Journal of Neurochemish-y Raven Press, Ltd., New York 0 1992 International Society for Neurochernistry
Rapid Communication
Phosphoinositide Turnover Associated with Synaptic Transmission *T. H. Murphy, *D. D. Wright, and *tJ. M. Baraban Departments of *Neuroscience and j-Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, Baltimore, Maryland, U.S.A
Abstract: Although pharmacological stimulation of a wide variety of transmitter receptors tnggers phosphoinositide (PI) turnover, little is known about the type of synaptic activity required to activate this system. To investigate this question, we have used primary cultures of embryonic cortical neurons, which develop functional glutamate and GABA synapses during maturation in vitro. Mature cultures display spontaneous synaptic activity that is totally s u p pressed by tetrodotoxin (TTX). PI turnover, assayed by the lithiumsensitive accumulation of [’H]CDPdiacylglycerol, was readily detected under basal conditions and was abolished by TTX.Increased excitatory synaptic activity induced by picrotoxin, an antagonist of GABA, receptor-mediated inhibition, further stimulated PI turnover. Similar results were obtained when PI turnover was assayed using [3H]inositol labeling. With either assay, the magnitude ofsynaptically induced PI turnover was comparable to maximal responses produced by muscarinic receptor stimulation. Although a component of the spontaneous synaptic currents is sensitive to Nmethyl-Baspartate (NMDA)-preferring glutamate receptor antagonists, blockade of NMDA receptors did not affect PI turnover associated with synaptic transmission. To assess the time course of synaptically mediated PI turnover, the amplitude and duration of spontaneous synaptic currents were reduced by lowering the extracellular Ca2+concentration from 2.25 to 0.5 mM, a maneuver that suppresses basal PI turnover. Increases in PI turnover were detected as early as 5 min following restoration of the extracellular Ca2’ concentration to 2.25 mM. Taken together, these findings indicate that activation of the PI system is associated with physiological levels of glutamatergic synaptic transmission. Key Words: Longterm potentiation-Phosphoinositide-N-Methyl-DaspartateMetabotropic receptors-Glutamate-Protein kinase C. Murphy T. H. et al. Phosphoinositide turnover associated with synaptic transmission. J . Nezrrochem. 59,2336-2339 (1992).
The activation of glutamate (Glu) receptor channels is a ubiquitous event at excitatory synapses in the CNS. Although the ionic components of fast excitatory synaptic transmission have been well described (Hestrin et al., 1990), whether concomitant activation of “slow” second messen-
ger systems, such as the phosphoinositide (PI) system (Fisher et al., 1992), also occurs is largely unknown. Using primary cultures of embryonic cortical neurons, which mature in vitro and develop spontaneous network synaptic activity (Dichter, 1978; Murphy et al., 19916), we have been able to monitor the effects of synaptic transmission on PI turnover. Previous studies have demonstrated that excitatory synaptic transmission in these cultures is mediated by both N-methyl-D-aspartate (NMDA) and non-NMDA-type Glu receptors (Murphy et al., 1991a,b). In addition, metabotropic Glu receptors linked to PI turnover have also been described in these primary cultures (Pate1et al., 1990). Neuronal activity within these cultures can be readily manipulated with conventional pharmacological agents. In particular, tetrodotoxin (TTX) totally suppresses neuronal activity, and picrotoxin, a GABA, receptor channel blocker, enhances neuronal firing (Murphy et al., 199 16, 1992a). Results obtained using these agents indicate that physiological patterns of glutamatergic synaptic activity elicit PI turnover.
MATERIALS AND METHODS Cortical cultures were made from gestational day 17 rat fetuses (Murphy et al., 1990) and kept in vitro for at least 2 1 days. Electrophysiological recordings were performed using whole-cell voltage clamp (Hamill et al., 198 1; Murphy et al., 199 1 b). [3H]CDP-diacylglycerol (CDP-DAG) levels (Godfrey, 1989; Hwang et al., 1990) were measured by prelabeling cells with 1.5 pCi/ml of [3H]cytidine (25 Ci/mmol; no. 32,206-7; Sigma Chemical Co.) in a Hanks’ balanced salt solution (HBSS) that contained 137 mM NaCl, 5.0 mM KCl, 2.5 mM CaCl,, 1 .O mM MgSO,, 0.44 mM KH2P0,, 0.34 mMNa2HP0,(7H,O), 10 M s o d i u m HEPES, 1 mh4 NaHCO,, and 5 mMglucose (pH 7.4 and 340 mOsm). Cells were then treated with agonists or agents that modify synapAbbreviations used: ACPD, IS,3R- I -aminocyclopentane- 1,3dicarboxylic acid; CDP-DAG, CDP-diacylglycerol; Glu, glutamate; HBSS, Hanks’ balanced salt solution; NMDA, N-methyl-Baspartate; PI, phosphoinositide; TTX, tetrodotoxin.
Resubmitted manuscript received September 8, 1992; accepted September 10, 1992. Address correspondence and reprint requests to Dr. T. H. Murphy at Department of Neuroscience, Johns Hopkins University School of Medicine, 725 North Wolfe Street, Baltimore, MD 2 1205-2185. U.S.A.
2336
SYNAPTIC ACTIVITY TRIGGERS PI TURNOVER tic transmission in the presence of 2.5 mM LiCl and 1.5 pCi/ml of ['Hlcytidine. Following treatment, cells were washed once with HBSS, lysed with 1 ml of ice-cold chloroform/methanol (1 :2 vol/vol), scraped off, and collected. Material from two duplicate 12-well culture wells was pooled for each condition (two culture wells contain 1 .5 mg of protein). CDP-DAG was extracted by adding 0.8 ml ofchloroform, vortex-mixing, adding 0.8 ml of water, vortex-mixing, and centrifuging at low speed to separate the phases. The organic phase (lower phase) was then collected, washed once by adding 2.0 ml of 1 M HCl/methanol ( 1:1, vol/vol), vortex-mixed, and centrifuged to separate the phases. The organic phase was evaporated to dryness in scintillation vials, fluor was added, and the amount of radioactivity corresponding to CDP-DAG was determined by liquid scintillation counting. PI turnover was also measured using ['HIinositol labeling as described (Murphy et al., 1990). To monitor PI turnover, both methods use 2-10 mM LiCl to inhibit inositol- 1-phosphate hydrolysis. In preliminary studies, we found that 10 mMLiCl inhibited spontaneous synaptic activity. However, a lower concentration of LiC1, 2.5 mM, did not affect spontaneous activity and was sufficient to allow detection of PI cycle intermediates. Accordingly, 2.5 mMLiCl was used in both assays of PI turnover.
233 7
0
- moo ; "
-
t
Plcrotorln
L
D PIC
RESULTS Mature, 2 I d a y in vitro, primary cortical cultures exhibit spontaneous synaptic transmission that can be enhanced by picrotoxin ( 1 0 &) or suppressed by TTX ( I p M ) . In the first set of experiments, we examined the effects of these treatments on PI turnover in LiC1-treated cultures that had been prelabeled with [3H]cytidine. Accumulation of radiolabeled material in the lipid phase corresponding to CDPDAG was monitored using the procedure described by Godfrey (1 989). Both control and picrotoxin-treated cultures showed marked increases in CDP-DAG content when compared with TTX-treated cultures (Fig. 1A). As expected, the increased labeling of material corresponding to CDP-DAG was totally dependent on the presence of Li+ during the incubation (data not shown). Furthermore, the response to picrotoxin was completely blocked by pretreatment with TTX, indicating that picrotoxin's stimulatory effect is mediated by its enhancement of excitatory synaptic activity (Fig. 1A). The rise in PI turnover does not appear to be mediated by NMDA receptors, as inclusion of MK-801 (3 p M ) , which blocks NMDA responses in these cultures (Murphy et al., 199 1 b), did not significantly reduce picrotoxin-induced PI turnover (Fig. 1A). Because muscarinic receptors strongly activate PI turnover in these cultures (Fig. l), we assessed whether a small number of cholinergic neurons may mediate this synaptically mediated PI response. However, atropine (1 p M ) , which blocks carbachol (1 mA4)-induced increases in PI turnover (n = 2), did not block PI turnover stimulated by picrotoxin (Fig. 1A). To confirm that the Li'-sensitive accumulation of CDPDAG reflects PI turnover, we also used ['H]inositol labeling and measured ['H]inositol- 1-phosphate accumulation. Using this method, we observed a fivefold difference between TTX- and picrotoxin-treated cultures (Fig. 1B). This difference in PI turnover associated with synaptic activity is similar in magnitude to that observed after adding the agonist carbachol to TTX-treated cultures.
FIG. 1. PI turnover associated with synaptic transmission.A Accumulation of r3H]CDP-DAG. Cultures were labeled during a 40min incubation with 1.5 pCi/ml of [3H]cytidine.and then 2.5 mM LiCl was added to all cultures. At this time, TTX (1 pM), MK-801 (3 pM). or atropine (Atrop.; 1 pM) was added in the continued presence of [3H]cytidine,for 20 min, before exposure to picrotoxin (10 pM; the four columns on the right). Control cultures (Con.) were only treated with either LiCl or LiCl and picrotoxin as indicated. After 60 min, cultures were then washed once with HBSS, and 1 ml of cold chloroform/methanol(1:2 vol/vol) was added. Data are mean SEM (bars) values of the total cpm associated with rH]CDP-DAG per culture (01.5 rng of total protein) from at least four separate experiments. Total incorporation of [3H]cytidine into washed cortical cultures was unaffected by TTX, carbachol, or picrotoxin. *p < 0.05 as determined by two-tailed t test cornparing [3H]CDP-DAG levels under control conditions with those with TTX or antagonists. The difference between the medium control (leftmost column) and the picrotoxin-treatedcontrol was also significant by the above-mentionedtest. B: lnositol phosphate levels in synaptically active and agonist-stimulated cultures. Cortical cultures were labeled overnight with 1 pCi/ml of [3H]inositol in inositol-free minimal essential medium. Cultures were then washed three times in HBSS. LiCl (2.5 mM) and 1 p M TTX (as indicated) were then added for 20 min before addition of carbachol (Carb.; 1 mM) or picrotoxin (Pic.; 10 pM). After an additional 45 min, the medium was removed, cells were washed once with HBSS. cold 3% perchloric acid was added, and inositol phosphate levels were determined as described (Murphy et al., 1990). Data are mean SEM (bars) values of the cprn associated with the inositol-1-phosphate fraction (0.2 M ammonium formate and 0.1 M formic acid were used for elution) per cortical culture from three separate experiments. None of the conditions used produced significant alterations in the total incorporation of [3H]inositol. *p < 0.05 as determined by two-tailed t test comparing inositol phosphate levels in TTX-treated cultures with the other conditions.
*
*
To investigate the time course of CDP-DAG accumulation resulting from synaptically induced PI turnover, we suppressed the amplitude and duration of spontaneous synaptic currents by lowering the extracellular Ca2+concentraJ. Neurochem., Vol. 59, No. 6, 1992
T. H. MURPHY ET AL.
2338
tion from 2.25 to 0.5 W a n d raising the extracellular Mg2+ concentration from 1 to 2 mM(Fig. 2A). Under these conditions, basal intracellular Ca2+levels were unchanged and rhythmic synaptic activity was preserved, although the amplitude and duration of Ca2+ transients were reduced (Murphy et al., 1992a). This treatment was as effective as TTX in blocking CDP-DAG accumulation but unlike TTX, which is only slowly reversible, could be readily reversed by adding an equal volume of medium containing 0 Mg2+and 4 mM CaCl,. Restoration of normal Ca2+and Mg2+levels results in a rapid (within 10s) increase in synaptic currents. This return in synaptic activity was associated with significant CDP-DAG accumulation within 5 rnin (Fig. 2B). To help define which receptors might be involved in PI turnover associated with synaptic transmission, we investigated in TTX-treated cultures the effect on PI turnover of agonists selective for Glu receptor subtypes. The metabotropic Glu receptor agonist 1S,3R- 1-aminocyclopentane- 1,3dicarboxylic acid (ACPD) (Palmer et al., 1989; Miller, 1991) produced significant stimulation of CDP-DAG accumulation. In contrast, agonists for ionotropic Glu receptors,
-
A
0.5mMCalcium
* i
TTX
5000rZOpA
B
t
d
1 0 2 0 30 4 0 S O 0 0 Tlme of Ca++ Exposure (mln)
FIG. 2. Effect on PI turnover on modulationof spontaneous synaptic activity by alteration of the extracellular Ca2+concentration. A Spontaneoussynaptic currents recorded under whole-cell voltage clamp at a holding potential of -60 mV in a picrotoxin (10 pV)-treated culture that had been switched to buffer containing 0.5 mM CaCI, and 2 mM MgSO,. The second trace was recorded 4 rnin after return to standard buffer containing 2.25 mM CaCI, and 1 mM MgSO,. This maneuver resulted in prolonged synaptic currents. Some of the large fast synaptic currents may be unclamped action potentials. Addition of TTX (1 p/W) suppresses these currents. However, at higher gain settings (20 pA calibration; lower trace) some miniature synaptic currents are still observed. B: Time course of 13H]CDP-DAG accumulation associated with augmented synaptic currents. Cortical neurons were labeled for 40 rnin with 1.5 pCi/ml of f'H]cytidine in the presence of 10 pA4 picrotoxin in HBSS containing 0.5 mM CaCI, and 2 mM MgSO,. LiCl (2.5 m/W) and TTX (as indicated) were then added. To assess the time course of CDP-DAG accumulation, Ca2' and Mg'+ concentrations were restored to normal levels by adding an equal volume of HBSS containing 4 rnM Ca" and 0 Mg2+,picrotoxin, and [3H]cytidine, which was then mixed and removed to restore the original volume. This solution was added 20. 50, 65, or 75 rnin after treatment with LiCI, and cultures were harvested after exposure for 60.30,15, or 5 min to 2.25 mM Ca2+and 1 mM Mg2+.Stimulation of PI turnover was not observed if low-Ca2+ medium or high-Ca" medium containing TTX (lTX/Ca2+) was added in this manner. Data are from a single representative time course. which was repeated twice with similar results.
J. Neurochem.. Vol. 59. No. 6. 1992
Agonist
% of control [ 'HICDP-DAG
ACPD (300 p M ) NMDA (10 f l ) Kainate (30 p M ) Carbachol(1,OOOp M )
1,020 f 370 (n = 6) 680 f 150 (n = 3) 250 -t 20 (n = 3) 2,800 + 510 (n = 6)
Cortical cultures were labeled with ['Hlcytidine as described in Fig. 1 and stimulated for 30-60 min with the indicated agonists in the presence of 1 pM TTX. Results are expressed as percentages of control ['HICDP-DAG accumulation (TTX-treated = 100%). All ofthe treatment conditions listed produced significant elevations in [3H]CDP-DAGcontent as determined by two-tailed t test. Data are mean f SEM values.
such as kainate, produced more moderate increases in PI turnover in cortical cultures (Table I), consistent with the findings of Patel et al. ( 1990).
DISCUSSION
P
4
TABLE 1. Agonist stirnulation of [3H]CDP-DAG accumulation
We report that spontaneous glutamatergic synaptic transmission is associated with robust activation ofthe PI system. Using the ['Hlcytidine labeling assay, PI turnover is readily detectable in untreated cultures that display spontaneous synaptic activity. Enhancement of excitatory synaptic activity by addition of picrotoxin increases PI turnover. The stimulation of PI turnover observed in the presence of picrotoxin is comparable in magnitude to that produced by continuous maximal stimulation of muscarinic receptors. Thus, even though synaptic activity produces only intermittent stimulation of neurons in these cultures, this pattern of synaptic stimulation is particularly effective at eliciting PI turnover. Experiments using reduced extracellular Ca2+ concentrations revealed that synaptic activity is not invariably associated with PI turnover. Under these conditions, the duration and amplitude of spontaneous synaptic currents are only partially reduced, yet PI turnover is reduced to levels found in TTX-treated cultures. This maneuver did not affect ACPDstimulated PI turnover (data not shown), indicating that this suppression is due to reduced synaptic activity rather than a direct effect of reduced Ca" concentrations on PI turnover. These findings suggest that there may be a threshold of synaptic stimulation needed to activate the PI system. As synaptic activation of the PI system has been implicated in eliciting long-term alterations in synaptic efficacy (Bashir and Collingridge, 1992), the threshold for eliciting synaptic plasticity may be determined, in part, by the requirements for inducing PI turnover. Both NMDA and amino-3-hydroxy-5-methylisoxazolepropionate (AMPA)-type Glu receptors contribute to spontaneous synaptic activity in these cultures (Murphy et al., 1991aJ). Blockade of non-NMDA Glu receptors with 6cyano-7-nitroquinoxaline-2,3-dione(Honor6 et al., 1988) resulted in a complete loss of synaptically mediated PI turnover (data not shown). However, this treatment also blocks network synaptic activity, making it difficult to assess directly the role of this receptor subtype in stimulating PI turnover. Addition of kainate, which activates these recep tors, produced only mild PI stimulation, suggesting that this receptor subtype may not account for the bulk of PI turnover associated with synaptic activity. As direct application
SYNAPTIC ACTIVITY TRIGGERS PI TURNOVER of NMDA elicits PI turnover, we also examined whether synaptic activation of this receptor mediates the PI response elicited by picrotoxin. In contrast to 6-cyano-7-nitroquinoxaline-2,3-dione, NMDA receptor antagonists block only a component of synaptic activity, leaving the pattern of spontaneous rhythmic activity intact (Murphy et al., 1991a,b, 19924. Incubation of cultures with the NMDA antagonist MK-80 1 (Wong et al., 1986) did not attenuate PI turnover associated with synaptic activity. In previous studies, we have found that the NMDA component of synaptic transmission is essential for coupling synaptic activity to expression of immediate early genes (Murphy et al., 1991a,b)and the generation of autonomous Ca*+/calmodulindependent protein kinase type I1 activity (Murphy et al., 19926) within these cultures. Thus, these signaling pathways appear to be differentially regulated and respond to distinct components of synaptic activity. The ability of ACPD to stimulate PI turnover in TTXtreated cultures suggests that activation of metabotropic receptors by Glu released during synaptic activity might contribute to the PI response associated with synaptic activity. This proposal is consistent with the observations that some synaptically mediated responses are mimicked by metabotropic Glu receptor agonists (Charpak and Gahwiler, 1991; Glaum and Miller, 1992). In initial experiments, we observed partial agonist properties with the metabotropic receptor antagonist 2-amino-3-phosphonopropionic acid (Schoepp et al., 1990), and therefore we were not able to evaluate its effects on synaptically stimulated PI turnover. Accordingly, we cannot exclude the possibility that nonglutamatergic transmitters or even growth factors released during this synaptic activity might mediate the PI response. Our finding that PI turnover is induced by synaptic activity in these cultures indicates that targets downstream of protein kinase C may be activated as well. Consistent with this prediction, we have found that microtubule-associated protein kinase, which is activated in cultures treated with phorbol esters, is also induced by synaptic stimulation (Fiore et al., 1992; Murphy et al., 19926). Thus, this culture preparation may also be useful in assessing which other targets of protein kinase C, such as neurotransmitter receptors (Huganir and Greengard, 1990), are dynamically regulated in response to synaptic activity. Acknowledgment: We thank Darla Rodgers for excellent secretarial assistance. This work was supported by an NRSA Fellowship (to T. H. Murphy) and U.S. Public Health Service grant DA-00266 and RSDA grant MH-00926 (to J. M. Baraban) and by the Lucille P. Markey Charitable Trust (to J. M. Baraban). J. M. Baraban is a Lucille P. Markey Scholar.
REFERENCES Bashir Z. I. and Collingridge G. L. (1992) Synaptic plasticity: longterm potentiation in the hippocampus. Curr. Opin. Neurobiol. 2, 328-335. Charpak S. and Gahwiler B. H. (1 99 1) Glutamate mediates a slow synaptic response in hippocampal slice cultures. Proc. R. SOC. Lond. [Biol.]243, 22 1-226. Dichter M. A. (1978) Rat cortical neurons in cell culture: culture
2339
methods, cell morphology, electrophysiology, and synapse formation. Brain Res. 149,279-293. Fiore R. S., Murphy T. H., Bayer V.,Pelech S. L., Cooper J. A., and Baraban J. M. ( I 992) Dendritic localization and synaptic regulation of MAP kinase in brain. SOC.Neurosci. Abstr. 18, 1348. Fisher S. K., Heacock A. M., and Agranoff B. W. (1992) Inositol lipids and signal transduction in the nervous system: an update. J. Neurochem. 58, 18-38. Glaum S. R. and Miller R. J. (1992) Metabotropic glutamate recep tors mediate excitatory transmission in the nucleus of the solitary tract. J. Neurosci. 12, 2251-2258. Godfrey P. P. ( 1 989) Potentiation by lithium of CMP-phosphatide formation in carbachol-stimulated rat cerebral-cortical slices and its reversal by myo-inositol. Biochem. J. 258, 621-624. Hamill 0. P., Marty A., Neher E., Sakmann B., and Sigworth F. J. (198 1 ) Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. PJugers Arch. 391, 85-100. Hestrin S., Perkel D. J., Sah P., Manabe T., Renner P., and Nicoll R. A. (1990) Physiological properties of excitatory synaptic transmission in the central nervous system. Cold Spring Harb. Symp. Quanl. Biol. 5587-94. Honor6 T., Davies S. N., Drejer J., Fletcher E. J., Jacobsen P., Lodge D., and Nielsen F. E. (1988) Quinoxalinediones: potent competitive non-NMDA glutamate receptor antagonists. Science 241,701-703. Huganir R. L. and Greengard P. ( 1990) Regulation of neurotransmitter receptor desensitization by protein phosphorylation. Neuron 5,555-567. Hwang P. M., Bredt D. S., and Snyder S. H. (1990) Autoradiographic imaging of phosphoinositide turnover in the brain. Science 249,802-804. Miller R. J. (1991) Metabotropic glutamate receptors reveal their true colors. Trends Pharmacol. Sci. 12, 365-368. Murphy T. H., Schnaar R. L., and Coyle J. T. (1990) Immature cortical neurons are uniquely sensitive to glutamate toxicity by inhibition of cystine uptake. FASEB J. 4, 1624-1633. Murphy T. H., Worley P. F., and Baraban J. M. (1991~)L-type voltage sensitive calcium channels mediate synaptic activation of immediate early genes. Neuron 7,625-635. Murphy T. H., Worley P. F., Nakabeppu Y . , Christy B., Gastel J., and Baraban J. M. (199 1b) Synaptic regulation of immediate early gene expression in primary cultures of cortical neurons. J. Neurochem. 57, 1862-1872. Murphy T. H., Blatter L. A., Wier W. G., and Baraban J. M. ( 1 9 9 2 ~Spontaneous ) synchronous synaptic calcium transients in cultured cortical neurons. J. Neurosci. 12, 4834-4845. Murphy T. H., Bhat R. V., Fiore R., and Baraban J. M. (19926) Calcium calmodulin dependent protein kinase and MAP kinase are differentially regulated by synaptic activity. Soc. Neurosci. Abstr. 18, 109 1 . Palmer E., Monaghan D. T., and Cotman C. W. (1989) transACPD, a selective agonist of the phosphoinositide-coupled excitatory amino acid receptor. Eur. J. Pharmacol. 166, 585587. Patel J., Moore W. C., Thompson C., Keith R. A,, and Salama A. I. (1990) Characterization of the quisqualate receptor linked to phosphoinositide hydrolysis in neurocortical cultures. J. New rochem. 54, 1461-1466. Schoepp D. D., Johnson B. G.,Smith E. C. R., and McQuaid L. A. (1990) Stereoselectivity and mode of inhibition of phosphoinositide-coupled excitatory amino acid receptors by 2-amino-3phosphonopropionic acid. Mol. Pharmacol. 38,222-228. Wong E. H. F., Kemp J. A., Priestly T., Knight A. R., Woodruff G. N., and Iversen L. L. (1 986) The anticonvulsant MK-80 1 is a potent N-methyl-Baspartate antagonist. Proc. Nad. Acad. Sci. USA 83,7104-7108.
J. Neurochem.. Vol. 59, No. 6, 1992
Rapid Communication
Phosphoinositide Turnover Associated with Synaptic Transmission *T. H. Murphy, *D. D. Wright, and *tJ. M. Baraban Departments of *Neuroscience and j-Psychiatry and Behavioral Sciences, Johns Hopkins University School of Medicine, Baltimore, Maryland, U.S.A
Abstract: Although pharmacological stimulation of a wide variety of transmitter receptors tnggers phosphoinositide (PI) turnover, little is known about the type of synaptic activity required to activate this system. To investigate this question, we have used primary cultures of embryonic cortical neurons, which develop functional glutamate and GABA synapses during maturation in vitro. Mature cultures display spontaneous synaptic activity that is totally s u p pressed by tetrodotoxin (TTX). PI turnover, assayed by the lithiumsensitive accumulation of [’H]CDPdiacylglycerol, was readily detected under basal conditions and was abolished by TTX.Increased excitatory synaptic activity induced by picrotoxin, an antagonist of GABA, receptor-mediated inhibition, further stimulated PI turnover. Similar results were obtained when PI turnover was assayed using [3H]inositol labeling. With either assay, the magnitude ofsynaptically induced PI turnover was comparable to maximal responses produced by muscarinic receptor stimulation. Although a component of the spontaneous synaptic currents is sensitive to Nmethyl-Baspartate (NMDA)-preferring glutamate receptor antagonists, blockade of NMDA receptors did not affect PI turnover associated with synaptic transmission. To assess the time course of synaptically mediated PI turnover, the amplitude and duration of spontaneous synaptic currents were reduced by lowering the extracellular Ca2+concentration from 2.25 to 0.5 mM, a maneuver that suppresses basal PI turnover. Increases in PI turnover were detected as early as 5 min following restoration of the extracellular Ca2’ concentration to 2.25 mM. Taken together, these findings indicate that activation of the PI system is associated with physiological levels of glutamatergic synaptic transmission. Key Words: Longterm potentiation-Phosphoinositide-N-Methyl-DaspartateMetabotropic receptors-Glutamate-Protein kinase C. Murphy T. H. et al. Phosphoinositide turnover associated with synaptic transmission. J . Nezrrochem. 59,2336-2339 (1992).
The activation of glutamate (Glu) receptor channels is a ubiquitous event at excitatory synapses in the CNS. Although the ionic components of fast excitatory synaptic transmission have been well described (Hestrin et al., 1990), whether concomitant activation of “slow” second messen-
ger systems, such as the phosphoinositide (PI) system (Fisher et al., 1992), also occurs is largely unknown. Using primary cultures of embryonic cortical neurons, which mature in vitro and develop spontaneous network synaptic activity (Dichter, 1978; Murphy et al., 19916), we have been able to monitor the effects of synaptic transmission on PI turnover. Previous studies have demonstrated that excitatory synaptic transmission in these cultures is mediated by both N-methyl-D-aspartate (NMDA) and non-NMDA-type Glu receptors (Murphy et al., 1991a,b). In addition, metabotropic Glu receptors linked to PI turnover have also been described in these primary cultures (Pate1et al., 1990). Neuronal activity within these cultures can be readily manipulated with conventional pharmacological agents. In particular, tetrodotoxin (TTX) totally suppresses neuronal activity, and picrotoxin, a GABA, receptor channel blocker, enhances neuronal firing (Murphy et al., 199 16, 1992a). Results obtained using these agents indicate that physiological patterns of glutamatergic synaptic activity elicit PI turnover.
MATERIALS AND METHODS Cortical cultures were made from gestational day 17 rat fetuses (Murphy et al., 1990) and kept in vitro for at least 2 1 days. Electrophysiological recordings were performed using whole-cell voltage clamp (Hamill et al., 198 1; Murphy et al., 199 1 b). [3H]CDP-diacylglycerol (CDP-DAG) levels (Godfrey, 1989; Hwang et al., 1990) were measured by prelabeling cells with 1.5 pCi/ml of [3H]cytidine (25 Ci/mmol; no. 32,206-7; Sigma Chemical Co.) in a Hanks’ balanced salt solution (HBSS) that contained 137 mM NaCl, 5.0 mM KCl, 2.5 mM CaCl,, 1 .O mM MgSO,, 0.44 mM KH2P0,, 0.34 mMNa2HP0,(7H,O), 10 M s o d i u m HEPES, 1 mh4 NaHCO,, and 5 mMglucose (pH 7.4 and 340 mOsm). Cells were then treated with agonists or agents that modify synapAbbreviations used: ACPD, IS,3R- I -aminocyclopentane- 1,3dicarboxylic acid; CDP-DAG, CDP-diacylglycerol; Glu, glutamate; HBSS, Hanks’ balanced salt solution; NMDA, N-methyl-Baspartate; PI, phosphoinositide; TTX, tetrodotoxin.
Resubmitted manuscript received September 8, 1992; accepted September 10, 1992. Address correspondence and reprint requests to Dr. T. H. Murphy at Department of Neuroscience, Johns Hopkins University School of Medicine, 725 North Wolfe Street, Baltimore, MD 2 1205-2185. U.S.A.
2336
SYNAPTIC ACTIVITY TRIGGERS PI TURNOVER tic transmission in the presence of 2.5 mM LiCl and 1.5 pCi/ml of ['Hlcytidine. Following treatment, cells were washed once with HBSS, lysed with 1 ml of ice-cold chloroform/methanol (1 :2 vol/vol), scraped off, and collected. Material from two duplicate 12-well culture wells was pooled for each condition (two culture wells contain 1 .5 mg of protein). CDP-DAG was extracted by adding 0.8 ml ofchloroform, vortex-mixing, adding 0.8 ml of water, vortex-mixing, and centrifuging at low speed to separate the phases. The organic phase (lower phase) was then collected, washed once by adding 2.0 ml of 1 M HCl/methanol ( 1:1, vol/vol), vortex-mixed, and centrifuged to separate the phases. The organic phase was evaporated to dryness in scintillation vials, fluor was added, and the amount of radioactivity corresponding to CDP-DAG was determined by liquid scintillation counting. PI turnover was also measured using ['HIinositol labeling as described (Murphy et al., 1990). To monitor PI turnover, both methods use 2-10 mM LiCl to inhibit inositol- 1-phosphate hydrolysis. In preliminary studies, we found that 10 mMLiCl inhibited spontaneous synaptic activity. However, a lower concentration of LiC1, 2.5 mM, did not affect spontaneous activity and was sufficient to allow detection of PI cycle intermediates. Accordingly, 2.5 mMLiCl was used in both assays of PI turnover.
233 7
0
- moo ; "
-
t
Plcrotorln
L
D PIC
RESULTS Mature, 2 I d a y in vitro, primary cortical cultures exhibit spontaneous synaptic transmission that can be enhanced by picrotoxin ( 1 0 &) or suppressed by TTX ( I p M ) . In the first set of experiments, we examined the effects of these treatments on PI turnover in LiC1-treated cultures that had been prelabeled with [3H]cytidine. Accumulation of radiolabeled material in the lipid phase corresponding to CDPDAG was monitored using the procedure described by Godfrey (1 989). Both control and picrotoxin-treated cultures showed marked increases in CDP-DAG content when compared with TTX-treated cultures (Fig. 1A). As expected, the increased labeling of material corresponding to CDP-DAG was totally dependent on the presence of Li+ during the incubation (data not shown). Furthermore, the response to picrotoxin was completely blocked by pretreatment with TTX, indicating that picrotoxin's stimulatory effect is mediated by its enhancement of excitatory synaptic activity (Fig. 1A). The rise in PI turnover does not appear to be mediated by NMDA receptors, as inclusion of MK-801 (3 p M ) , which blocks NMDA responses in these cultures (Murphy et al., 199 1 b), did not significantly reduce picrotoxin-induced PI turnover (Fig. 1A). Because muscarinic receptors strongly activate PI turnover in these cultures (Fig. l), we assessed whether a small number of cholinergic neurons may mediate this synaptically mediated PI response. However, atropine (1 p M ) , which blocks carbachol (1 mA4)-induced increases in PI turnover (n = 2), did not block PI turnover stimulated by picrotoxin (Fig. 1A). To confirm that the Li'-sensitive accumulation of CDPDAG reflects PI turnover, we also used ['H]inositol labeling and measured ['H]inositol- 1-phosphate accumulation. Using this method, we observed a fivefold difference between TTX- and picrotoxin-treated cultures (Fig. 1B). This difference in PI turnover associated with synaptic activity is similar in magnitude to that observed after adding the agonist carbachol to TTX-treated cultures.
FIG. 1. PI turnover associated with synaptic transmission.A Accumulation of r3H]CDP-DAG. Cultures were labeled during a 40min incubation with 1.5 pCi/ml of [3H]cytidine.and then 2.5 mM LiCl was added to all cultures. At this time, TTX (1 pM), MK-801 (3 pM). or atropine (Atrop.; 1 pM) was added in the continued presence of [3H]cytidine,for 20 min, before exposure to picrotoxin (10 pM; the four columns on the right). Control cultures (Con.) were only treated with either LiCl or LiCl and picrotoxin as indicated. After 60 min, cultures were then washed once with HBSS, and 1 ml of cold chloroform/methanol(1:2 vol/vol) was added. Data are mean SEM (bars) values of the total cpm associated with rH]CDP-DAG per culture (01.5 rng of total protein) from at least four separate experiments. Total incorporation of [3H]cytidine into washed cortical cultures was unaffected by TTX, carbachol, or picrotoxin. *p < 0.05 as determined by two-tailed t test cornparing [3H]CDP-DAG levels under control conditions with those with TTX or antagonists. The difference between the medium control (leftmost column) and the picrotoxin-treatedcontrol was also significant by the above-mentionedtest. B: lnositol phosphate levels in synaptically active and agonist-stimulated cultures. Cortical cultures were labeled overnight with 1 pCi/ml of [3H]inositol in inositol-free minimal essential medium. Cultures were then washed three times in HBSS. LiCl (2.5 mM) and 1 p M TTX (as indicated) were then added for 20 min before addition of carbachol (Carb.; 1 mM) or picrotoxin (Pic.; 10 pM). After an additional 45 min, the medium was removed, cells were washed once with HBSS. cold 3% perchloric acid was added, and inositol phosphate levels were determined as described (Murphy et al., 1990). Data are mean SEM (bars) values of the cprn associated with the inositol-1-phosphate fraction (0.2 M ammonium formate and 0.1 M formic acid were used for elution) per cortical culture from three separate experiments. None of the conditions used produced significant alterations in the total incorporation of [3H]inositol. *p < 0.05 as determined by two-tailed t test comparing inositol phosphate levels in TTX-treated cultures with the other conditions.
*
*
To investigate the time course of CDP-DAG accumulation resulting from synaptically induced PI turnover, we suppressed the amplitude and duration of spontaneous synaptic currents by lowering the extracellular Ca2+concentraJ. Neurochem., Vol. 59, No. 6, 1992
T. H. MURPHY ET AL.
2338
tion from 2.25 to 0.5 W a n d raising the extracellular Mg2+ concentration from 1 to 2 mM(Fig. 2A). Under these conditions, basal intracellular Ca2+levels were unchanged and rhythmic synaptic activity was preserved, although the amplitude and duration of Ca2+ transients were reduced (Murphy et al., 1992a). This treatment was as effective as TTX in blocking CDP-DAG accumulation but unlike TTX, which is only slowly reversible, could be readily reversed by adding an equal volume of medium containing 0 Mg2+and 4 mM CaCl,. Restoration of normal Ca2+and Mg2+levels results in a rapid (within 10s) increase in synaptic currents. This return in synaptic activity was associated with significant CDP-DAG accumulation within 5 rnin (Fig. 2B). To help define which receptors might be involved in PI turnover associated with synaptic transmission, we investigated in TTX-treated cultures the effect on PI turnover of agonists selective for Glu receptor subtypes. The metabotropic Glu receptor agonist 1S,3R- 1-aminocyclopentane- 1,3dicarboxylic acid (ACPD) (Palmer et al., 1989; Miller, 1991) produced significant stimulation of CDP-DAG accumulation. In contrast, agonists for ionotropic Glu receptors,
-
A
0.5mMCalcium
* i
TTX
5000rZOpA
B
t
d
1 0 2 0 30 4 0 S O 0 0 Tlme of Ca++ Exposure (mln)
FIG. 2. Effect on PI turnover on modulationof spontaneous synaptic activity by alteration of the extracellular Ca2+concentration. A Spontaneoussynaptic currents recorded under whole-cell voltage clamp at a holding potential of -60 mV in a picrotoxin (10 pV)-treated culture that had been switched to buffer containing 0.5 mM CaCI, and 2 mM MgSO,. The second trace was recorded 4 rnin after return to standard buffer containing 2.25 mM CaCI, and 1 mM MgSO,. This maneuver resulted in prolonged synaptic currents. Some of the large fast synaptic currents may be unclamped action potentials. Addition of TTX (1 p/W) suppresses these currents. However, at higher gain settings (20 pA calibration; lower trace) some miniature synaptic currents are still observed. B: Time course of 13H]CDP-DAG accumulation associated with augmented synaptic currents. Cortical neurons were labeled for 40 rnin with 1.5 pCi/ml of f'H]cytidine in the presence of 10 pA4 picrotoxin in HBSS containing 0.5 mM CaCI, and 2 mM MgSO,. LiCl (2.5 m/W) and TTX (as indicated) were then added. To assess the time course of CDP-DAG accumulation, Ca2' and Mg'+ concentrations were restored to normal levels by adding an equal volume of HBSS containing 4 rnM Ca" and 0 Mg2+,picrotoxin, and [3H]cytidine, which was then mixed and removed to restore the original volume. This solution was added 20. 50, 65, or 75 rnin after treatment with LiCI, and cultures were harvested after exposure for 60.30,15, or 5 min to 2.25 mM Ca2+and 1 mM Mg2+.Stimulation of PI turnover was not observed if low-Ca2+ medium or high-Ca" medium containing TTX (lTX/Ca2+) was added in this manner. Data are from a single representative time course. which was repeated twice with similar results.
J. Neurochem.. Vol. 59. No. 6. 1992
Agonist
% of control [ 'HICDP-DAG
ACPD (300 p M ) NMDA (10 f l ) Kainate (30 p M ) Carbachol(1,OOOp M )
1,020 f 370 (n = 6) 680 f 150 (n = 3) 250 -t 20 (n = 3) 2,800 + 510 (n = 6)
Cortical cultures were labeled with ['Hlcytidine as described in Fig. 1 and stimulated for 30-60 min with the indicated agonists in the presence of 1 pM TTX. Results are expressed as percentages of control ['HICDP-DAG accumulation (TTX-treated = 100%). All ofthe treatment conditions listed produced significant elevations in [3H]CDP-DAGcontent as determined by two-tailed t test. Data are mean f SEM values.
such as kainate, produced more moderate increases in PI turnover in cortical cultures (Table I), consistent with the findings of Patel et al. ( 1990).
DISCUSSION
P
4
TABLE 1. Agonist stirnulation of [3H]CDP-DAG accumulation
We report that spontaneous glutamatergic synaptic transmission is associated with robust activation ofthe PI system. Using the ['Hlcytidine labeling assay, PI turnover is readily detectable in untreated cultures that display spontaneous synaptic activity. Enhancement of excitatory synaptic activity by addition of picrotoxin increases PI turnover. The stimulation of PI turnover observed in the presence of picrotoxin is comparable in magnitude to that produced by continuous maximal stimulation of muscarinic receptors. Thus, even though synaptic activity produces only intermittent stimulation of neurons in these cultures, this pattern of synaptic stimulation is particularly effective at eliciting PI turnover. Experiments using reduced extracellular Ca2+ concentrations revealed that synaptic activity is not invariably associated with PI turnover. Under these conditions, the duration and amplitude of spontaneous synaptic currents are only partially reduced, yet PI turnover is reduced to levels found in TTX-treated cultures. This maneuver did not affect ACPDstimulated PI turnover (data not shown), indicating that this suppression is due to reduced synaptic activity rather than a direct effect of reduced Ca" concentrations on PI turnover. These findings suggest that there may be a threshold of synaptic stimulation needed to activate the PI system. As synaptic activation of the PI system has been implicated in eliciting long-term alterations in synaptic efficacy (Bashir and Collingridge, 1992), the threshold for eliciting synaptic plasticity may be determined, in part, by the requirements for inducing PI turnover. Both NMDA and amino-3-hydroxy-5-methylisoxazolepropionate (AMPA)-type Glu receptors contribute to spontaneous synaptic activity in these cultures (Murphy et al., 1991aJ). Blockade of non-NMDA Glu receptors with 6cyano-7-nitroquinoxaline-2,3-dione(Honor6 et al., 1988) resulted in a complete loss of synaptically mediated PI turnover (data not shown). However, this treatment also blocks network synaptic activity, making it difficult to assess directly the role of this receptor subtype in stimulating PI turnover. Addition of kainate, which activates these recep tors, produced only mild PI stimulation, suggesting that this receptor subtype may not account for the bulk of PI turnover associated with synaptic activity. As direct application
SYNAPTIC ACTIVITY TRIGGERS PI TURNOVER of NMDA elicits PI turnover, we also examined whether synaptic activation of this receptor mediates the PI response elicited by picrotoxin. In contrast to 6-cyano-7-nitroquinoxaline-2,3-dione, NMDA receptor antagonists block only a component of synaptic activity, leaving the pattern of spontaneous rhythmic activity intact (Murphy et al., 1991a,b, 19924. Incubation of cultures with the NMDA antagonist MK-80 1 (Wong et al., 1986) did not attenuate PI turnover associated with synaptic activity. In previous studies, we have found that the NMDA component of synaptic transmission is essential for coupling synaptic activity to expression of immediate early genes (Murphy et al., 1991a,b)and the generation of autonomous Ca*+/calmodulindependent protein kinase type I1 activity (Murphy et al., 19926) within these cultures. Thus, these signaling pathways appear to be differentially regulated and respond to distinct components of synaptic activity. The ability of ACPD to stimulate PI turnover in TTXtreated cultures suggests that activation of metabotropic receptors by Glu released during synaptic activity might contribute to the PI response associated with synaptic activity. This proposal is consistent with the observations that some synaptically mediated responses are mimicked by metabotropic Glu receptor agonists (Charpak and Gahwiler, 1991; Glaum and Miller, 1992). In initial experiments, we observed partial agonist properties with the metabotropic receptor antagonist 2-amino-3-phosphonopropionic acid (Schoepp et al., 1990), and therefore we were not able to evaluate its effects on synaptically stimulated PI turnover. Accordingly, we cannot exclude the possibility that nonglutamatergic transmitters or even growth factors released during this synaptic activity might mediate the PI response. Our finding that PI turnover is induced by synaptic activity in these cultures indicates that targets downstream of protein kinase C may be activated as well. Consistent with this prediction, we have found that microtubule-associated protein kinase, which is activated in cultures treated with phorbol esters, is also induced by synaptic stimulation (Fiore et al., 1992; Murphy et al., 19926). Thus, this culture preparation may also be useful in assessing which other targets of protein kinase C, such as neurotransmitter receptors (Huganir and Greengard, 1990), are dynamically regulated in response to synaptic activity. Acknowledgment: We thank Darla Rodgers for excellent secretarial assistance. This work was supported by an NRSA Fellowship (to T. H. Murphy) and U.S. Public Health Service grant DA-00266 and RSDA grant MH-00926 (to J. M. Baraban) and by the Lucille P. Markey Charitable Trust (to J. M. Baraban). J. M. Baraban is a Lucille P. Markey Scholar.
REFERENCES Bashir Z. I. and Collingridge G. L. (1992) Synaptic plasticity: longterm potentiation in the hippocampus. Curr. Opin. Neurobiol. 2, 328-335. Charpak S. and Gahwiler B. H. (1 99 1) Glutamate mediates a slow synaptic response in hippocampal slice cultures. Proc. R. SOC. Lond. [Biol.]243, 22 1-226. Dichter M. A. (1978) Rat cortical neurons in cell culture: culture
2339
methods, cell morphology, electrophysiology, and synapse formation. Brain Res. 149,279-293. Fiore R. S., Murphy T. H., Bayer V.,Pelech S. L., Cooper J. A., and Baraban J. M. ( I 992) Dendritic localization and synaptic regulation of MAP kinase in brain. SOC.Neurosci. Abstr. 18, 1348. Fisher S. K., Heacock A. M., and Agranoff B. W. (1992) Inositol lipids and signal transduction in the nervous system: an update. J. Neurochem. 58, 18-38. Glaum S. R. and Miller R. J. (1992) Metabotropic glutamate recep tors mediate excitatory transmission in the nucleus of the solitary tract. J. Neurosci. 12, 2251-2258. Godfrey P. P. ( 1 989) Potentiation by lithium of CMP-phosphatide formation in carbachol-stimulated rat cerebral-cortical slices and its reversal by myo-inositol. Biochem. J. 258, 621-624. Hamill 0. P., Marty A., Neher E., Sakmann B., and Sigworth F. J. (198 1 ) Improved patch-clamp techniques for high-resolution current recording from cells and cell-free membrane patches. PJugers Arch. 391, 85-100. Hestrin S., Perkel D. J., Sah P., Manabe T., Renner P., and Nicoll R. A. (1990) Physiological properties of excitatory synaptic transmission in the central nervous system. Cold Spring Harb. Symp. Quanl. Biol. 5587-94. Honor6 T., Davies S. N., Drejer J., Fletcher E. J., Jacobsen P., Lodge D., and Nielsen F. E. (1988) Quinoxalinediones: potent competitive non-NMDA glutamate receptor antagonists. Science 241,701-703. Huganir R. L. and Greengard P. ( 1990) Regulation of neurotransmitter receptor desensitization by protein phosphorylation. Neuron 5,555-567. Hwang P. M., Bredt D. S., and Snyder S. H. (1990) Autoradiographic imaging of phosphoinositide turnover in the brain. Science 249,802-804. Miller R. J. (1991) Metabotropic glutamate receptors reveal their true colors. Trends Pharmacol. Sci. 12, 365-368. Murphy T. H., Schnaar R. L., and Coyle J. T. (1990) Immature cortical neurons are uniquely sensitive to glutamate toxicity by inhibition of cystine uptake. FASEB J. 4, 1624-1633. Murphy T. H., Worley P. F., and Baraban J. M. (1991~)L-type voltage sensitive calcium channels mediate synaptic activation of immediate early genes. Neuron 7,625-635. Murphy T. H., Worley P. F., Nakabeppu Y . , Christy B., Gastel J., and Baraban J. M. (199 1b) Synaptic regulation of immediate early gene expression in primary cultures of cortical neurons. J. Neurochem. 57, 1862-1872. Murphy T. H., Blatter L. A., Wier W. G., and Baraban J. M. ( 1 9 9 2 ~Spontaneous ) synchronous synaptic calcium transients in cultured cortical neurons. J. Neurosci. 12, 4834-4845. Murphy T. H., Bhat R. V., Fiore R., and Baraban J. M. (19926) Calcium calmodulin dependent protein kinase and MAP kinase are differentially regulated by synaptic activity. Soc. Neurosci. Abstr. 18, 109 1 . Palmer E., Monaghan D. T., and Cotman C. W. (1989) transACPD, a selective agonist of the phosphoinositide-coupled excitatory amino acid receptor. Eur. J. Pharmacol. 166, 585587. Patel J., Moore W. C., Thompson C., Keith R. A,, and Salama A. I. (1990) Characterization of the quisqualate receptor linked to phosphoinositide hydrolysis in neurocortical cultures. J. New rochem. 54, 1461-1466. Schoepp D. D., Johnson B. G.,Smith E. C. R., and McQuaid L. A. (1990) Stereoselectivity and mode of inhibition of phosphoinositide-coupled excitatory amino acid receptors by 2-amino-3phosphonopropionic acid. Mol. Pharmacol. 38,222-228. Wong E. H. F., Kemp J. A., Priestly T., Knight A. R., Woodruff G. N., and Iversen L. L. (1 986) The anticonvulsant MK-80 1 is a potent N-methyl-Baspartate antagonist. Proc. Nad. Acad. Sci. USA 83,7104-7108.
J. Neurochem.. Vol. 59, No. 6, 1992
