuromuscular junction (Crawford 1975). In the present study, lithium action on field EPSPs was not independent of extracellular calcium concentration. Exti'acellular calcium may affect lithium action via presynaptic mechanisms. It is possible that the amplitude of population spikes may be dependent on the balance between excitatory and inhibitory postsynaptic potentials. If the facilitation of the former is overwhelming during lithium application, the amplitude of population spikes may be increased. Conversely, wl:en facilitation of the latter is dominant, the opposite result may be induced. This balance: may also be an important factor in the therapeutic action of lithium and manic-depressive illness. If lithium exerts its therapeutic action by affecting calcium-dependent processes (Allison and Stewart 1971, Hallcher and Sherman 1980; Aldenhoff and Lux 1985), some changes in intracellular calcium concentration might be observed. But the calcium concentration of the cell body at rest and its glutamate-induced increase were not influenced
Effects of Lithium on Synaptic Transmission
BIOLPSYCHIATRY 1990;27:174-182
181
by lithium when examined in cultured hippocampal neurones. These results suggest ~ a t calcium concentration in the soma of the neurone were not changed noticeably by acute expc~ure to lithium. This observation is ccns.:~tent with the results obtained from the after-hyperpolarization experiments in which calcium-activated potassium conductance was not affected by lithium. Rinaldi et al. (1986) has reported that the acute effects of lithium on population EPSP in CAI region was not consistent with chronic exposure results: the therapeutic effects of lithium require at least several days to develop. The acute effects of lithium in this study iavoIved higher concentrations (2-10 mM) than the therapeutic range ( < 2 raM), and may not account for the therapeutic effects. The most important finding in this experiment was the presynaptic involvement of lithium. There is a need for further investigation of t~ese presynaptic mechanisms. The authors are grateful to Professor S. Totsuka for continuous support and encouragement and acknowledge the skillful technical assistance of H. Miyakawa, K. Ito, and K. Kaneko.
References Aldenhoff JB, Lux HD (1985): Lithium slows neuronal calcium regulation in the snail Helix Pomatia. Neurosci Lett 54:103-108. Alger BE, Nicoll RA (1982): Feed-forward dendritic inhibition in rat hippocampal cells studied in vitro, l Physiol 328:105-123. Allison JH, Stewart MA (1971): Reduced brain inositol in lithium-treated rats. Nature New Biol 233:267-268. Banker GA, Cowan WM (1977): Rat hippocampal neurons in dispersed cell culture. Broi~ Res 126:397-425. Berridge MJ (1984) Inositol tfiphosphat, and diacylglycerol as second messengers. Biochem J 220:345-360. Brosemer RW (1985): Effect of inhibitors of Na+,K+-ATPase on *.he membr~e potentials and neurotransmitter efflux in ra', brain slices. Brain Res 344:125-137. Creager R, Dunwiddie T, Lynch G (1980): Paired-pulse and frequency facilitation in the CAI region of the in vitro rat hippocampus. J Physiol 299:409--424. Crawford AC (1975): Lithium ions and the release of transmitter at the frog neuromuscular junction. J Physiol 246:109-142. Dubovsky SL, Frank RD (1983): Intracellular calcium ions in affective disorders. Biol Psychiatry 19:781-797. Emlich HM, Aldenhoff JB, Lux JB (1982): Basic mechanisms in the action of lithium. Amsterdam: Excerpta Medica. Gryr~dewicz G, Poenie M, Tsien RY (1985): A new generation of Ca2+ indicators with greatly haproved fluorescence properties. J Biol Chem 260:3440-3450. Haas HL (1982): Lithium and synaptic transmission in the mammalian brain. In Emiich HM (ed), Basic Mechanisms in the Action of Lithium. Amsterdam: Elsevier, pp 71-79. Hallcher CM, Sherman WR (1980): The effects of lithium ion and other agents on the activity of myo-inositol one-r~osphatase from bovine brain. J Biol Chem 255:10896-10901. Hotson JR, Prince DA (1980): A calcium-activated hyperpolarization follows repetitive firing in hippocampal neurons, J Neurophysiol 43:409-419. Kaneda M, Higashitani Y, Ohtani R, Fujii S, Kato H (1986): A rotary slicer for brain research. Famagaca Med 1 4:81-85.
182
BIOLPSYCHIATRY 1990;27:174-182
Y. Higashitani et al.
Katz B, Miledi R (1968): The role of calcium in neuromuscular facilitation. J Physiol 195:481492. Kudo Y, Ogura A (1986): Glutamate-induced increase in intracellular Ca2+ concentration in isolated hippocampal neurones. Br J Pha~'macol 89:191-198. Meltzer HL (1986): Lithium mechanisms in bipolar illness and altered intracellular calcium functions. Biol Psychiatry 21:439-510. Murray JB (1985): Lithium therapy for mania and depression. J Gen Psychol 112:5-33. Newberry NR, Nicoll RA (1984): A bicuculine-resistant inhibitory post-synaptic potential in rat hippocampal pyramidal cells in vitro. J Physiol 348:239-254. Rinaldi PC, Fairchild MD, Kusske JA (1986): Perfusion with lithium modifies neurophysiological responses in the CAI region of the hippocampal slices preparation. Brain Res 375:313-319. Tsien RY, Rink TJ, Poenie M (1985): Measurement of cytosolic free Ca2+ in individual small cells using fluorescence microscopy with dual excitation wavelength. Cell Calcium 6:145-157. Ploeger EJ (1974): The effects of lithium on excitable cell membranes. On the mechanism of inhibition of the sodium pump of non-myelinated nerve fibers of the rat. Eur J Pharmacol 25:316-321. Ullrich A, Baierl P, ten Bmgge~cate G (1980): Extracellular potassium in rat cerebellar cortex during acute and chronic lithium application. Brain Res 192:287-290.
BIOL PSYCHIATRY 1990;27:174-182
Acute Effects of Lithium on Synaptic Transmission in Rat Hippocampus Studied in Vitro Yoshiaki Higashitani, Yoshihisa Kudo, Akihiko Ogura, and Hiroshi Kato
Acute effects of lithium on synaptic transmissions in the CAI neurones of rat hippocampus were examined. Perfusion of 2-10 mM lithium chloride (LiCl) produced a dose-dependent increase in the amplitude of field EPSPs, whereas change in the population spikes was variable. The increasing ratio of second field EPSP, which was examined by pairedpulse stimulation, was reduced about 10% by 5 mM LiCl. Intracellularly recorded EPSPs and IPSPs were facilitated by 10 mM LiCl, and the soma membrane was depolarized about 3.2 mV. IntraceUular calcium concentration was measured in single hippocampal neurones using fura-2. Although calcium concentration at rest was approximately 30 nM and was increased to an average of 220 nM by 10 -s M glutamate, 10 mM LiCl had no inflt, ence on it. The effects of Li on calcium-dependent processe~ were not manifested in this study. Variable changes of population spikes may be dependent on the balance between the excitatory and inhibitory postsynaptic potentials during lithium application.
Introduction Lithium has been used extensively in treating many psychiatric disorders, especially manic-depressive illness (Murray 1985), which is thought to be caused by the dysfunction of certain transndtter systems in the brain. Lithium is not only effective for manic episodes, but i~ is also used for some patients with depression. It is interesting that lithium itself has both antimanic and antidepressive properties. Although there are many reports dealing with the therapeutic effects of lithium (Emlich et al. 1982; Meltzer 1986), the mechanism for its clinical action is not understood. A number of investigators have suggested that lithium interfere~ with the Na-K electrogenic pump (Ploeger 1974; Ullrich et al. 1980). On the other hand it has been suggested that lithium modifies calcium-dependent processes and influences specific transmitter systems (Hallcher and Sherman 1980; Dubovsky and Frank 1983; Berridge 1984; Aldenhoff and Lux 1985). In this experiment the acute effects of Li on excitatory and inhibitory synaptic transmissions were examined, using hippocampal slice prepmations. We tried to find out whether these acute effects could be construed as a result of inhibition of the Na-K pump. From the Departments of Neuro-Psychiatry (Y. H.) end Physiology (H. K. ), Yamagata U,iversity School of Medicine, Yamagata; and Department of Neuroscience, Mitsubishi-Kasei Institute of Life Scic~::es, Machioa-shi, Tokyo (A.O., Y.K.) Japan. Address reprint requests to Yoshiaki Higashitani, Department of Neuro-Psychiatry, Yamagata Unive~'~it~,,School of Medicine, Yamagata 990-23, Japan. Received January 26, 1988; revised April 4, 1989. © 1990 Society of Biological Psychiatry
0006-3223/90/$03.50
Effects of Lithium on Synaptic Transmission
BIOL PSYCHIATRY 1990;27:174-182
175
The effects of lithium on intracellular calcium concentrations were investigated in single cultured hippocampal neurones by the use of the intracellular calcium indicator fura-2 (Grynkiewicz et al. 1985; Tsien et al. 1985). Fura-2 is loaded into intact cells by incubating them with a membrane permeant ester derivative. Cytosolic esterases split off the ester groups and leave the membrane-impermeant fura-2 in the cytosol. Increases in fura-2 fluorescence signal increased intracellular calcium.
Methods
Slice Technique Rat hippocampal slices (400 Ixm) were prepared from male Wister rats (150-300 g) with a rotary slicer with a round blade (Kaneda et al. 1986). Slices were maintained in an incubation chamber containing an artificial medium consisting of the folloiying (in mM): NaCI 124, KCI 5.0, NaHePO4 1.25, CaC12 2.0, MgSO4 2.0, NaHCO3 22.0, and glucose 10.0. During the intracellular recording, 2.5 mM calcium was used. The medium was kept at a temperature of 34 _+ I°C and at a pH of 7.4. After incubation for more than 1 hr, a slice was submerged in the medium for experiments. Extracellular potentials were obtained with a glass micropipette filled with the medium (10 MI'I). For intraceUular recordings, a pipette filled with 4 M K-acetate (30-60 M[I) was used. Arrangement of stimulating and recording electrodes is shown in Figure 1. To examine the amplitude change in potentials, the intensity of the stimulus was adjusted to get about 50% of the maximal response. Recordings were begun after responses were stabilized for at least 5 min, and the slice was exposed to a test solution for 10 rain. The test solution was prepared by exchanging an equimolar amount of NaCI in the medium with LiCI.
Fura-2 Technique Fura-2, a fluorescent intracellular calcium indicator, was applied to single-cultured hippocampal neurones. The hippocampus was isolated from a 16-20-day rat embryo following the Banker and Cowan (1977) method. After maintenance for 3-10 days, cells were loaded with fura-2 in a basal salt solution (BSS) consisting of the following (in mM): NaCI 130, glucose 5.5:. KCI 5.4, CaCl21.8, tetrodotoxin0.001, HEPES-NaOH 20.0, and 1.5 IxM fUra2 acetoxymethylester. After 00 rain incubation the cells were mounted on an inverted microscope. The visual field of the microscope was displayed on a cathode ray tube (CRT). The criteria for neurone identification were a large distinct cell body, presence of long neurites, and responsiveness to glutamate and 50 mM KCI stimulation. The intracellulm calcium concentration was estimated from the dual beam excitation ratios of emitted fluorescence (fluorescence intensity of 505 nm wavelength under an excitation light of 340 nm wavelength divided by that under wavelength of 360 nm) (Grynkiewicz et al 1985; Tsien et al 1985). The change of fluorescence was measured before, during, and after the well was perfused with 10 mM LiCI, 50 mM KCI, or IO-5 M glutamate. Results
Slice Technique Field potentials. The amplitude of field EPSPs increased gradually and maintained a high level during the 10-rain application of LiCI. There was a complete recovery, as
176
Y. Higashitani et ai.
BIOL PSYCHIATRY 1990;27:174-182
All
B.
field EPSP
._11my
st.
10ms
/7 population spike
J4mV
4ms intraceUular recording
I
I
1
2
I 2mV 50ms
Figure 1. (A) Diagram of a hippocampal slice with stimulating (st.) and recording (rec.) electrodes. alv., alveus; st. pyr., stratum pyramidale (pyramidal cell layer); st. rad., stratum radiatum. (B) Synaptically evoked potentials. The upper trace is a field EPSP recorded in the st. radiatum. The middle trace is a field population spike recorded in the st. pyramidale. The lower trace is a series of potentials recorded intraceUularly in the st. pyramidale. (*), EPSP; (1), GABA-PSP; (2), Lhp (late hyperpolarizing potential). The amplitude of field EPSPs was measured from the baseline to the point of the maximal negative deflection. The amplitude of the population spikes was measured from the tangent, joining spike onset and offset, to the spike peak. At intracellular recording, the amplitude of EPSPs was measured from the baseline to its peak and that of IPSPs was measured at a fixed latency, about 50 msec and 150 msec after the stimulus.
shown in Figure 2. The increase in amplitude was observed in all 26 slices examined during the 2-10 mM LiCl peffusion (see Figure 3A). Changes in the amplitude of population spikes, on the other hand, were variable without any specific relationship to LiC! concentration (see Figure 3B). The amplitude of the antidromic spikes was constant with less than 5 mM LiCI, but decreased by about 5% with 10 mM LiCI. Presynaptic fiber volley responses showed no significant changes during LiCl perfusion (2-10 mM). These data indicate that field EPSPs increase, but population spikes do not necessarily increase, depending on the concentration of LiCl.
lntraceUular recordings. Intracellular recordings were examined in 36 neurones from 12 rats. The mean membrane potential, spike-amplitude, and membrane resistance were 68 mV, 71 mV, and 49 MfL respectively. Stimulation of afferents to the CA1 neurones elicited a series of potentials (Newberry and Nicoll 1984). These include glutamatemediated EPSP, GABA-IPSP, and Lhp (Alger .,rod Nicoll I982). The membrane potential was depolarized by an average of 3.2 mV during perfusion with 10 mM LiCI (n = 24, p < 0.01; paired-t test), when the mean membrane potential,
Effects of Lithium on Synaptic Transmission
.~3E
BIOL PSYCHIATRY 1990;27:174-182
177
I LiCi 5mM I
v
n
m
2-
O.. hl
•
OoOOO00s
oO000000 • OOoI OOQ
O
..o
.~_ 1-
_
6
io
Figure 2. Example of the effects of 5 mM LiCi on field EPSPs. A stimulating pulse was delivered at a 5-see interval through bipolar, insulated tungsten wires. The amplitude of field EPSPs was increased by 37% and returned to control level within 10 min after removal of LiCI.
to
time ( rain )
spike-amplitude, and membrane resistance were 66 mV, 71 mV, and 50 MI2. But the membrane resistance measured by intracellular current injection showed no significant changes. The amplitude of EPSPs was increased by 31.1 _+ 17.7% (mean _+ SD, n = 8, p < 0.05; paired-t test) when measured from the depolarized membrane potential during perfusion with 10 mM LiCl. As shown in Figure 4, both GABE-IPSP (n = 4) and Lhp (n = 3) were increased in amplitude. As after-hyperpolarizafions (AHPs) are considered to be related to calcium-activated potassium conductance (Hotson and Prince 1980), the effects of lithium on AHPs were examined in three neurones. There was no change in their amplitude or time course, suggesting that calcium concentration may not be altered by lithium in postsynaptic neuroncs.
Paired-pulse facihtation. To examine whether LiCI acts presynaptically, the effects of lithium on paired-pulse facilitation of field EPSPs was measured before and during 5 A. f i e l d
EPSP
80
"(3
D
-40-
O.
-80-
E
B. pop. spike
O ~b OI ¢lu
8040-
•,[
. . . . '...I-. . . . . . .
0-
,I
:-
-40 -80-
2,
5l
LiCI
Ii0
(ram)
Figure 3. Effects of LiCI on field EPSPs (A) and population spikes (B). The amplitude of field EPSPs was increased by 2, 5, and l0 mM LiCl by 14. l 4- 7.7% (mean 4- SD, n = 7), 31.4 4- 14.3% (n = 12), and 49.8 4- 12.9% (n = 7), respectively ( * p < 0 . 0 1 ; **p < 0.005; paired-t test). The amplitude of the population spikes was changed by 3.1 _+ 32.2% (n = 6), 8.0 4. 35.0% (n = 12), and 10.1 4- 27.8% (n - 6), respectively (no significant differences).
178
Y. Higashitani et al.
BIOL PSYCHIATRY 1990;27:174-- 182
11/1
A. GABA-IPSP
1
membrane
potential , (my)
A
E 12.
0
Figure 4. Effects of 10 mM LiCI on two types of IPSPs. The amplitude of IPSPs was measured at a latency of 40 msec (A, GABA-IPSP) and 150 msec (B, Lhp) from the baseline membrane potential. A l0 mM LiCl perfusion caused the amplitude of IPSPs to increase at comparable membrane potentials.
: : control o--o LiCI lOrnM
10"
B. Lhp 1D :~ °D
5.
e~
E
0
:ho
-7O I
I
,L
-5
-10
mM LiCI perfusion. The resalts are shown in Figure 5, where facilitation ratios are plotted agemst first EPSP amplitudes. The facilitation ratio was reduced by about 10% during LiCI perfusion, which suggests that presynaptic mechanisms are involved in the effects of lithium.
Effects of extraceUular calcium concentration. Intracellular calcium concentration was not affected during LiCI peffusion as indicated by the AHP and fura-2 experiments which a
b
40msec =--e control o..o LiCI 5rnM o
150j . . . .
0-, .= . . " " 0
.D
~"
100a
a" 1st
EPSP
(mY)
Figure 5. Effects of LiCI on pairedpulse facilitation. Paired stimulation at 40 msec intervals induced an increase in amplitude of the second field EPSP, as shown in the upper trace. In the lowe~ graph, the facilitation ratio, (sec¢~nd EPSP amplitude/first EPSP amplitude) *100 (%), was plotted against the amplitude of the first EP3P. A 5 mM LiC1 perfusion reduced the facilitation ratio frGm 142.5 e 2.4% to 130.0 -4-_ 10.0% (mean _+ sv, n = 6, p < 0.05; paired-~, tes0, measured at 50% of maxim~d amplitude for the first EPSP.
Effects of Lithium on Synaptic Transmission
PSYCHIATRY 1990;27:174-182
BIOL
179
=I=
60
I
I LiCI 5rnM
0. Q. w
~0'
q,.
o o
20"
!
e-
Figure 6. Effects of extracellular calcium concentration on the action of LiCI. During 5 mM LiCI peffusion, the amplitude of field EPSPs was increased by 10.9 4- 7.1% (mean _ SD, n = 7) with 1.5 mM calcium, and that was increased by 33.2 _ 17.5% (n = 7) with 2.5 mM calcium (*p < 0.01; student's-t test).
lU
O.
!
I
2.5
1.5
[ Ca2+] (mM)
are discussed in the following section. The effects of LiC1, however, may be influenced by the extracellular calcium concentration through a presynaptic mechanism. This was examined using mediums containing 1.5 mM or 2.5 mM calcium. As shown in Figure 6, the facilitating effects of LiC1 on field EPSPs were dependent on extracellular calcium concentration. The higher the extracellular calcium concentration was the more effect LiCI exerted.
Fura-2 Technique After loading fura-2 in the cultured hippucampal neurones, the resting calcium concentration was measured; in the 18 neurones examined, it was approximately 30 nM. As shown in Figure 7A, 10 mM LiCl did not change the resting fluorescence whereas 50 mM KCI increased it in the same neurone. When IO-SM glutamate was applied, fluorescence was increased to 220 _ 74% (mean _+ so, n = 18), which corresponded to about 220 nM in calcium concentration (Figure 7B). As shown in the cell in Figure 7C, 10 mM LiCI had no effect on glutamate-induced fluorescence.
Discussion Haas (1982) has already reported that lithium ions had enhancing effects on field EPSPs and population spikes in rat hippocampal slices. However, our results indicated that the amplitude of field EPSPs was increased by lithium, but that of the population spikes was va.fiable. Two types of IPSPs, GABA-IPSP and Lhp, were observed to be facilitated. Paired-pulse facilitation is thought to result from the presynaptie mechanism (Katz and Miledi 1968; Creager et al. 1980), providing an indicator of the pre~ynaptic function for transmitter release. The changes in facilitation ratios caused by lithium indicate that lithium action may be involved in the presynaptic terminals. Thus, lithium might act on both excitatory and inhibitory presynaptic terminals independent of their functions. These results, that both excitatory and inhibitory transmissions were facilitated, agree with this hypothesis. A number of investigators have sugges ed that lithium interferes with the Na-K pump.
180
Y. Higashitani et al.
BIOL PSYCHIA,':'RY 1990;27:1"/4-182
A.
4..I -PI
i1o
B. 3
I
L iCl,., lOmM
'
I
~KCI
io
'
2o
ilG IG
(/)
BG
IG
,l.J .r-I
U
10
20
'
U
C.
3~
8G
~_ LiCI
lOmM
I
IG
0,-! 21.1.-i
°o
1o
.....
20
time
Figure 7. Effects of LiCI on intracellular calcium concentration. The ordinate shows the fluorescence intensity under an excitation light of 340 nm. (A) The fluorescence at rest was not changed by 10 mM LiCl, but was increased by depolarization with 50 mM KCI. The resting calcium concentration of this neurone was 55 nM. (B) Glutamate, 10-SM, applied for 30 sec (G) induced a remarkable increase in fluorescence. The gradual decline in maximal fluorescence after successive glutamate applications probably reflected desensitization. The glutamate-induced increase in fluorescence was not influenced by l0 mM LiCI.
4"o
( m~n )
Membrane depolarization can be regarded as a result of lithium-induced inhibition of the Na-K electrogenic pump, and this probably occurred more strongly in the presynaptic terminals than in the soma of neurones, due to the volume-surface ratio. Inhibition of the Na-K pump is known to facilitate transmitter release (Brosemer 1985), and does not conflict with the findings in this electrophysiological experiment. The facilitating effect of lithium without extracellular calcium was reposed in the neuromuscular junction (Crawford 1975). In the present study, lithium action on field EPSPs was not independent of extracellular calcium concentration. Exti'acellular calcium may affect lithium action via presynaptic mechanisms. It is possible that the amplitude of population spikes may be dependent on the balance between excitatory and inhibitory postsynaptic potentials. If the facilitation of the former is overwhelming during lithium application, the amplitude of population spikes may be increased. Conversely, wl:en facilitation of the latter is dominant, the opposite result may be induced. This balance: may also be an important factor in the therapeutic action of lithium and manic-depressive illness. If lithium exerts its therapeutic action by affecting calcium-dependent processes (Allison and Stewart 1971, Hallcher and Sherman 1980; Aldenhoff and Lux 1985), some changes in intracellular calcium concentration might be observed. But the calcium concentration of the cell body at rest and its glutamate-induced increase were not influenced
Effects of Lithium on Synaptic Transmission
BIOLPSYCHIATRY 1990;27:174-182
181
by lithium when examined in cultured hippocampal neurones. These results suggest ~ a t calcium concentration in the soma of the neurone were not changed noticeably by acute expc~ure to lithium. This observation is ccns.:~tent with the results obtained from the after-hyperpolarization experiments in which calcium-activated potassium conductance was not affected by lithium. Rinaldi et al. (1986) has reported that the acute effects of lithium on population EPSP in CAI region was not consistent with chronic exposure results: the therapeutic effects of lithium require at least several days to develop. The acute effects of lithium in this study iavoIved higher concentrations (2-10 mM) than the therapeutic range ( < 2 raM), and may not account for the therapeutic effects. The most important finding in this experiment was the presynaptic involvement of lithium. There is a need for further investigation of t~ese presynaptic mechanisms. The authors are grateful to Professor S. Totsuka for continuous support and encouragement and acknowledge the skillful technical assistance of H. Miyakawa, K. Ito, and K. Kaneko.
References Aldenhoff JB, Lux HD (1985): Lithium slows neuronal calcium regulation in the snail Helix Pomatia. Neurosci Lett 54:103-108. Alger BE, Nicoll RA (1982): Feed-forward dendritic inhibition in rat hippocampal cells studied in vitro, l Physiol 328:105-123. Allison JH, Stewart MA (1971): Reduced brain inositol in lithium-treated rats. Nature New Biol 233:267-268. Banker GA, Cowan WM (1977): Rat hippocampal neurons in dispersed cell culture. Broi~ Res 126:397-425. Berridge MJ (1984) Inositol tfiphosphat, and diacylglycerol as second messengers. Biochem J 220:345-360. Brosemer RW (1985): Effect of inhibitors of Na+,K+-ATPase on *.he membr~e potentials and neurotransmitter efflux in ra', brain slices. Brain Res 344:125-137. Creager R, Dunwiddie T, Lynch G (1980): Paired-pulse and frequency facilitation in the CAI region of the in vitro rat hippocampus. J Physiol 299:409--424. Crawford AC (1975): Lithium ions and the release of transmitter at the frog neuromuscular junction. J Physiol 246:109-142. Dubovsky SL, Frank RD (1983): Intracellular calcium ions in affective disorders. Biol Psychiatry 19:781-797. Emlich HM, Aldenhoff JB, Lux JB (1982): Basic mechanisms in the action of lithium. Amsterdam: Excerpta Medica. Gryr~dewicz G, Poenie M, Tsien RY (1985): A new generation of Ca2+ indicators with greatly haproved fluorescence properties. J Biol Chem 260:3440-3450. Haas HL (1982): Lithium and synaptic transmission in the mammalian brain. In Emiich HM (ed), Basic Mechanisms in the Action of Lithium. Amsterdam: Elsevier, pp 71-79. Hallcher CM, Sherman WR (1980): The effects of lithium ion and other agents on the activity of myo-inositol one-r~osphatase from bovine brain. J Biol Chem 255:10896-10901. Hotson JR, Prince DA (1980): A calcium-activated hyperpolarization follows repetitive firing in hippocampal neurons, J Neurophysiol 43:409-419. Kaneda M, Higashitani Y, Ohtani R, Fujii S, Kato H (1986): A rotary slicer for brain research. Famagaca Med 1 4:81-85.
182
BIOLPSYCHIATRY 1990;27:174-182
Y. Higashitani et al.
Katz B, Miledi R (1968): The role of calcium in neuromuscular facilitation. J Physiol 195:481492. Kudo Y, Ogura A (1986): Glutamate-induced increase in intracellular Ca2+ concentration in isolated hippocampal neurones. Br J Pha~'macol 89:191-198. Meltzer HL (1986): Lithium mechanisms in bipolar illness and altered intracellular calcium functions. Biol Psychiatry 21:439-510. Murray JB (1985): Lithium therapy for mania and depression. J Gen Psychol 112:5-33. Newberry NR, Nicoll RA (1984): A bicuculine-resistant inhibitory post-synaptic potential in rat hippocampal pyramidal cells in vitro. J Physiol 348:239-254. Rinaldi PC, Fairchild MD, Kusske JA (1986): Perfusion with lithium modifies neurophysiological responses in the CAI region of the hippocampal slices preparation. Brain Res 375:313-319. Tsien RY, Rink TJ, Poenie M (1985): Measurement of cytosolic free Ca2+ in individual small cells using fluorescence microscopy with dual excitation wavelength. Cell Calcium 6:145-157. Ploeger EJ (1974): The effects of lithium on excitable cell membranes. On the mechanism of inhibition of the sodium pump of non-myelinated nerve fibers of the rat. Eur J Pharmacol 25:316-321. Ullrich A, Baierl P, ten Bmgge~cate G (1980): Extracellular potassium in rat cerebellar cortex during acute and chronic lithium application. Brain Res 192:287-290.
