Calcium-sensitive cytosolic phospholipase A2 (cPLA2) is expressed in human brain astrocytes

BRAIN RESEARCH Brain Research 637 (1994) 97-105

ELSEVIER

Research Report

Calcium-sensitive cytosolic phospholipase A 2 (cPLA 2) is expressed in human brain astrocytes Diane T. Stephenson, Joseph V. Manetta, Donald L. White, X. Grace Chiou, Laura Cox, Bruce Gitter, Patrick C. May, John D. Sharp, Ruth M. Kramer, James A. Clemens * Ldly Research Laboratories, Ell Ldly and Company, Ldly Corporate Center, Indtanapohs, IN 46285, USA

(Accepted 28 September 1993)

Abstract

Calcium-sensitive cytosolic phospholipase A 2 (cPLA 2) is responsible for receptor-mediated liberation of arachidonic acid, and thus plays an important role in the initiation of the inflammatory lipid-mediator cascade generating eicosanoids and platelet-activating factor. In this study we have investigated the cellular distribution of cPLA 2 in brain using a monoclonal antibody raised against c P L A 2 t o immunostain tissue sections of human cerebral cortex. We have localized cPLA 2 in astrocytes of the gray matter. Colocalization with glial fibrillary acidic protein (GFAP) confirmed that cPLA 2 is associated predominantly with protoplasmic astrocytes. Astrocytes of the white matter, on the other hand, were not immunoreactive. In experiments using different human astrocytoma cell lines we found that c P L A 2 c a n be immunochemically localized in UC-11 MG cells, but cannot be detected in U-373 MG cells. This finding is consistent with the observation that c P L A 2 mRNA as well as cPLA 2 enzymatic activity can be readily measured in UC-11 MG astrocytoma cells, yet cannot be detected in U-373 MG cells. Our data suggest that the astrocyte is a primary source of cPLA 2 in the brain and provide further evidence for the importance of this cell type in inflammatory processes in the brain. Key words: Phospholipase A2; Astrocyte; Cortex; Astrocytoma; Eicosanoid; Inflammation

I. Introduction

Phospholipase A 2 catalyzes the hydrolysis of m e m brane phospholipids, liberating free fatty acid and lysophospholipid. Arachidonic acid is typically released from m e m b r a n e phospholipids when hydrolyzed by P E A 2 and serves as a precursor for several biologically active lipids including prostaglandins, leukotrienes and thromboxanes (collectively called eicosanoids). Moreover, PLAE-mediated hydrolysis of ether-linked phospholipids produces the precursor for platelet-activating factor. Activation of P L A 2 can also result in m e m brane destruction and is triggered by inflammation, ischemia or injury. Two types of mammalian P L A 2 have recently been purified, cloned and sequenced, namely the 14-kDa secretory P L A z (sPLA 2) [36,56] and the 85-kDa cy-

* Corresponding author. Fax: (1) (317) 276-9276. 0006-8993/94/$07 00 © 1994 Elsevier Science B.V. All nghts reserved SSDI 0006-8993(93)E1368-D

tosolic P L A 2 (cPLA 2) [13,14,37,58]. Although sPLA 2 and cPLA 2 perform the same catalytic function, their structural and biochemical properties are very different. Another 40-kDa cytosolic P L A 2 activity has been purified and characterized, but its primary structure has not yet been delineated [24]. cPLA 2 has been thought to control receptor-mediated eicosanoid production and to participate in intracellular signal transduction processes. Recent studies investigating the biochemical mechanisms responsible for the control and regulation of cPLA 2 provide further evidence for its crucial role in the cellular production of eicosanoids [38,39]. Direct investigations of P L A 2 and its role in the brain have been somewhat limited. The brain contains significant P L A 2 activity [66] and is rich in arachidonic acid-containing phospholipids [43]. Elevated levels of free fatty acids, arachidonic acid and its metabolites have been observed following cerebral ischemia [54], hypoglycemia [62] and seizures [5,59]. T h e r e is some evidence that P L A 2 may play a fundamental role in

98

D T Stephenson et al ,/Brain Research 637 (1994) 97-105

central nervous system (CNS) physiology. Arachidonlc acid and its eicosanoid metabohtes have been implicated in the control of behavior [11], regulation of blood flow [67] and modulation of neural [35] and immune [21] function. Specific CNS mechanisms in which arachidonic acid metabolites are critically involved include synaptic receptor signal transduction pathways [1], ion channel activities [33,50,55] and neurotransmitter release [48]. Many of these effects are thought to be mediated by neuronal PLA 2, yet a glial source of PLA 2 cannot be ruled out. Astroglia are poised to respond to alterations in free fatty acids in the brain. They express prostaglandin receptors [30] and produce eicosanoids [6,46] and arachidonic acid [61]. Importantly, these cells synthesize and secrete cytokines and other inflammatory mediators [reviewed by 4]. In addition, astrocytes can propagate calcium signals [15]. Since cytokines, inflammatory mediators and increments in cytosolic free calcium are factors known to affect levels of expression and/or enzymatic activity of PLA z, we set out to investigate the importance of the astrocyte in eicosanoid production. By employing immunohistochemical techniques, we examined whether astroglial ceils or other cell types in the human brain express cPLA 2, the receptor-regulated cytosolic form of PLA 2.

2. Materials and methods Immunocytochemlstry was performed on paraffin sections from h u m a n occipital cortex Four neurologically normal cases were evaluated. Three of the four cases were male and the ages ranged from 36-80 years (mean 66 years) Tissue was fixed only briefly (60-90 mln) and transferred to Tris-buffered sahne for several days prior to embeddmg. Routinely processed tissue (immersed m 10% buffered formahn for weeks) was not suitable for our investigations. The monoclonal antibody M12 was raised against purxfied cPLA 2 from U937 cells as reported previously [37]. Ascltes was produced in B A L B / c mice and antibodies were affinity purified using Protein A Fast Flow resin (Pharmacia, Piscataway, NJ) M12 recognizes the native form of cPLA 2 and is also a neutrahzing antibody [37]. H u m a n cPLA 2 was purified to homogeneity as previously described [37] A rabbit antiserum to glial fibrillary acidic protein (GFAP; Biogenex Labs, San Ramon, CA) was used to label astrocytes. Immunostainlng of tissue sections ( 1 0 / z m ) utd~zed conventional lmmunoperoxadase techmques and employed the avidm-biotin peroxldase system (ABC, Vector Labs, Burlingame, CA) Briefly, tissue sections were deparaffinlzed, blocked with serum and incubated m 1% H 2 0 2 in methanol (30 rain, 22°C) for quenching of endogenous peroxldase activity. Primary antibodies were layered over tissues and

were incubated overnight at 4°C For cPLA 2 Iocdhzatlon, 0 1 m g / m l M12 was used A n t l - G F A P was obtained as predduted antlsera Sections were stained w~tb the ~mmunoperoxidase system according to the recommendations of the manufacturer The reaction product was visualized by development with 3,3'-dmmlnobenzldme (DAB) and H~O~ Optimal staining of paraffin sections with M12 was observed following pretreatment of the sections with trypsin (0 3 m g / m l , 8 mln, 22°C) Sections were counterstalned with hematoxyhn prior to dehydration and covershppmg with permount Dual Iocahzatlon was carried out by sequential lmmunostalnlng An alkahne phosphatase-streptavldln system (Biogenex Labs ) using Fast Red as chromogen was used to locahze the rabbit antibody (GFAP) and mckel chloride-enhanced D A B (Vector Labs ) was used to detect the peroxldase-labeled mouse anti-cPLA 2 For localization of cPLA 2 in vitro, astrocytoma cells were grown in chambers on glass mlcroshdes, fixed for 30 m m m 4% buffered paraformaldehyde, washed in PBS and stained with M12 ( + 0 lC,k Triton X-100) by lmmunofluorescence techniques. Two h u m a n astrocytoma cell lines were evaluated UC-11 M G [40, gift from Dr C L Johnson] and U-373 M G [52, ATCC] For localization of the astrocytoma cells, staining was carried out by incubating cells for 1 h in biotlnylated anti-mouse IgG and for 1 h in avldln-Texas red (5 ~xg/ml, Vector Labs ) Cells were covershpped m 70% glycerol and photographed on a Nlkon mlcrophot equipped with eplfluorescence and Nomarskl optics Positive and negative controls were conducted in parallel with cPLA 2 stained sections or cells In each experiment AntI-GFAP staining served as positive control. Negative controls included staining cells or tissue sections with omission of the primary antibody In addition, sections were stained with anti-cPLA2 antibody that had been preadsorbed with an excess of purified cPLA 2 m R N A levels were measured by a reverse transcrlptase-polymerase chain reaction (RT-PCR) method using Southern hybridization [57] Total R N A from h u m a n tissues and cell hnes was extracted by the A G P C mlcroprocedure [12]. Total R N A was reverse transcribed using the Perkln-Elmer Cetus R T / P C R kit according to the manufacturer's recommendations. Following c D N A synthesis, each sample was sermlly diluted and amplified vm PCR for 25 cycles. Oligonucleotides used for P C R amplification were synthesized on an Applied Blosystems D N A synthesizer. Following synthesis, the ohgonucleotldes were gel purified and resuspended in distilled water Two 18-nucleotlde primers amplified a 938 bp c D N A fragment, bases 1178-2115 of the h u m a n cPLA 2 c D N A [58]. Subsequent to amphficatlon, the samples were fractlonated using agarose gel electrophoresls and transferred to a Zeta-Probe m e m b r a n e (B~o-Rad, CA) via alkaline blotting. The blots were probed with a 32p_labeled D N A fragment corresponding to bases 1708-2065 of cPLA 2 cDNA. The probe used for hybrl&zatlon was completely internal to the sites used for amplification to ensure the authenticity of the amplified bands To assay for cPLA 2 activity, astrocytoma cells were collected by scraping with a rubber policeman, suspended at 107/ml in buffer (150 m M NaC1, 1 0 0 / z m leupeptin, 1 m M PMSF, 50 m M Hepes, pH 7.5) and lysed by somcation with a probe sonlcator The assay was performed using sonicated liposomes containing 1-palmltoyl-2[14C]arachidonoyl-sn-glycero-3-phosphochohne as substrate as prewously described [37]

Fig. 1 Specific cPLA 2 immunoreactlvlty is detected in h u m a n cortical gray matter. A: cPLA 2 immunoperoxtdase staining in the gray matter from occipital cortex. N u m e r o u s immunoreactive structures are scattered throughout the gray matter (arrows) The purple structures distributed throughout the micrograph are nuclei stained with hematoxylin as counterstain. B: absence of cPLA 2 lmmunoreactlvity in a section stained with antibody preadsorbed with purified cPLA 2. Only hematoxylin stained nuclei are observed C. high power magnification of cPLA 2 staining in the gray matter reveals punctate, globular deposits within presumptive cell bodies (arrows) Scale bar = 150/~m (A,B), 16/.~m (C)

D.T. Stephenson et al /Bram

Research 637 (1994) 97-105

99

100

D, T Stephenson et al /Bram Research 037 (1994) 97-105

3. Results

Table 1 Phosphohpase A 2

Immunocytochemical localization of cPLA 2 in the human occipital cortex revealed a distinctive pattern of staining. Numerous structures were observed scattered throughout the gray matter (Fig. 1A). The average size of immunoreaetive deposits was approximately 15-25 /zm. Often there were greater numbers of stained deposits in the subpial lamina I; in other gray matter regions, no laminar distribution was observed. Tissue sections from all four human cases demonstrated this staining pattern, although some cases had greater numbers of immunoreactive deposits than others. This staining was identified to be specifically associated with cPLA z because staining was almost completely abolished when performing localization with anti-cPLA 2 antibody that had been preadsorbed with purified cPLA z (Fig. 1B). High magnification of cPLA 2 staining showed punctate, globular-like deposits of stain that appeared to be localized within the cytoplasm of particular cells (Fig. 1C). Structures identified as glycogen deposits (corpora amylacea) were also stained (not shown); however, this staining was not found to be specific for c P L A 2 since immunoadsorbed sections showed an equal n u m b e r of these deposits with the same staining intensity as in nonadsorbed sections. No cellular immunostaining for cPLA 2 was detected in the white matter. Based upon the size of the stained cells, the immunoreactive cell type was presumed to be glial yet the identity of the cell type could not be determined on the basis of cPLA 2 staining. To determine whether the cPLA 2 staining was associated with astrocytes, sections were counterstained with antisera to GFAP. As shown in Fig. 2A, cPLA 2 staining colocalized with G F A P immunoreactive cells. Astrocyte processes were predominantly stained with anti-GFAP while the cell body possessed cPLA 2 immunoreactivity. In all cases evaluated, only a subset of protoplasmic astrocytes were found to stain positively for cPLA 2. The percentage of immunoreactive astrocytes could not be determined definitively due to the lack of robust G F A P staining in protoplasmic astrocytes. However, we estimate that cPLA 2 positive astrocytes comprise 10-30% of total gray matter astrocytes in normal human occipital cortex. Virtually all cPLA 2 immunoreactive structures were co-localized with anti-GFAP. G F A P immunoreactive astrocytes within the white matter did not colocalize with c P L A 2 staining (Fig. 2B).

U-373 MG UC-1I MG

activityof cell line somcates cPLA 2 actwlty(pmol/mm/mg) ND * 67_+8

* Not significantly above background level vartatlon (_+14 pmol/ mm/mg) m three separate experiments To verify the presence of cPLA 2 within astrocytes and to develop an in vitro system for future investigations, we determined whether astrocytoma cell lines likewise express cPLA 2 immunoreactivity. Two human astrocytoma lines, U-373 M G and UC-11 M G cells, were tested for cPLA e immunoreactivity by fluorescence localization techniques. UC-11 M G cells stained prominantly for cPLA 2 (Fig. 3A), while the U-373 M G cells were immunonegative (Fig. 3B). Interestingly, the staining pattern observed in the UC-11 M G cells was somewhat similar to that observed in human cortical astrocytes in situ. An uneven, globular-like staining was associated with the cell body (Fig. 3A, arrows) and was not localized to processes emanating from the cells. Virtually all UC-11 M G cells were immunoreactive although some cells were stained more intensely than others. The difference in immunoreactivity between these two astrocytoma cell lines is consistent with measurements made on their cPLA e enzyme activity and m R N A levels, cPLA 2 enzymatic activity was detected in homogenates from UC-11 M G but not U-373 M G cells (Table 1). R T - P C R analyses showed abundant expression of cPLA 2 m R N A in human hippocampus and UC-11 M G cells but not in U-373 M G cells (Fig. 4). Thus by three criteria, there are great differences in expression of cPLA 2 by these two human astrocytoma cell lines.

4. Discussion The results of this study demonstrate that a subset of human brain gray matter astrocytes contain cPLA 2These cells appear to be an abundant source of the enzyme in the human brain. The lack of immunoreactivity in neurons, microglia, oligodendrocytes or vascular endothelial cells indicates that either these cells do not contain cPLA 2 or have cPLA 2 in amounts below the limit of detection. The cPLA 2 imrnunoreactivity that was observed in this study was deemed specific for the following reasons: firstly, the immunoreactivity

Fig 2 cPLA2 lmmunoreactivityis localized to astrocytes A. m the gray matter of human ocopltal cortex, GFAP and cPLA 2 lmmunoreactmties are co-locahzed within the same cell. c P L A 2 was locahzed with Ni-enhanced DAB as chromogen for peroradase staining, producing a brown/black stare (arrow); anti-GFAP was locahzed with alkahne-phosphatase/Fast red detection system producing a red reaction product (open arrow). Asterisks delineate blood vessels B. In the white matter, GFAP tmmunoreactlVlty(open arrow) is observed m the absence of cPLA2 immunoreactmty. Scale bar = 16/~m.

D T. Stephenson et al. / Brarn Research 637 (1994) 97-105

101

1(}2

D T Stephemon et al /Brain Research 637 (1994) 97-105

D T. Stephenson et al. ~Brain Research 637 (1994) 97-105

Fig. 4. Expression of cPLA 2 mRNA in human brain and astrocytoma cell lines. Total RNA from each sample was reversed transcribed and amplified via PCR for 25 rounds. For each sample three concentrations of cDNA (100, 50 and 25 ng) were amplified and are illustrated in the three replicate lanes shown for each sample. The samples were fractlonated, blotted and probed with a cPLA2 restriction fragment internal to the sites used for amplification The UC-11 MG sample was from cells stimulated with recombinant human IL-1/3, however, unstlmulated cells produced similar levels of cPLA 2 mRNA (data not shown). U-373 MG cells either unlnduced (shown here) or induced with IL-1/3 (data not shown) do not make detectable levels of cPLA 2 mRNA.

could be virtually eliminated by preadsorption of the antibody with purified cPLA 2. Secondly, the astrocytoma cells U-373 MG and UC-11 MG differed in their expression of cPLA 2 immunoreactivity and the presence of immunoreactivity correlated with cPLA 2 enzyme activity and mRNA levels. Finally, when immunocytochemistry was performed utilizing the monoclonal antibody M3-1 that recognizes the denatured form of cPLA 2 [37], no detectable staining was observed (unpublished observations). The difference between cortical gray matter protoplasmic astrocytes and white matter fibrous astrocytes was striking. Although GFAP immunoreactivity was present in both gray matter and white matter astrocytes, cPLA 2 immunoreactivity was absent in white matter astrocytes. This finding suggests that the functional roles of these two populations of astrocytes may be different. There is increasing evidence that astrocytes are a heterogeneous cell type within the CNS [for review see ref. 64]. Our study suggests that gray matter astrocytes in the cortex may have the ability to abundantly produce products derived from the arachidonic acid cascade, while white matter astrocytes may not

103

possess this property. The differential expression of cPLA 2 by two different astrocytoma cell lines suggests that these two lines may be derived from different populations of astrocytes. An association between PEA 2 and astrocytes has been observed previously [7]. A secreted form of PLA 2 (sPLA 2) has been reported to be expressed in primary cultures of rat astrocytes after stimulation with inflammatory agents [49]. In that study, levels of the sPLA 2 were undetectable in unstimulated cells; cPLA 2 expression was not investigated. Recently, Sun et al. demonstrated that ATP depletion in the human astrocytoma UC-11 MG leads to breakdown of membrane phospholipids [60]. They also reported that addition of PLA 2 inhibitors reduced the severity of cell injury due to energy depletion and suggested that PLA 2 activation plays an important role in membrane injury to astrocytes. Other studies suggest that PLA 2 may be activated by receptor-mediated events in astrocytes under various conditions. In fact, arachidonic acid release was induced in rat cerebral cortical astrocytes in primary culture by bradykinin [8], substance P [41] and ATP [23]. Collectively, these studies support the notion that astrocytes are a critical cell type for PLA 2 activity in the brain. Although neurons are thought to be a source of arachidonic acid and its metabolites, astrocytes may also be an important or even more abundant source of these substances. The observations of this study are consistent with reports that astrocytes can produce eicosanoids [22,46] and also have receptors for eicosanoids [30,47]. In addition, multiple neurotransmitter and neuropeptide receptors are located on astrocytes [reviewed by 3,28,34,45,63]; therefore, it is likely that activation of certain of these receptors, especially those associated with calcium entry, will lead to activation of cPLA 2. In support of this view are the findings that arachidonic acid and its metabolites are released upon stimulation of certain neurotransmitter receptors including NMDA [16,51], GABA [5], m5 muscarinic [19], alpha-1 adrenergic [31] and serotonin [20] receptors. In addition, it has been suggested that arachidonic acid is involved in long term potentiation [2,65]. Some evidence also exists that astrocytes play a role in neurotransmission [reviewed by 25]. The results of the present study, which demonstrate substantial cPLA 2 immunoreactivity in gray matter astrocytes pro-

Fig. 3. Human astrocytoma cell lines differentially express cPLA 2 lmmunoreactlvlty. A: lmmunofluorescent staining of cPLA 2 m UC-11 MG cells. Staining is localized intraceUularly in a globular-like staining pattern (arrows). The mlcrograph shows two prominently stained cells and one presumptive mitotic cell with moderate staining intensity. B: U-373 MG astrocytomas are immunonegative. UC-11 MG cells were grown at lower density than the U-373 MG cells, which were allowed to grow to confluency One cell in B appears stained with a greater intensity than the other cells in this field of confluent cells; however, this cell is not stained above backround levels (compare with C). C: negatwe control showing backround levels of staining in UC-11 MG cells stained with omission of the primary antibody. All cells were stained m parallel and photographed at the same magmflcat~on, time and exposure settings. Scale bar = 32/zm.

104

D T Stephenson et al /Bram Research 637 (1994) 97-105

vides strong evidence that astrocytes m the human brain can participate in the generation of phospholipid-derived products and may have an important, heretofore unrecognized role in brain function includmg neurotransmission. Furthermore, if astrocytlc cPLA 2 were involved in synaptic function, cPLA 2 might only need to be present in the gray matter, the locus of the somatodendritic domain. This view is consistent with our observations. The presence of significant amounts of cPLA 2 in GFAP-containing astrocytes is suggestive of another phystological role of the astrocyte. Astrocytes become reactive to almost any form of brain insult. For example, astrocytic reactivity is observed after CNS trauma, infections, tumors, inflammation a n d / o r ischemia [for review see refs. 18, 27]. Increases in PLA 2 activity and free fatty acid levels have been observed in several forms of brain injury. Evidence exists for PLA: activation after brain damage induced by free radicals [9], complete and severe incomplete cerebral ischemia [17,29,32,53,54], and cold-injury [10]. Astrocytes are known to produce several inflammatory cytokines [for review see ref. 4], as well as growth factors [for review see ref. 44] and products of phospholipid metabolism [for review see ref. 42]. The astrocytes containing cPLA z that we observed in the present study may be well-poised to react to inflammatory stimuli. Future studies in our laboratory will investigate alterations in cPLA 2 of the astrocyte under different inflammatory conditions and its potential role in neurotransmission. Acknowledgements The authors thank Dr C.L Johnson, University of Cincinnati for providing the UC-11 MG astrocytoma cells

5. References [1] Axelrod, J., Burch, R.M. and Jelsema, C.L, Receptor-mediated actwatlon of phosphohpase A2 via GTP-bindmg proteins arach~donlc acid and its metabohtes as second messengers, Trends Neurosct, 11 (1988) 117-123 [2] Barbour, B, Szatkowskl, M., Ingledew, N and Attwell, D , Arachldonic acid reduces a prolonged inhibition of glutamate uptake into ghal cells, Nature, 34 (1989) 918-920 [3] Barres, BA., New roles for glia, J Neuroscz, 11 (1991) 36853694. [4] Benvemste, E N.. Inflammatory cytokmes within the central nervous system' sources, function, and mechanism of action. Am J Phystol, 263 (Cell Physlol, 32) (1992) C1-C16. [5] Btrkle, D L and Bazan, N.G., Effect of blcuculhne-lnduced status epllepticus on prostaglandlns and hydroxyeicosatetraenoic acids in rat brain subcellular fractions, J Neurochem, 48 (1987) 1768-1778 [6] Brenner, T., Boneh, A , Shohaml, E., Abramsky, O and Weldenfeld, J , GlucocortlCOld regulation of elcosanold production by ghal cells under basal and shmulated conditions, J Neurotmmunol, 40 (1992) 273-280. [7] Brenner, T., Yedgar, S, Weldenfeld, J and Abramsky, O , Phospholipase A 2 activity in cultured ghal cells. Correlation with appearance of myehn markers and effects of steroid hormones, Ann N Y A e a d Scl, 540 (1988) 403-404

[8] Burch, R M and Knlss, D A , Modulation of receptor-mediated signal transductlon by dlacylglycerol mlmet~cs m astrocytes, Cell Mol Neurobtol, 8 (1988) 251-257 [9] Chan, P H , Yurko, M. and Flshman. R A , Phospholipld degradahon and cellular edema reduced by free radicals in brain cortical slices, J Neurochem, 38 (1982) 525-531 [10] Chan, P H , Longar, S and Flshman, R A,, Phosphohp~d degradation and edema development m cold-injured rat brain, Brain Res, 277 (1983) 329-337 [11] Chlu, E K.Y. and Richardson, J S., Behavloural and neurochemical aspects of prostaglandins m brain function, Gen Pharmacol, 16 (1985) 163-175 [12] Chomczynskl, P and Sacchl, N., Single step method of RNA isolation by acid guanldinlum thiocyanate-phenol-chloroform extraction, Anal Btochem, 162 (1987) 156-159 [13] Clark, J.D, Lln, L L., Dnz, R W , Ramesha, C S, Sultzman, L A , Lln, A Y, Mllona, N and Knopf, J L, A novel arachldonlc acld-selectwe cytosohc PLA 2 contains a Ca(2 +)-dependent translocahon domain with homology to PKC and GAP, Cell, 65 (1991) 1043-1051 [14] Clark, J D., Mllona, N and Knopf, J L, Purification of a ll0kilodalton cytosohc phospholipase A2 from the human monoCytlc cell hne U937, Proc Natl Acad Sct USA, 87 (1990) 7708-7712. [15] Dam, J W, Chernjavsky, A and Smith, S.J, Neuronal activity triggers calcium waves In hlppocampal astrocyte networks, Neuron, 8 (1992) 420-440 [16] Dumuls, A . Sebben, M., Haynes, L Pin, J -P and J Bockaert, NMDA receptors activate the arachldonlc acid cascade system in strlatal neurons, Nature, 336 (1988) 68-70 [17] Edgar, A.D, Strosznajder, J. and Horrocks, L A , Activation of ethanolamlne phosphohpase A2 in brain during lschemla, J Neurochem,, 39 (1982) 1111-1116 [18] Eng, L F., Ye, A C H and Lee, Y L. Astrocytlc response to injury, Prog Brain Res, 94 (1992) 353-365, [19] Felder, C.C, Dieter, P, KinseUa, J , Tamura, K., Kanterman, R Y and Axelrod, J , A transfected m5 muscarlmc acetylchohne receptor stimulates phosphohpase A2 by inducing both calcium reflux and activation of protein kinase C, J Pharmacol Exp Ther, 255 (1990) 1140-1147 [20] Felder, C.C, Kanterman, R Y., Ma, A L and Axelrod, J , Serotonln stimulates phospholipase A 2 In the release of arachidomc acid m hlppocampal neurons by a type 2 serotonm receptor that is independent of mosltol phosphohpld hydrolysis, Proc Natl Acad Sct USA, 87 (1090) 2187-2191 [21] Flerz, W and Fontana, A , The role of astrocytes m the interaction between the immune and nervous system In S. Fedoroff and A Vernadakls (Eds.), Astrocytes, Academic Press, New York, Vol 3, 1986, pp 203-230. [22] Fontana, A., Krlstensen, F, Dubs, R , Gemsa, D. and Weber, E., Production of prostaglandm E and an lnterleukm-1 hke factor by cultured astrocytes and C 6 ghoma cells, J Immunol, 129 (1982) 2413-2419 [23] Geblcke-Haerter, P J , Wurster, S, Schobert, A and Herttmg, G., P-2 purmoceptor induced prostaglandm synthesis in primary rat astrocyte cultures, Naunyn Schmtedeberg's Arch Pharmacol, 338 (1988) 704-707 [24] Gross, R W, Myocardial phosphohpases A 2 and their membrane substrates, Trends Cardtovaseular Med, 2 (1992) 115-121 [25] Hansson, E. and Ronnback, L , Astrocytes m neurotransmlsslon, Cell and Mol Btol , 36 (1990) 487-496. [26] Hartung, H , Helnlnger, K, Schafer, B and Toyka, K, Substance P and astrocytes stimulation of the cyelooxygenase pathway of arachldonlc acid metabolism, FASEB J , 2 (1988) 48-51 [27] Hatten, M E , Llem, R K.H, Shelanskl, M L and Mason, C A., Astrogha m CNS Injury, Gha, 4 (1991) 233-243 [28] Hosll, E and Hosh, L , Receptors for neurotransmltters on

D T. Stephenson et al ~Brain Research 637 (1994) 97-105 astrocytes in the mammalian central nervous system, Prog. Neurobml., 40 (1993) 477-506. [29] Ikeda, M., Yoshlda, S, Busto, R., Santiso, M., Martinez, E. and Ginsberg, M.D., Cerebral phosphoinosltlde and energy metabohsm during and after msuhn-induced hypoglycemia, J Neurochem., 49 (1987) 100-106 [30] Ito, S., Sugama, K., Inagaki, N., Fukui, H , Giles, H , Wada, H. and Hayalshl, O., Type-1 and type-2 astrocytes are distract targets for prostaglandins D2, E2 and F2 alpha, Gha, 6 (1992) 67-74 [31] Kanterman, R.Y., Felder, C.C., Brenneman, D.E., Ma, A.L, Fitzgerald, S. and Axelrod, J., Alphal-adrenergic receptor mediates arachldonic acid release m spinal cord neurons independent of lnoslto1 phosphohpld turnover, J Neurochem 54 (1990) 1225-1232. [32] Kempskl, O , Shohaml, E , von Lubitz, D., Hallenbeck, J M, and Feuersteln, G., Postischemlc production of elcosanolds in gerbil brain, Stroke, 18 (1987) 111-119. [33] Kim, D and Clapham, D E., Potassmm channels in cardiac cells activated by arachldonic acid and phosphohplds, Science, 244 (1989) 1174-1176 [34] Kimelberg, H.K, Ghal Cell Receptors, Raven Press, New York, 1988 [35] KImura, H , Okamoto, K. and Sakal, Y, Modulatory effects prostaglandm D2, E2 and F2alpha on the postsynaptlc actions of inhibitory and excitatory amino acids in cerebellar Purkmje cell dendrites in vitro, Brain Res., 330 (1985) 235-244 [36] Kramer, R.M., Hession, C., Johansen, B, Hayes, G., McGray, P, Chow, E.P., Tlzard, R. and Pepmsky, R.B., Structure and properties of a human non-pancreatic phospholipase A2, J. Btol Chem., 264 (1990) 5768-5775. [37] Kramer, R.M., Roberts, E.F., Manetta, J. and Putnam, J.E., The Ca++-sensltwe cytosolic phosphohpase A 2 is a 100-kDa protein m human monoblast U937 cells, J B~ol Chem, 266 (1991) 5268-5272. [38] Kramer, R M, Roberts, E.F., Manetta, J.V., Sportsman, J.R. and Jakubowskl, J A., Ca 2÷ sensitive cytosolic phospholipase A 2 (cPLA 2) in human platelets, J Lipid Me&ators, 6 (1993) 206-216 [39] Kramer, R M, Structure, function and regulation of mammahan phosphohpases A 2 In B L Brown and P.R M. Dobson (Eds), Advances in Second Messenger and Phosphoproteln Research, Vol 28,, Raven Press, New York, 1993. [40] Liwmcz, B H , Archer, G , Soukay, S W and Llwnlcz, B H , Continuous human ghoma-derwed cell lines UC-11MG and UC-302MG, J. Neurooncol, 3 (1986) 373-385. [41] Marriott| D.R., Wllkm, G.P and Wood, J.N., Substance P-reduced release of prostglandms from astrocytes, regional speoalization and correlation with phosphomosltol metabolism, J. Neurochem., 56 (1991) 259-265. [42] Martin, D.L, Synthesis and release of neuroactive substances by glial cells, Gha, 5 (1992) 81-94. [43] Mead, J F., Dhopeshwarkar, G A and Elepano, M.G., Transformations of the polyunsaturated fatty acids in the bram. In Functions and Bmsynthests of Llplds, Plenum Press, New York, 1977, pp 318-328 [44] Merrill, J E , Macrogha: neural cells responswe to lymphoklnes and growth factors, Immunol Today, 5 (1987) 146-150 [45] Murphy, S. and Pearce, B., Functional receptors for neurotransmltters on astroglial cells, Neuroscwnce, 22 (1987) 381-394. [46] Murphy, S, Pearce, B., Jeremy, J. and Dandona, P., Astrocytes as elcosanold producing cells, Glia, 1 (1988) 241-245. [47] Nakahata, N, Ishlmoto, H , Masatake, K., Ohmon, K., Takahashi, A. and Nakamshl, H , The presence of thromboxane A 2 receptors in cultured astrocytes from rabbit brain, Brain Res, 583 (1992) 100-104. [48] Ohmlchl, M., Hirota, K., Kolki, K, Kadowakl, K., Miyake, A., Klyama, H , Tohyama, M. and Tanizawa, O., Involvement of

105

extracellular calcium and arachidonate in 3H dopamlne release from rat tuberomfundibular neurons, Neuroendocrmology, 50 (1989) 481-487 [49] Oka, S. and Anta, H., Inflammatory factors stimulate expression of group II phosphohpase A2 in rat cultured astrocytes, J. Btol Chem., 266 (1991) 9956-9960. [50] Ordway, R W, Walsh, J.V J and Singer, J.J., Arachldomc acid and other fatty acids directly actwate potassium channels in smooth muscle cells, Science, 244 (1989) 1176-1179 [51] Pellerm, L. and Wolfe, L.S., Release of arachldomc acid by NMDA-receptor activation in the rat hlppocampus, Neurochem Res. 16 (1991) 983-989 [52] Ponten, J and Maclntyre, E H , Long term culture of normal and neoplastic human gila, Acta Pathol Mwrobtol Scand, 74 (1968) 465-486. [53] Rehncrona, S, Westberg, E , Akesson, B and Slesjo, B.K., Brain cortical fatty acids and phosphohplds during and followmg complete and incomplete lschemla, J Neurochem, 38 (1982) 84-93. [54] Rordorf, G , Uemura, Y. and Bonventre, J V, Characterization of phosphohpase A2 activity m gerbd brain" enhanced activities of cytosohc, mitochondnal and mlcrosomal forms after ischemia and reperfuslon, J Neurosct., 11 (1991) 1829-1836. [55] Schweltzer, P, Madamba, S and Slgglns, G R , Arachldonlc aod metabohtes as mediators of somatostatln-induced increase of neuronal M-current, Nature, 346 (1990) 464-467. [56] Sellhamer, J.J., Pruzanski, W., Vadas, P, Plant, S, Miller, J A , Kloss, J and Johnson, L K., Cloning and recombinant expression of phosphohpase A 2 present m rheumatoid arthritic synovial fluid, J Blol Chem, 264 (1990) 5335-5338 [57] Sharp, J D. and White, D L, Cytosollc PLA 2 mRNA levels and potential for transcriptional regulation, J Lipid Mediators, in press [58] Sharp, J.D, White, D L, Chiou, X G., Goodson, T , Gamboa, G.C., McClure, D., Burgett, S., HoskIns, J., Skatrud, P.L., Sportsman, J.R., Becker, G.W., Kang, L.H, Roberts, E.F and Kramer, R M, Molecular cloning and expression of human Ca2+-sensltive cytosohc phospholipase A2, J Btol Chem, 266 (1991) 14850-14853. [59] Slesjo, B K, Ingyar, M and Westerberg, E , The influence of blcuculhne-induced seizures on free fatty acid concentrations in cerebral cortex, hlppocampus and cerebellum, J Neurochem, 39 (1982) 796-802 [60] Sun, F.F., Fleming, W.E. and Taylor, B M, Degradation of membrane phosphohpids in the cultured human astroghal cell hne UC-11 MG during ATP depletion, Blochem Pharmacol, 45 (1993) 1149-1155. [61] Tence, M, Cordier, J., Glowlnskl, J and Premont, J, Endothelin-evoked release of arachidonlc aod from mouse astrocytes in primary culture, Eur J Neurosct., 4 (1992) 993-999 [62] Wieloch, T., Hams, R.J, Symon, L and Slesjo, B K, Influence of severe hypoglycemia on brain extracellular calcium and potassium actwitles, energy, and phosphohpld metabolism, J Neurochem, 43 (1984) 160-168. [63] Wllkln, G.P. and Cholewinski, A J , Peptlde receptors in astrocytes. In H K Kamelberg (Ed.), Ghal Cell Receptors, Raven Press, New York, 1988, pp. 223-241. [64] Wllkm, G P, Marriott, D.M. and Cholewinskl, A.J., Astrocyte heterogeneity, Trends in Neurosci, 13 (1990) 43-46 [65] Wilhams, J H., Errmgton, M L., Lynch, M.A. and Bliss, T.V P, Arachldonic acid Induces a long-term activity-dependent enhancement of synaptic transmission in the hxppocampus, Nature, 341 (1989) 739-742 [66] Wltter, B. and Kanfer, J.N., Hydrolysis of endogenous phosphohplds by rat brain microsomes, J Neurochem., 44 (1985) 155-162 [67] Wolfe, L.S., Eicosanolds prostaglandlns, thromboxanes, leukotnenes, and other derivatives of carbon-20 unsaturated fatty acids, J Neurochem, 38 (1982) 1-14