Neulvchemical Research, VoL 15, No. 7, 1990, pp. 743-750

Presynaptic Facilitation of Glutamate Release from Isolated Hippocampal Mossy Fiber Nerve Endings by Arachidonic Acid Ernest J. Freeman 1, David M. Tertian 2, and Robert V. Dorman 3 (Accepted April 9, 1990)

Hippocampal mossy fiber synaptosomes were used to investigate the role of arachidonic acid in the release of endogenous glutamate and the long-lasting facilitation of glutamate release associated with long-term potentiation. Exogenous arachidonate induced a dose-dependent efflux of glutamate from the hippocampal mossy fiber synaptosomes and this effect was mimicked by melittin. Neither treatment induced the release of occluded lactate dehydrogenase at the concentrations used in these experiments. In each case, removal of the biochemical stimulus allowed for glutamate efflux to return to spontaneous levels. However, there was a persistent effect of exposure to either arachidonate or melittin, since these compounds facilitated the glutamate release induced by the subsequent addition of 35 mM KCt. This facilitation of glutamate release resulted from an enhancement of both the magnitude and duration of the response to depolarization. Although exogenous prostanoids were also able to stimulate the release of glutamate, they appeared to play no direct role in secretion processes, since inhibition of eicosanoid synthesis potentiated the glutamate efflux in response to membrane depolarization or exogenous arachidonic acid. We suggest that the calciumdependent accumulation of arachidonic acid in presynaptic membranes plays a central role in the release of endogenous glutamate and that the persistent effects of arachidonic acid may be related to the maintenance of long-term potentiation in the hippocampal mossy fiber-CA3 synapse.

KEY WORDS: Arachidonic acid; glutamate release; hippocampus; presynaptic facilitation; mossy fiber synaptosomes; phospholipase.

INTRODUCTION

transmitters from synaptosomes (2-5). We previously showed that the presynaptic accumulation of arachidonate is attenuated by calcium entry blockers (6) and it is known that the arachidonate-induced release of neurotransmitters is calcium-independent (2-4). On the basis of these findings, we proposed that the opening of presynaptic, voltage-sensitive calcium channels may play a role in the lipase-dependent liberation of arachidonate from membrane phospholipids (3). The resultant accumulation of unesterified arachidonate may then act to facilitate the stimulus-secretion process. It has been suggested that the induction of longterm potentiation (LTP) in the dentate gyrus is associated with the release of arachidonic acid from postsynaptic

Depolarization of the presynaptic plasma membrane stimulates the release of neurotransmitters from isolated nerve endings and the liberation of unesterified arachidonic acid (1-3). These biochemical responses to membrane depolarization may be functionally related, since exogenous arachidonate induces the release of neuroDepartment of Biological Sciences, Kent State University, Kent, OH 44242, U.S.A. 2 Department of Anatomy and Cell Biology, School of Medicine, East Carolina University, Greenville, NC 27858 3 Correspondence: Robert V. Dorman, Ph.D., Department of Biological Sciences, Kent State University, Kent, OH 44242, U.S.A.

743 0364-3190/90/0700-0743506.00/09 1990PlenumPublishingCorporation

744 phospholipids and that this arachidonate m a y then function as a retrograde messenger to enhance the release of glutamate and ensure the m a i n t e n a n c e of LTP (7,8). A similar role for lipoxygenase products of arachidonate m e t a b o l i s m has b e e n established in Aplysia (9,10). However, it is also possible that the Ca2+-dependent release of arachidonate from presynaptic phospholipids might act locally to increase glutamate release, since calcium entry into presynaptic terminals is increased following the induction of LTP (11). This suggestion is supported b y the observations that exogenous arachidonate stimulates neurotransmitter release from hippocampal slices (8), as well as from isolated nerve endings (2-5). In the present study, we report that the activation of e n d o g e n o u s phospholipase A2 and the addition of unesterified arachidonic acid stimulated the release of e n d o g e n o u s glutamate from hippocampal m o s s y fiber synaptosomes. These treatments enhanced glutamate release upon subsequent exposure of the nerve endings to depolarizing conditions. In addition, both the depolarization- and arachidonate-induced release of glutamate were potentiated w h e n the conversion of arachidonate to eicosanoids was inhibited. These results are consistent with a direct role for arachidonic acid in the evoked release of glutamate, as well as the suggestion that the liberation of arachidonate from presynaptic phosphoIipids m a y be i n v o l v e d in the m e c h a n i s m s of LTP in the m o s s y fiber-CA3 synapse.

EXPERIMENTAL PROCEDURE Hippocampal Mossy Fiber Synaptosomal Preparation. Mossy fiber terminals were isolated from hippocampi dissected from SpragueDawley rats. The homogenization and centrifugation procedures have been described in detail (12). The final pellets (P3) were resuspended in a calcium-free Krebs-bicarbonate medium and preincubated for 5 rain at 30~ The Warburg-Christian method (t3) was routinely used to estimate the protein concentration of the freshly prepared mossy fiber synaptosomal suspension. Fractions of each preparation were also taken for subsequent protein measurements according to Lowry et al. (14). Aliquots of the mossy fiber synaptosomal suspension (2 mg protein) were applied to water-jacketed superfusion columns, fitted with glass fiber filters (Whatman GFB), that had been equilibrated at 30~ with control medium (final concentrations in raM: 127 NaC1, 3.9 KC1, 1.2 MgS04, 1.8 CaCI2, 1.8 KH2PQ, 20 NaHCO3, 11 D-glucose, 15 sodium N-tris[hydroxymethyl]-methyl-2-aminoethanesulfonicacid (NaTES); pH 7.4). Release of Endogenous Glutamate. Preincubated synaptosomes were maintained at 30~ in parallel chambers and superfused at a rate of 0.5 ml/min with a Krebs-bicarbonate buffer containing 1.8 mM CaClz (12). The synaptosomes were superfused for 16 rain with control buffer, in order to allow spontaneous glutamate effiux to reach a steady baseline level. Nerve terminals were depolarized with a 2 rain pulse of 35 mM KCI and the cyclooxygenase and lipoxygenase inhibitors were added 10 rain prior to depolarization. Two rain exposures were

Freeman, Tertian, and Dorman also used for 4-aminopyridine (4-AP), melittin, purified bee venom phospholipase A2 and Triton X-100, while exogenous fatty acids and eicosanoids were present for 10 min, in order to allow sufficient time for absorption by the synaptosomes. All treatment compounds were purchased from Sigma Chemical Co. The fatty acids and inhibitors of cyclooxygenase and lipoxygenase were solubilized in dimethyl sulfoxide (DMSO; M melittin, respectively. These LDH activities were not significantly different from that obtained from unstimulated synaptosomes (15.9 + 2.3 txmol/h/mg protein). Treatment with 10 ~M melittin caused a significant decrease in the occluded LDH (11.8 + 2.1 ~mol/h/rng protein) to 74% of that observed in unstimulated controls. By comparison, a 2 min pulse of 0.4% Triton X-100 caused the release of 89% of the intrasynaptosomal LDH, reducing the occluded activity to 1.8 + 0.5 txmol/h/mg protein.

32

56

Table I. Effects of Fatty Acids and Eicosanoids on the Spontaneous Release of Glutamate from Resting, Polarized Hippocampal Mossy Fiber Synaptosomes in

4-0

44

Evoked Glutamate Release (pmol/min/mg protein)

48

SUPERFUSION TIME (rain)

Fig. 5. Effects of melittin on the depolarization-evoked release of endogenous glutamate from hippocampal mossy fiber synaptosomes. Superfused synaptosomes were exposed to a 2 rain pulse with 1.0 IxM melittin (MEL), followed by a 14 min exposure to control buffer. A 2 rain pulse of 35 mM KC1 (K +) was applied at the time shown and parallel superfusion chambers received only the K+ stimulus. Released glutamate was quantitated as described. The results were obtained from 7 separate experiments and are shown as pmol glutamate released/rain/ mg protein • SEM. (0) = unstimulated controls; (li,) = 35 mM KC1 only; ( I ) = melittin followed by 35 mM KCI; * indicates significantly different from unstimulated controls; x indicates significantly different from 35 mM KCI only; P < 0.05; Newman-Keuls test.

reversible (Figure 5). There was no change in the release of glutamate relative to spontaneous efflux when 0.6

Fatty Acids (200 txM) Arachidonic Acid Linoleic Acid Oleic Acid Palmitic Acid

59 • 6* 29 __ 8* NS NS

Eicosanoids (100 ~M) PGE 2 PGF2~ 12-HETE

38 _+ 6* 18 + 6* NS i

i

i

Superfused hippocampal mossy fiber synaptosomes were exposed to free fatty acids or eicosanoids at the concentrations shown. Their effects on glutamate efflux were determined as described. The results were obtained from at least four separate experiments and are expressed as pmol glutamate released/min/mg protein - SEM. * indicates significantly different from spontaneous efflux; _P < 0.05; NewmanKeuls test; NS indicates not sigificantly different from spontaneous efflux.

Arachidonic Acid and Glutamate Release

efflux are also shown in Table I. A 10 rain exposure to prostaglandins E2 (PGE2; 100 jxM) and F~ (PGF2~ ; 100 ~M) induced glutamate effluxes that were 27 and 12% above spontaneous levels, respectively. The lipoxygenase product 12-HETE (100 ~M) did not stimulate the release of glutamate from the mossy fiber terminals. Inhibition of eicosanoid synthesis affected the K +and arachidonate-induced release of glutamate from the mossy fiber synaptosomes (Figure 3), but had no effect on the spontaneous release of glutamate (unpublished results). Depolarization of the terminals with 35 mM KC1 increased glutamate release by 255 _+ 6 pmol/min/ mg protein. The response to K +-induced depolarization was increased 73% in the presence of indomethacin (10 ~,M). Pretreatment with 10 ~M ibuprofen potentiated the K+-evoked release by 35%, while the lipoxygenase inhibitor nordihydroguaiaretic acid (NDGA, 10 ~M) increased the K+-evoked response by 83%. The glutamate release evoked by 200 txM arachidonic acid (59 _ 6 pmol/min/mg protein) was also potentiated by 10 ~M ibuprofen and 10 ~M NDGA. Ibuprofen caused a 33% increase in evoked release when compared to that in the presence of arachidonate alone, while efflux was increased 113% in the presence of NDGA. The addition of 10 ~M indomethacin, however, had no effect on arachidonic acid-stimulated glutamate release. Effects of Arachidonate and Melittin on Glutamate Release Induced by Membrane Depolarization. The effects of 35 mM KCl-induced membrane depolarization on glutamate release were determined following the addition and wash-out of either arachidonate or melittin. The glutamate efflux induced by 200 ~M arachidonate returned to the control level of spontaneous efflux following removal of the free fatty acid from the superfusion buffer. A subsequent 2 rain stimulation of these synaptosomes with 35 mM KC1 resulted in a significantly prolonged release of glutamate, when compared to the K+-evoked release from synaptosomes that were not previously treated with arachidonate (Figure 4). Exposure of the synaptosomes to 1.0 ~M melittin also caused a significant enhancement of the subsequent K+-stimulated glutamate release when compared to previously unstimulated synaptosomes (Figure 5). Again, wash-out of the stimulus allowed for glutamate release to return to spontaneous levels, while a subsequent exposure of the synaptosomes to 35 mM KC1 resulted in a 20% increase in peak release, when compared to the K+-induced release from synaptosomes that had not been pretreated with melittin. Melittin pretreatment also increased the duration of the release profile in response to K+-induced depolarization.

747

DISCUSSION Mossy fiber nerve endings were isolated from rat hippocampus, in order to assess the relationships between arachidonic acid metabolism and glutamate secretion. The enhanced release of glutamate (17,18) and the activity dependent mobilization of unesterified arachidonic acid (7,8,19) have been implicated in the expression of hippocampal long-term potentiation. It has been established that LTP can be elicited by repetitive stimulation of the mossy fibers (20-22) and glutamate is thought to mediate the excitatory, mossy fiber synaptic input (23). In addition, glutamate is the amino acid transmitter that is selectively released from the isolated nerve endings in response to depolarization (24). The relative homogeneity of this synaptosomal fraction is indicated by the inability of sustained depolarization to evoke the release of 7-aminobutyric acid or glycine (24). The stimulated release of these transmitters would be expected if the mossy fiber synaptosomal preparation were contaminated with smaller, inhibitory nerve endings that are present in intact hippocampal tissue. Thus, the mossy fiber synaptosomal preparation can be used to correlate the metabolism of arachidonic acid with the release of endogenous glutamate from a population of presynaptic terminals known to express LTP. Unesterified arachidonic acid appears to play a central role in the evoked release of endogenous glutamate from rat hippocampal mossy fiber nerve endings. Exogenous arachidonic acid stimulated the dose-dependent release of glutamate from mossy fiber synaptosomes, as did the phospholipase A 2 activator melittin. Neither of these treatments adversely affected the integrity of the synaptosomal membranes. Although the arachidonate effect was significant at 100 p,M, the concentration of free arachidonate associated with the membranes was shown to be much less. We found that only 3.5% of the added arachidonate was able to gain access to the membranes during 10 rain of superfusion. Thus, superfusion of the terminals with 200 ~M arachidonate is estimated to result in an actual concentration change of 7 ~M. These results are consistent with the observation that addition of arachidonic acid to in vitro systems results in a rapid and marked reduction in the concentration of available arachidonate due to adsorption (25). The suggestion that local concentrations of free archidonate remained in the low micromolar range is also supported by the observation that exogenous arachidonate did not liberate intrasynaptosomal LDH activity. Melittin also significantly enhanced the release of glutamate and had no effect on occluded LDH activities

748 when present at 0.1 or 1.0 ~xM. In contrast, 10 ~M melittin caused the loss of 26% of the occluded LDH, while 0.4% Triton X-100 reduced synaptosomal LDH by 89%. The viability of the presynaptic membranes is further indicated by the observations that removal of either stimulus allowed for the return to spontaneous levels of glutamate efflux. In addition, a depolarizing stimulus applied following exposure of the synaptosomes to arachidonate or melittin was still able to induce the release of endogenous glutamate. It has been suggested that certain eicosanoids, rather than arachidonate, are responsible for the modulation of transmitter release. Lipoxygenase products affect release mechanisms in Aplysia (9,10) and may be involved in the maintenance of LTP in the hippocampus (7,19). Also, prostaglandins enhance the release of adrenal catecholamines (26) and inhibition of the cyclooxygenase pathway blocks the depolarization-induced release of neurotransmitters from cortical synaptosomes (27). We observed that high concentrations of PGE2 and PGF2a were able to stimulate glutamate release, while 12-HETE had no effect. In contrast, inhibition of the cyclooxygenase or lipoxygenase pathways not only failed to block glutamate efflux, but even potentiated the release induced by membrane depolarization or exogenous arachidonate. These results can be used to suggest that eicosanoids have no direct effect on glutamate secretion and that the depolarization-induced accumulation of free arachidonate is sufficient to stimulate release of this excitatory amino acid. However, an indirect inhibitory effect of lipoxygenase products on glutamate release can not be excluded, since 12-HETE did not stimulate release, yet NDGA potentiated the depolarization- and arachidonate-evoked efflux of glutamate. The apparent conflict between the stimulation of glutamate release with exogenous prostanoids and the potentiation of glutamate release when their synthesis was inhibited may be explained if both conditions increase the availability of unesterified arachidonic acid. It might be expected that cyclooxygenase and lipoxygenase inhibitors block the oxidative removal of free arachidonate. Ibuprofen did increase the recovery of unesterified arachidonate from cerebellar mossy fiber terminals (unpublished results). In turn, exogenous prostanoids may also increase available free arachidonate by interfering with its reacylation or oxidative removal. It has been shown that fatty acid peroxides inhibit the incorporation of arachidonate into synaptosomal phospholipids (28) and it is possible that endogenous prostanoids have a similar effect. Alternatively, they may activate endogenous phospholipase A2, since prostaglandins have

Freeman, Terrian, and Dorman been shown to mobilize intracellular free calcium and induce cGMP production (29). The evidence presented in this paper may be used to suggest that any treatment that inhibits the removal of the unesterified arachidonate that accumulates in response to (1) membrane depolarization, (2) the activation of phospholipase A 2 or (3) the addition of exogenous arachidonate will potentiate glutamate efflux. This proposal is supported by the observation that linoleic acid was the only tested fatty acid that was able to stimulate glutamate release. Linoleate, but not other fatty acids, has been shown to displace arachidonate from the glycerophospholipids of endothelial cell membranes and compete with arachidonate in the cyclooxygenase pathway (30,31). Thus, linoleate appears to act as an analogue of arachidonic acid and may increase free arachidonate concentrations by competing with it in reacylation and oxidation pathways. Arachidonic acid may also be involved in the presynaptic facilitation of glutamate release associated with the potentiation of excitatory hippocampal synapses. We observed that the K+-evoked release of glutamate was significantly increased following a brief exposure to exogenous arachidonate. Prior treatment of the mossy fiber synaptosomes with melittin, which activates phospholipase A 2 (32), also significantly potentiated the glutamate release induced by membrane depolarization. The effect of melittin appeared to be specific for the activation of endogenous phospholipase A2, since bee venom phospholipase A2 had no effect on glutamate efflux. The observation that 1 ~M melittin was more potent than 200 ~M arachidonate at both stimulating glutamate release and potentiating the release in response to membrane depolarization is consistent with the suggestion that the activation of endogenous phospholipase A a was more effective at raising free arachidonate levels. With either stimulus, however, the degree of facilitation of glutamate release is likely to be underestimated, since the controls did not receive a stimulus equivalent to arachidonate or melittin prior to depolarization and the evoked release of glutamate is known to diminish with repeated stimulations. The mechanisms underlying this presynaptic facilitation may depend on an increase in the availability of cytosolic free Caa§ The induction of LTP is associated with increased availability of calcium in presynaptic terminals (11) and this additional calcium may be expected to activate Ca2+-dependent lipases in presynaptic terminals and result in increasedconcentrations of free arachidonate. This suggestion is consistent with the observed effects of Ca a+ antagonists and the Ca 2+ ionophore A23187 on the accumulation of unesterified arachidonic

Araehidonic Acid and Glutamate Release acid in cerebellar mossy fiber terminals (3,6). However, unesterified arachidonate has been shown to have mixed metabolic effects that m a y be related to the mechanisms of synaptic transmission. Although arachidonic acid has been shown to increase the outward potassium conductance (9,33) and inhibit Ca2+/calmodulin-dependent kinase II activity (34), it does stimulate increases of intracellular sodium in synaptosomes, while reducing the high-affinity uptake of glutamate and 7-aminobutyric acid in rat brain slices and synaptosomes (35). W e have observed tffat free arachidonate is a powerful inhibitor of the N a + - K + - A T P a s e associated with the hippocampal mossy fiber synaptosomes with 10 p,M arachidonate inhibiting the enzyme by more than 50% under static incubation conditions. It has been implicated in calpactindependent exocytosis (36), the activation of protein kinase C (37) and the metabolism of cyclic nucleotides (38). Regardless of the mechanisms involved, we have demonstrated that unesterified arachidonic acid enhances gIutamate release from hippocampaI mossy fiber syn~ aptosomes. In conclusion, we observed that membrane depolarization, 4-aminopyridine, arachidonic acid and melittin stimulated the dose-dependent release of endogenous glutamate from hippocampal mossy fiber synaptosomes. Both the arachidonate- and depolarization-evoked release of glutamate were potentiated in the presence of cyclooxygenase and lipoxygenase inhibitors. These results m a y be used to suggest that arachidonate acts directly on the presynaptic membrane rather than through the formation of eicosanoids. W e also observed that arachidonate and melittin caused a persistent facilitation of glutamate release from the mossy fiber synaptosomes. These effects are consistent with the hypothesis that unesterified arachidonate is involved in the sustained increase in glutamate release associated with LTP. We propose, therefore, that the availability of unesterified arachidonic acid plays a central roIe in the presynaptic mechanisms related to glutamate release from the mossy fiber terminals and m a y also be related to the expression of LTP in the hippocampal mossy fiber-CA3 synapse.

ACKNOWLEDGMENTS This research was supported by AFOSR grants 89-0245 (RVD) and 89-0531 (DMT).

REFERENCES 1. Lazarewicz, J. W., Leu, V., Sun, G. Y., and Sun, A. Y. 1983. Arachidonic acid release from K+-evoked depolarization of brain synaptosomes. Neurochem. Int. 5:471-478.

749 2. Asakura, T., and Matsuda, M. 1984. Efflux of y-aminobutyric acid from and appearance of free arachidonic acid inside synaptosomes. Biochim. Biophys. Acta 773:301-307. 3. Dorman, R. V., Schwartz, M. A., and Tertian, D. M. 1986. Prostaglandin involvement in the evoked release of D-aspartate from cerebellar mossy fiber terminals. Brain Res. Bull. 17:243248. 4. Rhoads, D. E., Osburn, L. D., Peterson, N. A., and Raghupathy, E. 1983. Release of neurotransmitter amino acids from synaptosomes: enhancement of calcium-independent efflux by oleic and arachidonic acids. J. Neurochem. 41:531-537. 5. Frye, R. A., and Holz, R. W. 1984. The relationship between arachidonic acid release and catect~olaminesecretion from cultured bovine adrenal chromaffin cells. J. Neurochem. 43:146150. 6. Dorman, R. V., Schwartz, M. A., and Tertian, D. M. 1988. Depolarization-induced [3H]arachidonic acid accumulation: effects of external Ca2+ and phospholipase inhibitors. Brain Res. Bull. 21:445-450. 7. Lynch, M. A., Errington, M. L., and Bliss, T. V. P. 1989. Nordihydroguaiareticacid blocks the synaptic component of longterm potentiation and the associated increases in release of glutamate and arachidonate: an in vivo study in the dentate gyrus of the rat. Neuroseience 30:693-701. 8. Williams, J. H., Errington, M. L., Lynch, M. A., and Bliss, T. V. P. 1989. Arachidonic acid induces a long-term activity-dependent enhancement of synaptic transmission in the hippocampus. Nature 341:739-742. 9. Piomelli, D., Volterra, A., Dale, N., Siegelbaum, S. A., Kandel, E. R., Schwartz, J. H., and Belardetti, F. 1987. Lipoxygenase metabolites of arachidonic acid are second messengers for presynaptic inhibition of Aplysia sensory cells. Nature 328:38-43. 10. Piomelli, D., Shapiro, E., Zipkin, R., Schwartz, J. H. and Feinmark, S. J. 1989. Formationand action of 8-hydroxy-11,12-epoxy5,9,14-icosatrienoic acid in Aplysia: a possible second messenger in neurons. Proc. Natl. Acad. Sci. USA 86:1721-1725. 11. Agoston, D. V., and Kuhnt, U. 1986. Increased Ca-uptake of presynaptic terminals during long-term potentiation in hippocampal slices. Exp. Brain Res. 62:663-668. 12. Tertian, D. M., Johnston, D., Claiborne, B. J., Ansah-Yiadom, R., Strittmatter, W. J., and Rea, M. A. 1988. Glutamate and dynorphin release from a subcellualar fraction enriched in hippocampal mossy fiber synaptosomes. Brain Res. Bull. 21:343351. 13. Layne, E. 1957. Spectrophotometric and turbidimetric methods for measuring protein. Pages 447-454, in Colowick, S. P., and Kaplan, N. O. (eds.), Methods in Enzymology. VoI. III, Academic Press, New York. 14. Lowry, O. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. 195i. Protein measurement with Folin phenol reagent. J. Biol. Chem. 193:265-275. 15. Graham, L. T., Jr., and Aprison, M. H. 1966. Fluorometric determination of aspartate, glutamate and gamma-aminobutyrate in nerve tissue using enzymic methods. Anal. Biochem. 15:487497. 16. Dagani, F., and Erecinska, M. 1987. Relationships among ATP synthesis, K § gradients, and neumtransmitter amino acid levels in isolated rat brain synaptosomes. J. Neurochem. 49:1229-1240. 17. Dolphin, A. C., Errington, M. L., and Bliss, T. V. P. 1982. Long-term potentiation of the perforant path in vivo is associated with increased glutamate release. Nature 297:496-498. 18. Bliss, T. V. P., Douglas, R. M., Errington, M. L. and Lynch, M. A. 1986. Correlation between long-term potentiation and release of endogenous amino acids from dentate gyms of anaesthetized rats. J. Physiol. (Lond.) 377:391-408. 19. Lynch, M. A. 1989. Mechanisms underlying induction and maintenance of long-term potentiation in the hippocampus. Bioessays 10:85-90. 20. Barrionuevo, G., Kelso, S. R., Johnston, D., and Brown, T. H.

750

21. 22. 23. 24.

25. 26.

27.

28. 29.

30.

1986. Conductance mechanism responsible for long-term potentiation in monosynaptic and isolated excitatory synaptic inputs to hippocampus. J. Neurophysiol. 55:540-550. Harris, E. W, and Cotman, C. W. 1986. Long-term potentiation of guinea pig mossy fiber responses is not blocked by N-methylD-aspartate antagonists. Neurosci. Lett. 70:i32-137. Ito, I., Okada, D., and Sugiyama, H. 1988. Pertussis toxin suppresses Iong-term potentiation of hippocampal mossy fiber synapses. Neurosci. Lett. 90:181-185. Crawford, I. L., and Conner, J. D. 1973. Localization and release of glutamic acid in relation to the hippocampal mossy fiber pathway. Nature 244:442-443. Terrian, D. M., Gannon, R. L., and Rea, M. A. 1990. Glutamate is the endogenous amino acid selectively released by rat hippocampal mossy fiber synaptosomes concomitantly with pro@norphin-derived peptides. Neurochem. Res. 15:1-5. Samples, D. R., Sprague, E. A., Harper, M. 1. K. and Herlihy, J. T. 1989. In vitro adsorption losses of arachidonic acid and calcium ionophore A23187. Am. J. Physiol. 257:Cl166-Cl170. Yokohama, H., Tanaka, T., Ito, S., Negishi, M., Hayashi, H., and Hayaishi, O. 1988. Prostaglandin E receptor enhancement of catecholamine release may be mediated by phosphoinositide metabolism in bovine adrenal chromaffin ceils. J. Biol. Chem. 263:1119-1122. Bradford, P. G., Marinetti, G. V., and Abood, L. G. 1983. Stimulation of phospholipase A 2 and secretion of catecholamines from brain synaptosomes by potassium and A23187. J. Neurochem. 41:1684-1693. Zaleska, M. M., and Wilson, D. F. 1989. Lipid hydroperoxides inhibit reacylation of phospholipids in neuronal membranes. J. Neurochem. 52:255-260. Miwa, N., Sugino, H., Ueno, R., and Hayaishi, O. 1988. Prostaglandin induces Ca2+ influx and cyclic GMP formation in mouse neuroblastoma x rat glioma hydrid NG108-15 cells in culture. J. Neurochem. 50:1418-1424. Spector, A. A., Hoak, H. C., Fry, G. L., Denning, G. M., Stoll,

Freeman, Terrian, and Dorman

31. 32. 33.

34.

35.

36. 37.

38.

L. L., and Smith, J. B. 1980. Effect of fatty acid modification on prostacyclin production by cultured human endothelial cells. J. CIin. Invest. 65:1003-1012. Kaduce, T. L., Spector, A. A., and Bar, R. S. 1982. Linoleic acid metabolism and prostaglandin production by cultured bovine pulmonary artery endothelial ceils. Arteriosclerosis 2:380-389. Conricode, K. M., and Ochs, R. S. 1989. Mechanism for the inhibitory and stimulatory actions of proteins on the activity of phospholipase A~. Biochim. Biophys. Acta 1003:36-43. Volterra, A., and Siegelbaum, S. A. 1989. Antagonistic modulation of S-K + channel activity by cyclic AMP and arachidonic acid metabolites: role for two G proteins. Pages 219-236, in Barkai, A. I, and Bazan, N. G. (eds.), Arachidonic acid metabolism in the nervous system. Physiological and pathological significance. Annals N.Y. Acad. Sci., Vol. 559, New York. Piomelli, D., Wang, J. K. T., Sihra, T. S., Nairn, A. C., Czernik, A. J., and Oreengard, P. 1989. Inhibition of a Ca~+/calmodulindependent protein kinase II by arachidonic acid and its metabolites. Proc. Natl. Acad. Sci. USA 86:8550-8554. Chan, P. H., Kerlan, R., and Fishman, R. A. 1983. Reductions of gamma-aminobutyric acid and glutamate uptake and (Na +-K+)ATPase activity in brain slices and synaptosomes by arachidonic acid. J. Neurochem. 40:309--316. All, S. M., Geisow, M. J., and Burgoyne, R. D. 1989. A role for calpactin in calcium-dependent exocytosis in adrenal chromaffin cells. Nature 340:313-315. Murakami, K., and Routtenberg, A. 1985. Direct activation of purified protein kinase C by unsaturated fatty acids (oleate and arachidonate) in the absence of phospholipids and Ca2+. FEBS Lett. 192:189-193. Duman, R. S., Karbon, E. W., Harrington, C., and Enna, S. J. 1986. An examination of the involvement of phospholipases A2 and C in the [3-adrenergic and -,/-aminobutyric acid receptor modulation of cyclic AMP accumulation in rat brain slices. J. Neurochem. 47:800-810.