Prostaglandin E2 and 4-Aminopyridine Prevent the Lipopolysaccharide-Induced Outwardly Rectifying Potassium Current and Interleukin-1β Production in Cultured Rat Microglia

Abstract: Brain inflammation includes microglial activation and enhanced production of diffusible chemical mediators, including prostaglandin E₂. Prostaglandin E₂ is generally considered a proinflammatory molecule, but it also promotes neuronal survival and down-regulates some aspects of microglial activation. It remains unknown, however, if and how prostaglandin E₂ prevents microglial activation. In primary culture, microglial activation is predicted by a characteristic pattern of whole-cell potassium currents and interleukin-1β production. We investigated if prostaglandin E₂ could alter these currents and, if so, whether these currents are necessary for microglial activation. Microglia were isolated from mixed cell cultures prepared from neonatal rat brains and exposed to 0–10 µ M prostaglandin E₂ and lipopolysaccharide for 24 h. Currents were elicited by using standard patch-clamp technique, and interleukin-1β production was measured by ELISA. Peak outward current densities in microglia treated with lipopolysaccharide plus prostaglandin E₂ (10 n M ) were reduced significantly from those of cells treated with lipopolysaccharide alone. Prostaglandin E₂ and 4-aminopyridine (a blocker of outward potassium currents) also significantly reduced interleukin-1β production. Thus, although prostaglandin E₂ is classified generally as a proinflammatory chemical, it has complex roles in brain inflammation that include preventing microglial activation, perhaps by reducing the outward potassium current.     

Interferon-γ and Nitric Oxide Down-Regulate Lipopolysaccharide-Induced Prostanoid Production in Cultured Rat Microglial Cells by Inhibiting Cyclooxygenase-2 Expression

Abstract: We have used purified microglial cultures obtained from neonatal rat brains to study some aspects of the regulation of prostanoid production in these cells. We found that nitric oxide, which is released into the culture medium along with prostanoids when the cells are exposed to lipopolysaccharide (1–100 ng/ml), can down-regulate prostanoid biosynthesis. Indeed, the abrogation of endogenous nitric oxide production, obtained by depleting the medium of the precursor l -arginine or by blocking the activity of nitric oxide synthase by the specific inhibitor N^G-monomethyl-l -arginine, led to a remarkable increase in lipopolysaccharide-induced prostanoid release. Moreover, the nitric oxide-generating compound 3-morpholinosydnonimine caused a substantial reduction of prostanoid formation, in the absence of endogenous nitric oxide, suggesting that both endogenous and exogenous nitric oxide may inhibit the induced prostanoid production. We also found that interferon-γ potentiated the effect of lipopolysaccharide on nitrite accumulation as previously described by others and depressed the lipopolysaccharide-evoked production of prostaglandin E₂, prostaglandin D₂, and thromboxane. It is interesting that the inhibitory effect of interferon-γ on prostanoid production did not appear to depend on the potentiation of NO release, as it was present also when the endogenous synthesis of nitric oxide was abrogated. Additional experiments showed that interferon-γ and nitric oxide act mainly by down-regulating the lipopolysaccharide-induced enzymatic activity and expression (analyzed by western blot) of cyclooxygenase-2. Our data indicate that microglial prostanoid biosynthesis induced by proinflammatory stimuli, such as lipopolysaccharide, is tightly regulated by nitric oxide. Interferon-γ appears to affect the balance between these local mediators by favoring nitric oxide production and inhibiting the prostanoid cascade and may thus contribute to the modulation of inflammation, local immune reactivity, and neuronal damage.     

Prostaglandin Induces Ca2+ Influx and Cyclic GMP Formation in Mouse Neuroblastoma × Rat Glioma Hybrid NG108–15 Cells in Culture

Various prostaglandins (PGs) (10 nM-30 microM) were added to NG108-15 cells in culture, and changes in the levels of intracellular cyclic GMP and Ca2+ were investigated. Exposure of the cells to PGF2 alpha, PGD2, and PGE2 (10 microM) transiently increased the cyclic GMP content 7.5-, 3.9-, and 3.1-fold, respectively. Furthermore, the increased levels of cyclic GMP correlated well with the rise in cytosolic free Ca2+ concentrations induced by the PGs. Other PGs (10 microM), including metabolites and synthetic analogs, which had no effect on intracellular Ca2+, failed to increase the cyclic GMP content in the cells. When extracellular Ca2+ was depleted from the culture medium, the PG-induced increase in cyclic GMP level was almost completely abolished. In addition, treatment of the cells with quin 2 tetraacetoxymethyl ester dose-dependently inhibited the PG-induced cyclic GMP formation. The increase in cyclic GMP content caused by treatment of the cells with a high K+ level (50 mM) was completely blocked by voltage-dependent Ca2+ entry blockers, such as verapamil (10 microM), nifedipine (1 microM), and diltiazem (100 microM); however, the PG (10 microM)-induced increase in cyclic GMP content was not affected by such Ca2+ entry blockers. These findings indicate that PG-induced cyclic GMP formation may require the rise in intracellular Ca2+ level and that the voltage-dependent Ca2+ channels may not be involved in the PG-induced rise in Ca2+ content.     

Changes in β-Adrenergic Receptor Subtypes in Alzheimer-Type Dementia

Using ligand binding techniques, we studied beta-adrenergic receptor subtypes in brains obtained at autopsy from seven histologically normal controls and seven histopathologically verified cases with Alzheimer-type dementia (ATD). Inhibition of [3H]dihydroalprenolol [( 3H]DHA) binding by the selective beta 1 antagonist, metoprolol, results in nonlinear Hofstee plots, suggesting the presence of the two receptor subtypes in the human brain. The calculated ratios of beta 1/beta 2-adrenergic receptors in control brains are as follows: frontal cortex, 49:51; temporal cortex, 31:69; hippocampus, 66:34; thalamus, 23:77; putamen, 70:30; caudate, 48:52; nucleus basalis of Meynert (NbM), 43:57; cerebellar hemisphere, 25:75. Compared with the controls, total concentrations of beta-adrenergic receptors were significantly reduced only in the thalamus of the ATD brains. beta 1-Adrenergic receptor concentrations were significantly reduced in the hippocampus and increased in the NbM and cerebellar hemisphere, whereas beta 2-adrenergic receptor concentrations were significantly reduced in the thalamus, NbM, and cerebellar hemisphere and increased in the hippocampus and putamen of the ATD brains. These results suggest that beta 1- and beta 2-adrenergic receptors are present in the human brain and that there are significant changes in both receptor subtypes in selected brain regions in patients with ATD.     

Intracerebral NMDA Injection Stimulates Production of Interleukin-1β in Perinatal Rat Brain

Abstract: Susceptibility to NMDA neurotoxicity peaks in the early postnatal period in rats. Although indirect evidence suggests that interleukin-1β is a mediator of NMDA neurotoxicity in perinatal rats, direct confirmation of NMDA-induced interleukin-1β production in the brain has not been reported previously. The primary goal of this study was to determine if intracerebral injection of a neurotoxic dose of NMDA stimulates interleukin-1β production acutely. We used a rat-specific interleukin-1β ELISA to quantify brain tissue homogenate interleukin-1β content, and an immunocytochemical assay with a monoclonal anti-rat interleukin-1β antibody to visualize its distribution. NMDA (10 nmol) was injected stereotaxically into 7-day-old rats, using coordinates that targeted the striatum and overlying dorsal hippocampus. Interleukin-1β concentrations were measured in samples from the injected and contralateral cerebral hemispheres 0–12 h later; in addition, the impact of treatment with the noncompetitive NMDA antagonist MK-801 on interleukin-1β production was assessed. We found marked increases in tissue content of interleukin-1β in the lesioned hemisphere; values peaked at 6 h post injection. Treatment with MK-801 (1 mg/kg) blocked NMDA-induced increases in interleukin-1β. Preliminary immunocytochemical analysis demonstrated high concentrations of interleukin-1β-immunoreactive cells in the lesioned hippocampus, and concurrent increases in interleukin-1β immunoreactivity diffusely in the ependyma at 6 h after NMDA administration. Our data provide the first direct evidence that NMDA-induced excitotoxic injury stimulates interleukin-1β production in vivo.     

Lipopolysaccharide and Interleukin-1β Augmented Histidine Decarboxylase Activity in Cultured Cells of the Rat Embryonic Brain

Abstract: We investigated the effect of lipopolysaccharide (LPS) and various inflammatory cytokines on the histidine decarboxylase (HDC) activity in cultured cells of the rat embryonic brain. Histaminergic neuronal cell bodies were supposed to exist in cultured cells of the diencephalon but not in those of the cortex. The HDC activity was elevated by adding LPS and interleukin-1 β (IL-1β) but not by tumor necrosis factor-α (TNF-α) and IL-6 to the mixed primary cultures of diencephalon. In the adherent cell fraction of the cultured diencephalon cells, HDC activity was also enhanced by LPS and IL-1β. In a similar manner, LPS augmented HDC activity in the mixed primary culture of cerebral cortical cells and in its adherent cell fraction. The effects of IL-1β but not LPS in the mixed primary culture of diencephalon were canceled by a prior exposure to cytosine-β- d -arabinofuranoside. The changes in HDC activity after exposure to LPS for 12 h were not accompanied by increased mRNA levels. In these cell cultures, mast cells were not detected by Alcian Blue staining. These results indicated the presence of the third type of HDC-bearing cell besides neurons and mast cells in the brain. The increase of HDC activity by IL-1β might be due to cell proliferation.     

Alzheimer's β-Amyloid Peptide 1–42 Induces a Phagocytic Response in Murine Microglia

Abstract: β-Amyloid (Aβ) peptides are a key component of the senile plaques that characterize Alzheimer's disease. Cytokine-producing microglia have been shown to be intimately associated with amyloid deposits and have also been implicated as scavengers responsible for clearing Aβ deposits. However, little is known about the initial activation of these microglia or the effect of Aβ on phagocytosis. Murine BV-2 microglia were used to assess the effect of synthetic Aβ 1–42 on phagocytosis by quantifying uptake of fluorescent microspheres, acetylated low-density lipoproteins, and zymosan particles by flow cytometry. Aβ 1–42 stimulated microglial phagocytosis in a time- and dose-dependent manner. Aβ fibrils produced the greatest potentiation, and once activated, phagocytosis remained elevated after removal of Aβ from the cultures. Aβ-stimulated phagocytosis could be blocked if proteoglycans were first complexed to Aβ fibrils. These data suggest that Aβ fibrils act as an immune signal to stimulate microglial phagocytosis and that extracellular matrix molecules may modify Aβ function.     

Regulation of Growth Inhibitory Factor Expression by Epidermal Growth Factor and Interleukin-1β in Cultured Rat Astrocytes

Growth inhibitory factor (GIF) is highly expressed in the CNS under physiological conditions, but its expression is reduced in neurodegenerative diseases, such as Alzheimer's disease. The results of this study show that the levels of GIF and GIF mRNA were not influenced by neuroglial interactions. GIF was highly expressed in confluent astrocytes, but the expression was down-regulated in low-density growing astrocytes. A high level of GIF was not observed in serum-starved low-density cultures. These findings suggest that GIF is a quiescent state-specific protein and that two different mechanisms may exist for the cells to enter the quiescent state. Among interleukin-1beta (IL-1beta), fibroblast growth factor-2, epidermal growth factor (EGF), amyloid beta1-42, and 50% O2, only EGF and IL-1beta altered the level of GIF in confluent astrocytes: EGF increased both GIF mRNA and protein, and IL-1beta decreased GIF mRNA, but did not alter GIF protein. Kinetic analysis of the GIF mRNA level revealed the biphasic regulation of GIF mRNA expression by IL-1beta, i.e., a transient up-regulation followed subsequently by down-regulation, explaining in part the discrepancy between the levels of GIF mRNA and protein in astrocytes treated with IL-1beta.     

Interleukin-1β-Mediated Regulation of μ-Opioid Receptor mRNA in Primary Astrocyte-Enriched Cultures

Abstract: Opioids have been found to modulate the immune system by regulating the function of immunocompetent cells. Several studies suggest that the interaction between immune and opioid systems is not unidirectional, but rather reciprocal, in nature. In the CNS, one cellular target of immune system activation is the astrocytes. These glial cells have been shown to produce the opioid peptide, proenkephalin, to express the μ-, δ-, and κ-opioid receptors, and to respond to the immune factor interleukin-1β (IL1β) with an increased proenkephalin synthesis. To characterize more completely the astrocytic opioid response to immune factor stimulation, we examined the effect of IL1β (1 ng/ml) on the μ-receptor mRNA expression in primary astrocyte-enriched cultures derived from rat (postnatal day 1–2) cortex, striatum, cerebellum, hippocampus, and hypothalamus. A 24-h treatment with IL1β produced a 70–80% increase in the μ-receptor mRNA expression in the striatal, cerebellar, and hippocampal cultures but had no effect on this expression in the cortical and hypothalamic cultures. This observation represents one of the few demonstrated increases in levels of the μ-receptor mRNA in vitro or in vivo, since the cloning of the receptor. The enhanced μ-receptor mRNA expression, together with the previous observation that IL1β stimulates proenkephalin synthesis in astrocytes, supports the IL1β-mediated regulation of an astroglial opioid peptide and receptor in vitro, a phenomenon that may be significant in the modulation of the gliotic response to neuronal damage. Therefore, the astroglial opioid "system" may be important in the IL1β-initiated, coordinated response to CNS infection, trauma, or injury.     

Phloretin as an Antagonist of Prostaglandin F2α Receptor in Cultured Rat Astrocytes

Abstract: The effect of phloretin on prostaglandin (PG) F_2α-induced phosphoinositide hydrolysis and elevation of intracellular Ca²⁺ concentration was examined in cultured rat astrocytes. Phloretin inhibited PGF_2α (1 μ M )-induced phosphoinositide hydrolysis in a concentration-dependent manner with an IC₅₀ value of 16 μ M . The inhibitory action of phloretin was specific for PGs. The addition of increasing concentrations of phloretin caused progressive shifts of the dose-response curves of PGF_2α to the right. In digitoninpermeabilized astrocytes, phloretin (100 μ M ) inhibited the stimulation induced by PGF_2α (1 μ M ) plus GTPγS (50 μ M ) without affecting that induced by GTPγS alone. PGF_2α at 1 μ M transiently increased astrocytic intracellular Ca²⁺ concentration in 39% of the cells tested. The response was completely blocked by 100 μ M phloretin and the calcium response recovered again after washing out phloretin. These results suggest that phloretin is an antagonist of PGF_2α receptor linked to phospholipase C in astrocytes.     

Immunohistochemical Localization of G Protein β1, β2, β3, β4, β5, and γ3 Subunits in the Adult Rat Brain

Abstract: The regional distributions of the G protein β subunits (Gβ1–β5) and of the Gγ3 subunit were examined by immunohistochemical methods in the adult rat brain. In general, the Gβ and Gγ3 subunits were widely distributed throughout the brain, with most regions containing several Gβ subunits within their neuronal networks. The olfactory bulb, neocortex, hippocampus, striatum, thalamus, cerebellum, and brainstem exhibited light to intense Gβ immunostaining. Negative immunostaining was observed in cortical layer I for Gβ1 and layer IV for Gβ4. The hippocampal dentate granular and CA1–CA3 pyramidal cells displayed little or no positive immunostaining for Gβ2 or Gβ4. No anti-Gβ4 immunostaining was observed in the pars compacta of the substantia nigra or in the cerebellar granule cell layer and Purkinje cells. Immunoreactivity for Gβ1 was absent from the cerebellar molecular layer, and Gβ2 was not detected in the Purkinje cells. No positive Gγ3 immunoreactivity was observed in the lateral habenula, lateral septal nucleus, or Purkinje cells. Double-fluorescence immunostaining with anti-Gγ3 antibody and individual anti-Gβ1–β5 antibodies displayed regional selectivity with Gβ1 (cortical layers V–VI) and Gβ2 (cortical layer I). In conclusion, despite the widespread overlapping distributions of Gβ1–β5 with Gγ3, specific dimeric associations in situ were observed within discrete brain regions.     

Regional Variations in α1-Adrenergic Receptor Subtypes in Rat Brain

alpha 1-Adrenergic receptor subtypes were differentiated by their affinities for the competitive antagonist WB 4101 and their sensitivities to inactivation by chlorethylclonidine (CEC) in eight rat brain regions. WB 4101 showed low Hill coefficients for inhibition of specific 125I-[2-beta-(4-hydroxyphenyl)ethylaminomethyl]tetralone (125IBE) binding in all regions. Nonlinear regression analysis showed that there were two binding sites with different affinities for WB 4101 in each region. The proportions of these sites varied among regions, although the affinity of WB 4101 for each site remained constant. Thalamus and cerebral cortex had the highest proportion of low-affinity sites, whereas hippocampus and pons-medulla had the highest proportion of high-affinity sites. Pretreatment with CEC in hypotonic buffer significantly reduced the density of 125IBE binding sites in all brain regions. Cerebral cortex and cerebellum had the highest proportion of CEC-sensitive sites, whereas hippocampus and spinal cord had the highest proportion of CEC-insensitive sites. There was a significant correlation between the proportion of binding sites with a low affinity for WB 4101 and those sensitive to inactivation by CEC.     

Distribution of mRNAs Encoding the Peroxisome Proliferator-Activated Receptor α, β, and γ and the Retinoid X Receptor α, β, and γ in Rat Central Nervous System

Abstract: We report the isolation, by RT-PCR, of partial cDNAs encoding the rat peroxisome proliferator-activated receptor (PPAR) isoforms PPARα, PPARβ, and PPARγ and the rat retinoid X receptor (RXR) isoforms RXRα, RXRβ, and RXRγ. These cDNAs were used to generate antisense RNA probes to permit analysis, by the highly sensitive and discriminatory RNase protection assay, of the corresponding mRNAs in rat brain regions during development. PPARα, PPARβ, RXRα, and RXRβ mRNAs are ubiquitously present in different brain regions during development, PPARγ mRNA is essentially undetectable, and RXRγ mRNA is principally localised to cortex. We demonstrate, for the first time, the presence of PPAR and RXR mRNAs in primary cultures of neonatal meningeal fibroblasts, cerebellar granule neurons (CGNs), and cortical and cerebellar astrocytes and in primary cultures of adult cortical astrocytes. PPARα, PPARβ, RXRα, and RXRβ mRNAs are present in all cell types, albeit that PPARα and RXRα mRNAs are at levels near the limit of detection in CGNs. PPARγ mRNA is expressed at low levels in most cell types but is present at levels similar to those of PPARα mRNA in adult astrocytes. RXRγ mRNA is present either at low levels, or below the level of detection of the assay, for all cell types studied.     

Compartments of Labeled and Endogenous γ-Aminobutyric Acid Giving Rise to Release Evoked by Potassium or Veratridine in Rat Cortical Slices

Abstract— To establish compartments involved in depolarization-induced release of γ-aminobutyric acid (GABA) in rat brain slices, the amount of exogenous labeled and endogenous GABA released and retained was followed during 48 min exposure to 50 m m -K⁺ or to 50 μ m -veratridine. Endogenous GABA was measured with high performance liquid chromatography. The presence of 10 μ m -aminooxyacetic acid throughout prevented both the metabolism of GABA and the formation of endogenous GABA due to depolarization. During super-fusion with 50 m m -K⁺ and 2.6 m m -Ca²⁺ the efflux of labeled and endogenous GABA after an initial large increase declined to 10% of the highest value with constant and identical rates. Kinetic analysis of efflux showed that 10% of endogenous and 25% of labeled GABA present is available for release by high K⁺ and Ca²⁺. In the absence of Ca²⁺, release by high K⁺ of both labeled and endogenous GABA was nearly suppressed. Veratridine, unlike high K⁺, caused an efflux which declined with an initial fast and late very slow phase. The slow efflux by veratridine was doubled in the absence of Ca²⁺. Exposure to veratridine in the absence of Ca²⁺ during 120 min released nearly 70% of labeled and endogenous GABA present. Results suggest that only about 0.25 μmol g⁻¹ endogenous GABA is the source of physiological Ca²⁺-dependent release, while much of the remaining GABA present is released only under unphysiological conditions.     

Interleukin-1β Induces Prostaglandin G/H Synthase-2 (Cyclooxygenase-2) in Primary Murine Astrocyte Cultures

Abstract: Activation of glial cells and the consequent release of cytokines, proteins, and other intercellular signaling molecules is a well-recognized phenomenon in brain injury and neurodegenerative disease. We and others have previously described an inducible prostaglandin G/H synthase, known as PGHS-2 or cyclooxygenase-2, that is up-regulated in many cell systems by cytokines and growth factors and down-regulated by glucocorticoid hormones. In cultured mouse astrocytes we observed increased production of prostaglandin E₂ (PGE₂) after stimulation with either interleukin-1β (IL-1β) or the protein kinase C activator phorbol 12-myristate 13-acetate (TPA). This increase in PGE₂ content was blocked by pretreatment with dexamethasone and correlated with increases in cyclooxygenase activity measured at 4 h. Northern blots revealed concomitant increases in PGHS-2 mRNA levels that peaked at 2 h and were dependent on the dosage of IL-1β. Dexamethasone inhibited this induction of PGHS-2 mRNA by IL-1β. TPA, basic fibroblast growth factor, and the proinflammatory factors tumor necrosis factor α and lipopolysaccharide, but not interleukin-6, also stimulated PGHS-2 mRNA expression. Relative to IL-1β, the greater increases in PGE₂ production and cyclooxygenase activity caused by TPA correlated with a greater induction of PGHS-2 mRNA. Furthermore, NS-398, a specific inhibitor of cyclooxygenase-2, blocked >80% of the cyclooxygenase activity in TPA-treated astrocytes. These findings indicate that increased expression of PGHS-2 contributes to prostaglandin production in cultured astrocytes exposed to cytokines and other factors.     

Evidence that the Peripheral-Type Benzodiazepine Receptor Ligand Ro 5–4864 Inhibits β-Endorphin Release from AtT-20 Cells by Blockade of Voltage-Dependent Calcium Channels

The demonstrations that Ro 5-4864, a ligand selective for the peripheral-type benzodiazepine (BZD) binding site, inhibited cellular differentiation and proliferation and that occupancy of the peripheral-type BZD binding site likely mediated the observed BZD effects on diverse endocrine tissues suggested that Ro 5-4864 disrupted a common cellular regulatory event. Using a well-characterized anterior pituitary-derived tumor cell line (AtT-20 cells), which synthesizes and secretes adrenocorticotropic hormone (ACTH), beta-lipotropin hormone (beta-LPH), and beta-endorphin (BE), we have investigated the molecular mechanism of action of Ro 5-4864's capacity to alter BE secretion. Ro 5-4864 inhibits basal and induced BE release from AtT-20 cells, through a cyclic AMP-independent mechanism. Ro 5-4864 completely blocked the corticotropin-releasing hormone and forskolin-induced release of BE without altering the concomitant production of cyclic AMP. The addition to AtT-20 cells of CGP 28392, a dihydropyridine that has been demonstrated in other systems to specifically activate voltage-dependent Ca2+ channels, resulted in a cyclic AMP-independent, dose-related increase in BE secretion. This CGP-induced BE release was blocked by increasing concentrations of Ro 5-4864. In contrast to the capacity of Ro 5-4864 to block CGP-induced BE release, Ro 5-4864 lacked the capacity to block enhanced BE secretion due to the calcium ionophore A23187, which increases intracellular Ca2+ levels independent of the voltage-dependent Ca2+ channels. Our findings suggest that Ro 5-4864 inhibits BE secretion from AtT-20 cells through a blockade of the voltage-dependent membrane Ca2+ channels and this mechanism of action may be responsible for Ro 5-4864's diverse effects observed on other cell types.     

Glutamate Receptor Subtypes in Cultured Cerebellar Neurons: Modulation of Glutamate and γ-Aminobutyric Acid Release

Using cerebellar, neuron-enriched primary cultures, we have studied the glutamate receptor subtypes coupled to neurotransmitter amino acid release. Acute exposure of the cultures to micromolar concentrations of kainate and quisqualate stimulated D-[3H]aspartate release, whereas N-methyl-D-aspartate, as well as dihydrokainic acid, were ineffective. The effect of kainic acid was concentration dependent in the concentration range of 20-100 microM. Quisqualic acid was effective at lower concentrations, with maximal releasing activity at about 50 microM. Kainate and dihydrokainate (20-100 microM) inhibited the initial rate of D-[3H]aspartate uptake into cultured granule cells, whereas quisqualate and N-methyl-DL-aspartate were ineffective. D-[3H]Aspartate uptake into confluent cerebellar astrocyte cultures was not affected by kainic acid. The stimulatory effect of kainic acid on D-[3H]aspartate release was Na+ independent, and partly Ca2+ dependent; the effect of quisqualate was Na+ and Ca2+ independent. Kynurenic acid (50-200 microM) and, to a lesser extent, 2,3-cis-piperidine dicarboxylic acid (100-200 microM) antagonized the stimulatory effect of kainate but not that of quisqualate. Kainic and quisqualic acid (20-100 microM) also stimulated gamma-[3H]-aminobutyric acid release from cerebellar cultures, and kynurenic acid antagonized the effect of kainate but not that of quisqualate. In conclusion, kainic acid and quisqualic acid appear to activate two different excitatory amino acid receptor subtypes, both coupled to neurotransmitter amino acid release. Moreover, kainate inhibits D-[3H]aspartate neuronal uptake by interfering with the acidic amino acid high-affinity transport system.