9 1995 by Humana Press Inc. All rights of any nature whatsoeverreserved. 1044-7393/95/2501--0001 $07.40

Stimulation of Group II Phospholipase A2 mRNA Expression and Release in an Immortalized Astrocyte Cell Line (DITNC) by LPS, TNF~, and IL-I~ Interactive Effects WEI TONG, ZHONG-YI HU, AND GRACE Y. SuN* Biochemistry Department, M 121 Medical Sciences Building, University of Missouri, Columbia, MO 6 5 2 1 2 Received January 17, 1995; Accepted March 9, 1995

ABSTRACT Astrocytes are immunoactive cells in brain and have been implicated in the defense mechanism in response to external injury. Previous studies using cultured glial ceils indicated the ability of astrocytes to respond to bacteria endotoxin and cytokines, resulting in the release of phospholipase A2. In this study, we examined the interactive effects of lipopolysaccharides (LPS), interleukin 1/3 (IL-I~) and tumor necrosis factor (TNFc~) to stimulate phospholipase A2 (PLA2) in an immortalized astrocyte cell line (DITNC) with many properties of type I astrocytes. Northern blot analysis using oligonucleotide probes derived from the cDNA encoding the rat spleen group II PLA2 indicated the ability of DITNC ceils to respond to all three factors in the induction of gene expression and the release of PLA2. After an initial lag time of 2 h, PLA2 release was proportional to time, reaching a plateau by 12 h. This event occurred at a time period preceding any signs of cell death. Cycloheximide at 1.25 /d~ completely inhibited cytokine-induced PLA2 release. When suboptimal amounts of TNFa were added to the DITNC culture together with IL-1/~ or LPS, a synergistic increase in the induction of PLA2 release could be observed. On the other hand, combination of IL-I~ and LPS resulted only in an additive increase in PLA2 release. Antibodies to IL-1/~ and TNFo~ completely neutralized *Author to whom all correspondence and reprint requests should be addressed.

Molecularand ChemicalNeuropathology

1

Vol.25, 1995

2

Tong, Hu, and Sun the effects of these two agents on PLA2 release. However, neither antibody was able to inhibit the PLAz release induced by LPS, suggesting that the effect of LPS was not complicated by the release of IL-lfl or TNFo~. Taken together, results show that the immortalized astrocyte cell line (DITNC) can be used for studies to elucidate the molecular mechanism underlying the cytokine signaling cascade and subsequent induction of PLA2 synthesis. Index Entries: Astrocytes; injury; glial cells; endotoxin; cytokines; phospholipase A2; lipopolysaccharides; interleukin lfi; tumor necrosis factor; signaling cascade. Abbreviations: PC, phosphatidylcholine; PE, phosphatidylethanolamine; PS, phosphatidylserine; PI, phosphatidylinositol; DG, diacylglycerol; FFA, free fatty acid; AA, arachidonic acid; PLA2, phospholipase A2; PLC, phospholipase C; FBS, fetal bovine serum.

INTRODUCTION

Besides supplying nutrients to the neurons, astrocytes are immunologically competent and play an important role in the defense mechanism in the nervous system (Kimelberg and Norenberg, 1989). There is evidence that on exposure of astrocytes to bacterial endotoxins (lipopolysaccharides, LPS), proinflammatory cytokines, and neurotrophic viruses, these cells are capable of releasing cytotoxic factors which may have autocrine and paracrine functions (Lieberman et al., 1989; Sawada et al., 1989; Chung et al., 1990). In cultured glial cells, these factors were shown to cause a proliferative response (Selmaj et al., 1990) as well as alteration in cellular metabolism (Burch and Tiffany, 1989; Merrill, 1991). Consequently, cytokine responses in astrocytes have been implicated in the etiology and pathogenesis of immune-mediated diseases in brain including multiple sclerosis, AIDS, and bacterial meningitis (Fontana et al., 1987). A recognized event associated with the inflammatory effects of cytokines is their ability to stimulate the phospholipase A2 (PLA2) cascade leading to synthesis and release of prostaglandins (Chung et al., 1986; Oka and Arita, 1991; Shinohara et al., 1992). Recent studies further reveal the presence of several types of PLA2 in mammalian cells; among them, group II PLA2 is a secretory type (Mayer and Marshall, 1993). Cultured glial cells are known to contain a heterogeneous mixture of glial cells including microglial cells which are also responsive to cytokines (Sawada et al., 1989; Chao et al., 1992; Ganter et al., 1992). Depending on the purity of the culture, studies to examine molecular mechanisms underlying the cytokine induction pathway are limited in scope. Recently, Radany et al. (1992) successfully developed several immortalized astrocyte cell lines from rat diencephalon and these cells exhibit many of the characteristic properties of type I astrocytes. In the present study, we examined the effects of LPS, IL-1t3 and TNFo~ on the mRNA expression and release of Molecular and Chemical Neuropathology

Vol. 25, 1995

Cytokine Effects on Phospholipase A2 in Astrocytes

3

PLA2 in one of the immortalized astrocyte cell lines (DITNC) provided by these investigators (Radany et al., 1992).

MATERIALS AND METHODS Materials An immortalized astrocyte cell line (DITNC) from rat diencephanon was provided by Dr. C. F. Deschepper (Clinical Research Institute of Montreal, Montreal, Canada). This cell line has been shown to exhibit many of the characteristic properties of type I astrocytes (Randany et al., 1992). Cultured rat pheochromocytoma (PC-12) cells used for preparation of phospholipid substrates were originally from American Type Culture Collection (ATCC, Rockville, MD) and were provided by Dr. Albert Sun (Pharmacology Department, University of Missouri, Columbia, MO). Polyclonal antibody to glial fibriliary acidic protein (GFAP) was a gift from Dr. L. F. Eng (Pathology Department, Stanford University, Palo Alto, CA). TNFor (recombinant human, 2 x 107 U/mg), LPS (from E. coli), PLA2 (from Naja naja naja), protease inhibitors, and cyclohexamide were purchased from Sigma (St. Louis, MO). Initially, recombinant human ILlfl (specific activity > 1 x 107 U/mg, 10,000 U/mL) was purchased from Mallinckrodt (Chesterfield, MO). In later studies, recombinant murine TNFor and IL-lfl (specific activity 1 U/5-10 pg) were obtained from R & D Systems (Minneapolis, MN). Antihuman IL-lfl rabbit polyclonal antibody (1 mg/mL IgG solution in PBS) was obtained from GIBCO (Gaithersburg, MD). Antimouse TNFoc goat polyclonal antibody was obtained from R & D Systems. Two 48 oligomers (sequences 70-118 and 391-439) derived from the cDNA of rat spleen group II PLA2 (Ishizaki et al., 1989) were synthesized by the DNA Core Facilities, University of Missouri (Columbia, MO). An oligonucleotide probe for fl-actin corresponding to position 135 to 150 of the cDNA sequence (Nudel et al., 1983) was obtained from Dr. R. T. Zoeller (University of Massachusetts, Amherst, MA). DNA 3'-end-labeling kit was purchased from Beohringer Mannheim (Indianapolis, IN). QuikHYB solution was purchased from Stratagene (LaJolla, CA). Hybridization transfer membrane was from DuPont (Boston, MA). [14C]arachidonic acid (specific radioactivity, 55 mCi/mmol) was purchased from American Radiolabeled (St. Louis, MO). [oe-32p]-dATPwas purchased from New England Nuclear (Boston, MA). Cell culture reagents were obtained from Cell and Immunobiology Core Facilities (University of Missouri, Columbia, MO). Other chemicals were of the highest purity commercially available.

Cell Culture Glial cell culture was prepared from cerebral hemispheres of new-born Sprague Dawley rats according to the procedure described by Murphy (1990) with minor modifications. Briefly, brain tissue from rat pups was Molecular and Chemical Neuropathology

Vol. 25, 1995

4

Tong, Hu, and Sun

minced and dissociated in DMEM containing 0.25% trypsin. Cells were collected by centrifugation at 200g for 10 rain and were resuspended in DMEM medium with 10% FBS, 1% penicillin, 1% streptomycin, and 1% fungizone. Cell were placed on 100 mm poly-L-lysine-coated dishes and incubated with 10 mL of the earlier described medium at 37~ in SteriCult 200 Incubator (Forma Scientific, Marietta, OH) under a humidified atmosphere of 95% air and 5% CO2. Fresh culture medium was added twice per week and, after reaching confluency, cells were subcultured in 35-mm dishes with I mL of DMEM, 10% FBS, and antibiotics. In most instances, cells cultured after 2-4 passages were used for experiments. Microscopic examination showed the presence of a heterogeneous mixture of cells in this culture. Immunostaining showed that approximately 90% of cells in this type of culture were GFAP positive. The immortalized astrocytes (DITNC, passage 49) were initially cultured in T75 cm 2 flasks containing 20 mL DMEM medium with 10% FBS, 1% penicillin, 1% streptomycin, and 1% fungizone. Cells were maintained at 37~ in the incubator under a humidified atmosphere of 95% air and 5% CO2. Culture medium was changed every third day. After cells reached 90% confluency, they were briefly trypsinized and subcultured to 35-mm dishes with 1 rnL of the same medium. Cells in 35-ram dishes were used for experiments after reaching confluency. Microscopic examination showed a homogeneous population of cells in this culture. Immunostaining showed that all the cells were GFAP positive.

Exposure of Cells to LPS, TNFc~, and It,-I DITNC cells and cultured glial cells were washed with DMEM medium twice and incubated in a serum-free medium containing 1 mg/mL BSA and with or without LPS (10 and 50 #g/mL), TNFor (200 U/mL), and IL-lfl (100 U/mL) for 24 h or time specified. At the end of the incubation period, the culture medium was removed and centrifuged at 10,000g for 15 min to sediment cell debris. Aliquots of the supernatant were taken for assay of PLA2 activity. For studies to assay PLA2 activity in cell cytosol and membranes, cells were recovered from the culture dishes and homogenized in a buffer containing 0.32M sucrose with I mM EDTA and 50 mM Tris-HCl (pH 7.4). The cell homogenates were centrifuged at 100,000g for 60 min to yield the cytosol and the membrane fractions.

Assay of PLA2 Activity PLA2 activity was assayed using [14C]arachidonoyl-labeled phosphatidylcholine (PC), phosphatidylinositol (PI), phosphatidylserine (PS), and phosphatidylethanolamine (PE) as substrates. Labeled phospholipids were prepared by incubating [14C]arachidonic acid, ATP, CoASH, and the respective lyso-phospholipids with rat liver microsomes or membranes

Molecular and Chemical Neuropathology

Vol. 25, 1995

Cytokine Effects on Phospholipase A2 in Astrocytes

5

from pheochromocytoma (PC-12) cells according to the procedure described by Corbin and Sun (1978) with minor modifications. In initial experiments, rat liver microsomes were used because these membranes contained high levels of acyltransferase activity especially for transfer of labeled arachidonate to lyso-PC and PI. For preparation of liver microsomes, fresh liver (10 gm) was briefly sliced and homogenized in 10X buffer containing 0.32M sucrose, I mM EDTA, and 50 mM Tris-HC1 (pH 7.4). The liver homogenate was centrifuged at 12,000 rpm for 20 min to sediment the mitochondria and cell debris. The supernatant was further centrifuged at 100,000g for 60 min to obtain the microsomal pellet. In subsequent experiments, we found that it was more convenient to use membranes from PC-12 cells, especially for the synthesis of labeled PE. For isolation of cell membranes, PC-12 cells grown to cordluency in 100-mm dishes were harvested in phosphatebuffered saline (Cell and Immunology Core Facilities, University of Missouri, Columbia, MO). Cells were disrupted by sonication and a crude membrane preparation was obtained by centrifuging the suspension at 38,000 rpm for 15 min. The membrane pellets were resuspended in buffer and stored frozen in aliquots until use. For preparation of labeled phospholipid substrates, the incubation mixture contained cell membranes (100/~g protein), 5 ag LPE (1-palmitoyl glycerophosphoethanolamine), LPC (1-palmitoyl glycerophosphocholine) or LPI (extracted from brain), 0.2/~Ci [14C]arachidonic acid, 2.5 mM ATP, 10 mM MgCl2, 1 mM CoASH, and 0.32M sucrose with 50 mM Tris-HCl (pH 7.4). Labeled fatty acid was originally dissolved in chloroform and was evaporated in the test tube prior to adding other reagents. Incubation was carried out at 37~ for 30 min. Reactions were terminated by adding chloroform-methanol 2:1 (v/v) to the incubation mixture. Lipids in the organic layer were removed, evaporated to dryness, and applied to Whatman HP-K high performance thin-layer chromatography (HPTLC) silica gel plates (10 x 10 cm, 200 ~m thick) (Fisher, St. Louis, MO). Individual phospholipids were separated by a two-dimensional solvent system with a brief exposure to HC1 fumes between the first and second solvent system for hydrolysis of the alkenylether bonds of plasmalogens (Sun, 1988). This HPTLC system was important for the separation of diacyl-PE species from PE plasmalogens as well as PS from PI. Phospholipids were recovered from the HPTLC plates after spraying with 2',7'-dichlorofluorescein and viewing under a UV lamp. The amount of arachidonoyl group in each phospholipid was determined by converting the fatty acids to their methyl esters and then quantitative analysis of the methyl esters by gas-liquid chromatography (Sun, 1988). For assay of PLA2 activity, aliquots of the labeled phospholipids (approximately 8000 to 10,000 dpm) were dried and suspended in 100 #L of incubation buffer containing 0.25M sucrose, I mM EDTA, I mM EGTA, 6 mM CaC12, 1 mg/mL of BSA, and 50 mM Hepes (pH 7.4). The lipids

Molecular and Chemical Neuropalhology

Vol. 25, 1995

Tong, Hu, and Sun

6

were sonicated in the buffer to form a uniform suspension. Aliquots of the labeled suspension were added to test tubes containing the enzyme source. The enzyme reaction was initiated by adding 100/~g of protein from cell homogenates or cultured medium. Incubation was carried out at 37~ for 30 rain. The reaction was terminated by adding 2 mL of chloroform/methanol (4:1, v/v) and 0.3 mL of water. The FFA released from the phospholipids were separated using a solvent system containing hexaneethyl ether-acetic acid (85:15:2, by vol). After exposure of the HPTLC plates to iodine vapors, lipid bands corresponding to phospholipids (origin), diacylglycerol (DG) and free fatty acids (FFA) were removed and transferred to scintillation vials for measurement of radioactivity. In each experiment, a positive control for testing the labeled phospholipids was carried out by incubating the substrates with 25 U of the snake venom PLA2. The venom PLA2 should give a 75% hydrolysis of the substrate under the incubation condition described above. Background radioactivity in the FFA band was determined by incubating labeled substrate with buffer alone. PLA2 activity in each sample was calculated after subtraction from the background.

Assay of Lactate Dehydrogenase (LDH) Activity in Culture Medium At different times after exposure of DITNC cells to cytokines, aliquots of the culture medium were taken for assay of lactate dehydrogenase (LDH) activity according to the procedure described by Moldeus et al. (1978).

Northern Blot Analysis of Group II PLA2 rnRNA Total RNA was extracted from cultured cells according to the procedure described by Chomczynski and Sacchi (1987). The RNA sample (30/~g) was applied to 1% agarose gel with 2.2M formaldehyde. After electrophoresis, the RNA bands were transferred onto a nylon membrane (DuPont, Boston, MA). Two oligomer probes were synthesized based on the cDNA sequence of group II PLA2 in rat spleen (Ishizaki et al., 1989). These two probes were labeled with [32p]-dATP using the DNA 3'- endlabeling Kit (Boehringer Mannheim, Indianapolis, IN). After crosslinking the RNA to the membrane for 4-6 h at 80~ the membrane was subjected to hybridization using the QuikHYB solution (Stratagene, LaJolla, CA). After hybridization and washing, membranes were exposed to Kodak X-OMAT-AR imaging film (Eastman Kodak, Rochester, NY) and radioactive bands were detected by autoradiography. After exposure, the membrane was stripped and washed and then reprobed with the oligomer probe for ~-actin, which was used as an internal standard.

Molecular and Chemical Neuropathology

Vol. 25, 1995

Cytokine Effects on Phospholipase A2 in Astrocytes

RESULTS

Induction of mRNA Expression of Group fl PLA2 by LPS, TNF~, and IL 1 Initial testing of the PLA2 oligomer probe by Northern blot analysis indicated specific hybridization of the probe with total RNA isolated from rat spleen with one labeled RNA species (Sun and Hu, 1995). After exposing DITNC cells to LPS (10/~g/mL). TNFc~ (200 U/mL), and IL-lfl (100 U/mL) for 24 h, control and exposed cells were harvested for extraction of RNA. RNA samples were separated by gel electrophoresis followed by transfer onto a nylon membrane which was subsequently used for RNA hybridization with the oligomer probes for group II PLA2. Results of Northern blot analysis indicate that all three factors could induce group II PLA2 rnRNA in DITNC cells (Figs. 1A and B). Similar results were obtained for the cultured glial cells derived from newborn rat brain although only data for LPS was shown (Fig. 1C). Results in Fig. 1A and 1B show differences in levels of mRNA expression with cells exposed to two sources of TNFo~ (200 U/mL). It is interesting that cells exposed to the recombinant human TNFo~ (Fig. 1A) actually showed higher induction of PLA2 mRNA as compared to those exposed to the recombinant murine TNFc~ (Fig. 1B).

Incubation of Brain and Cell Homogenates with [14C]Arachidonoyl.Phosphollplds Initially, different phospholipids (PC, PE, PI, and PS) labeled with [14C]arachidonic acid (AA) were incubated with brain and cell homogenates (cultured glial cells and DITNC cells) in order to test the substrate specificity of PLA2. As shown in Table 1, incubation of all three types of homogenates with labeled phospholipids resulted in the release of labeled AA. Data in Table 1 further show that PLA2 activity was greater in cell homogenates than in brain homogenates. In both types of cell homogenates, labeled PE was a more effective substrate as compared to PC and PI. Among different labeled phospholipid substrates, [14C]arachidonoyl-PI was the only phospholipid showing a simultaneous release of labeled DG and AA, indicating the presence of PI-specific phospholipase C and PLA2. In fact, in all three types of homogenates, incubation with labeled PI yielded a greater amount of DG than AA (Table 1).

Effects of LPS, TNFc~, and IL. 1 fl on PLA2 Release in DITNC Cells When the cultured glial cells or DITNC cells were changed to a serumfree medium containing bovine serum albumin (BSA), increases in PLA2 activity in the cultured medium could be observed upon incubation of

Molecular and Chemical Neuropathology

Vol. 25, 1995

8

Tong, Hu, and Sun A

B

-Actin

9 "r~ i~:

?

:

PIA 2

e

r !i: ":j

li

..u~

Fig. 1. Northern blot analysis of group II PLA2 mRNA expression in DITNC cells and in cultured glial cells. (A) DITNC cells were exposed to LPS (10/~g/mL, E. coli, Sigma) and TNF~ (200 U/mL, recombinant human, Sigma) for 24 h. (B) DITNC cells were exposed to IL-1/~ (100 U/mL, recombinant murine, R & D System), LPS (10/~g/mL, E. coli, Sigma) and TNFol (200 U/mL, recombinant murine, R & D System). (C) Cultured glial cells exposed to LPS (10 #g/mL, E. coli, Sigma). Conditions for hybridization have been described in text. Cultured glial cells were isolated from newborn rat pups (3-4 passages). RNA from cells were hybridized with the oligomer probes constructed from rats spleen cDNA (Ishizaki et al., 1989). After hybridization, the same membrane was stripped and rehybridized with the oligomer probe for/3-actin, which was used as a positive control. Molecular and Chemical Neuropathology

Vol. 25, 1995

Cytotdne Effects on Phospholipase Az in Astrocytes Table 1 PLA2 Activity in Rat Brain Homogenates and Homogenates from Cultured Glial Cells and DITNC Cells ~ Brain PI

DG

PC PI PE PS

1 . 1 + 0.3 11.3+ 2.1 0.5 + 0.1 -

Glial cells AA

1.3• 4.8+ 5.9 + 1.1 +

DG 0.6 0.5 1.1 0.1

DITNC cells AA

0.5 + 0.1 21.6 + 2.2 0.5 + 0.1 -

3.6+ 4.7+ 12.7• 6.2 •

DG 0.6 1.3 1.0 0.4

0 . 6 • 0.1 3 3 . 9 • 2.2 0 . 7 + 0.1 -

AA 8.7• 6.9+ 16.9• 12.3 :t:

1.1 0.8 0.8 0.4

aCells were harvested from culture dishes and homogenized in a buffer with protease inhibitor as described in text. PLA2 activity is expressed as the percent of labeled FFA released from [14C]arachidonoyl-labeled PI, PC, PE, or PS/100 ag protein after subtraction of background activity. Values are mean + SD from three samples with duplicate determinations for each sample. Table 2 Induction of PLA2 Release by LPS, TNFo~, and IL-lfl: Comparison Between Cultured Glial Cells and DITNC Cells a Glial cells Treatment Control LPS TNFc~ IL-1/3

DG 1.0 1.6 (n 1.5 (n 1.8 (n

(1.88) • 0.9 = 4) • 0.5 -- 4) • 0.6 -- 4)

DITNC cells AA

DG

AA

1.0 (0.36) 8.7 + 1.6 (n = 4) 3.8 • 1.4 (n = 4) 13.6 + 1.7 (n = 4)

1.0 (0.45) 1.7 • 0.3 ( n - - 4) 1.8 + 0.0 (n = 4) 2.9 • 0.6 (n = 3)

1.0 (0.12) 7.6 • 1.3 (n =4) 3.3 • 0.5 (n --4) 9.3 + 0.7 ( n - - 3)

a PLA2 activity was measured in the culture medium of cultured gliai cells and DITNC cells using [14C]arachidonoyi-PI as substrate. Culture medium was obtained 24 h after incubating cells with LPS (10 ag/mL, E. coli, Sigma), TNFc~ (200 U/mL, recombinant human, Sigma), or IL-1/5 (100 U/mL, recombinant human, Mallinkrodt). PLA2 activity is expressed relative to controls, which is set at 1.0. Values are fold differences from control (mean :t: SD) taking the mean control values as 1 (values in parenthesis are expressed as nmol/mg protein/min.).

cells with LPS (10 #g/mL), TNFoe (200 U/mL), a n d IL-lfl (100 U / m L ) for 24 h (Table 2). U s i n g [14C]aracidonoyl-PI as substrate, i n c u b a t i o n of the culture m e d i u m after e x p o s i n g cells to c y t o k i n e s r e s u l t e d in o n l y A A release. A m o n g the three a g e n t s tested, TNFo~ w a s least effective in t h e i n d u c t i o n of PLA2 as c o m p a r e d to LPS a n d IL-lfl (Table 2). Examination of the time course for PLA2 activity in the culture m e d i u m after e x p o s u r e of D I T N C cells to LPS, TNFo~, a n d IL-lfl indicated a lag time of 2 h prior to the s t e a d y increase, reaching a plateau b e t w e e n 12 a n d 24 h (Fig. 2). U n d e r the s e r u m - d e p r i v e d incubation condition, the culture Molecular and Chemical Neuropathology

Vol. 25, 1995

10

Tong, Hu, and Sun O Ctrl oft

E

15

9

LPS

9 TNI:e

0

L GO E

1.0 ps

E r" s S

m

.....--"

j 4"

0.5

......

(..

"''''''"

~~

.o.."

lllJU.....0-',' "" 0

lO

I

30 20 Incubation time [hours]

I

I

40

50

Fig. 2. Time course for effects of LPS, TNFoe, and IL-lfl on PLA2 release in DITNC cells. Conditions for exposing cells to LPS (10 #g/mL, E. coli, Sigma), TNFa (200 U/mL, recombinant human, Sigma), and IL-I~ (100 U/mL, recombinant human, Mallinckrodt) were same as described in Table 2. Aliquots of the culture medium were taken for assay of PLA2 activity, which is expressed as nmol of arachidonic acid released from [14C]arachidonoyl-PI/mg protein/min. Values are mean + SD from three cultured dishes for each time point with duplicate determinations for sample in each cultured dish. m e d i u m from control cells also showed a small increase in PLA2 release with time. W h e n a portion of the same culture m e d i u m was taken for assay of LDH release as a measure of the extent of cell lysis, results indicated that no LDH release could be observed during the time of PLA2 release. Furthermore, the increase in LDH between 24 and 48 h was found in control as well as in cells exposed to cytokines (Fig. 3). In an experiment to examine the dose-response of these cytokines on the release of PLA2 from DITNC cells, the increase in PLA2 release correlated well with the levels of LPS, IL-lfl, and TNFor added to the culture m e d i u m although different doseresponse curves for each cytokine were observed (Fig. 4). Pretreatment of cells with cyclohexamide (1.25/aM) prior to addition of LPS, IL-lfl and TNFa resulted in complete inhibition of the PLA2 release by all three agents. Using DITNC cells, an experiment was carried out to examine effects of cytokines on induction of PLA2 mRNA expression as well as release of PLA2 into the culture m e d i u m . Using [14C[arachidonoyl-PE as substrate, exposure of cells to LPS and IL-lfl resulted in a 14-15-fold increase in PLA2 activity in the culture m e d i u m whereas exposure of cells to TNFc~ resulted in only a 2.7-fold increase (Table 3). Northern blot analysis using the same batch of cells indicated a correlative induction of m R N A expression by the

Molecular and Chemical Neuropathology

Vol. 25, 1995

Cytokine Effects on PhosphoUpase A2 in Astrocytes

11

600 O

...d

Ctrl

9 TNFa

9 LPS ji

9 IL-i fl

C- 40o E~ -~.-4 -'O

I

,

0

,

10 20 30 Incubation time [hours]

,

,

40

50

Fig. 3. Time course for effects of LPS, TNFo~, and IL-lfl on lactate dehydrogenase (LDH) release in DITNC cells. Conditions for exposing cells to LPS, TNF, and IL-lfl were the same as in Fig. 2. Aliquots of the culture medium were taken for measurement of LDH activity, which is expressed as U/dL of medium. Values are means from three dishes for each time point. Mean + SD for 48-h samples are: control, 456 + 228; LPS, 348 + 54; TNF, 654 • 211; IL-1, 546 + 130 U/L.

same agents (Fig. 1B). When cell cytosol and membranes were assayed using [14C]arachidonoyl-PI as substrate, endogenous PLA2 as well as phospholipase C (data not shown) activity was observed in both cell fractions. Exposure of cells to LPS and IL-lfl resulted in a small increase (40-50%) in PLA2 activity in both the cytosol and membrane fractions (Table 3). When cell cytoso! was incubated with labeled PI, PE, and PC in the presence or absence of Ca 2., both Ca2§ and i n d e p e n d e n t PLA2 activities could be observed in this cell fraction (data not shown).

Interactive Effects of Cytokines on PLA2 Release in DITNC Cells In order to assess the interactive effects of LPS, IL-lfl, and TNFc~ on the induction of PLA2 release in DITNC cells, suboptimal amounts of these agents (25-50% of original) were used either singly or in combination. Under these conditions, the fold increases in PLA2 release for LPS (2.5 #g/mL), IL-I~ (25 U/mL) and TNFor (100 U/mL) were: 22.7, 14.1 and 1.5, respectively. When this amount of TNFcx was added together with LPS or IL-lfl to the culture medium, the fold increase in PLA2 release was 64.8 for TNFcx plus LPS and 42.0 for TNFo~ plus IL-lfl (Fig. 5). On the other hand, w h e n IL-lfl was added to the culture m e d i u m together with LPS, the combined effect was the sum of the two (31.9) (Fig. 5).

Molecular and Chernical Neuropathology

Vol. 25, 1995

12

Tong, Hu, and Sun c o

E

6

d.1

i

~b

16o

LPS [p~ml]

o

o

3

c.

g

1

0

20

200

2000

TNFe [ueiUmll

0

I

10

100

1000

IL-Ip [uoil/ml]

Fig. 4. Dose response for LPS, TNFcr and IL-lfl on PLA2 release from DITNC cells. Conditions for exposure of different concentrations of cytokines were similar to that in Fig. 2. PLA2 activity is expressed as nmol of arachidonic acid released from [14C]arachidonoyl-PI/mg protein/rain. Values are means • SD from four dishes for each point. We also tested the ability of antibodies to IL-1/3 and TNFc~ to neutralize the effects of IL-I~ and TNFc~ on induction of PLA2 release in DITNC cells. Data in Fig. 6 show that the antibodies completely neutralized the effects of IL-I~ and TNFc~ on PLA2 release. However, w h e n these antibodies were a d d e d to the culture m e d i u m containing LPS, neither antibody was able to alter PLA2 release due to induction by LPS (Fig. 6).

DISCUSSION Previous studies have demonstrated the effects of proinflammatory factors and cytokines (e.g., IL-lfi and TNFo~) on enhancing the release of

Molecular and Chemical Neuropatholog~

Vol. 25, 1995

Cytoklne Effects on Phospholipase A2 in Astrocytes

13

Table 3 Effects of LPS IL-lfl and TNFc~ on PLAz Activity in Culture M e d i u m a n d in Cytosol and M e m b r a n e Fractions from DITNC Cells a Medium % FFA Control LPS IL-lfi TNFo~

3.3 47.5 51.8 8.9

Cytosol Fold

• 0.26 • 5.6 b

• 4.3 b + 1.8

1.0 14.3 15.6 2.7

% FFA 19.7 28.4 27.3 18.2

• + • •

Membrane Fold

1.1 5.3 2.8 1.8

% FFA

1.0 1.4 1.4 0.9

14.1 19.3 21.4 10.7

• :t: • •

Fold

1.3 0.3 2.8 3.1

1.0 1.4 1.5 0.8

aCondition for treatment was same as in Fig. lB. After exposure of DITNC cells to LPS (10 ag/mL, E. coli, Sigma), IL-1/~ (100 U/mL, recombinant murine from R & D), and TNFc~ (200 U/mL, recombinant routine from R & D System) for 24 h, cells were harvested and the homogenates were centrifuged to yield a cytosolic and membrane fraction. For assay of PLA2 activity in the culture medium, [l~C]arachidonoyl-PE was incubated with 100 aL of the medium. For assay of PLAz in the cytosol (300 ag protein) and membrane (120 ag protein) fractions, [14C]arachidonoyl-PI was used as substrates. PLA2 activity is expressed as percent of total radioactivity recovered from the TLC plate and the fold increase from control. Values are mean • SD from three samples. bDenotes values that are significantly different from controls based on ANOVA followed by the Bonferroni t-test, p < 0.05. 70 60 50 k-

cu

40

-~ -8

30

LL

20

q-

Ctrl

TNFa IL-IB

LPS IL-tB IL-IB LPS +

LPS

+

+

TNFa TNFa

Fig. 5. Interactive effects of LPS, TNFo~, and IL-lfl on induction of PLA2 release from DITNC cells. DITNC cells were exposed to suboptimal a m o u n t s of the cytokines: LPS, 2.5 ~g/mL (E. coli, Sigma), IL-1/3, 25 U/mL (murine recombinant, R & D System), and TNFol, 100 U/mL (murine recombinant, R & D System). These factors were a d d e d singly or in combination to the culture m e d i u m and PLA2 release was m e a s u r e d 24 h after e x p o s u r e to cytokines or antibody. Results are e x p r e s s e d as fold increase (mean + SD from four cultured dishes except IL-lfi + TNFcr which is n --- 3) using controls as 1 (61 • 39 d p m / m g protein, n = 4).

Molecular and Chemical Neuropathology

Vol. 25, 1995

14

Tong, Hu, and Sun 50

40

30 LL LL

210,-0I o

L

CtrlTNFTNF +

TNF-Ab

Ctrl

IL IL + IL-Ab

Ctrl LPS LPSLPS +

+

IL-Ab TNF-Ab

Fig. 6. Effectof antibodies to IL-lfl (antihuman rabbit polyclonal, 25/~g/mL) and antibodies to TNFa (antimurine goat polyclonal, 10/~g/mL) on ability of IL-lfl (25 U/mL), TNFol (200 U/mL), and LPS (2.5 #g/mL) to induce PLA2. Values are percent labeled FFA released (mean + SD from four cultured dishes) for IL-1 and LPS stimulation and average of two dishes of TNF stimulated samples. group II PLA2 in a number of non-neural cells, e.g., rabbit chondrocytes (Chang et al., 1986), cultured rat gingival fibroblasts (Shinohara et al., 1992), rat renal mesangial cells (Pfeilschifter et al., 1989), fetal rat calvarial bone-forming cells (FRCC) (Vadas et al., 1991), and rat vascular smooth muscle cells (Nakano et al., 1990). Studies with rat mesangial cells further indicate that cytokine stimulation of PLA2 was marked by a parallel increase in prostaglandin Ez (Pfeilschifter et al., 1993), suggesting that one of the physiological consequences of the cytokine action is to mediate cell response through synthesis of specific eicosanoids. Similar to nonneural cells, cultured glial cells from brain are responsive to cytokines on the induction of group II PLA2 mRNA and its release (Oka and Arita, 1991). Despite the presence of endogenous PLA2 in these cells, cytokine induction seems to be limited to the secretory PLA2, which is not constitutively expressed in the cells. Similar to the cultured glial cells obtained from newborn rat brain, the immortalized astrocytes (DITNC) responded to cytokines and showed an induction of type II PLA2 mRNA and a correlative increase in release of this enzyme to the culture medium. Nevertheless, subtle differences in the response could be observed comparing the DITNC cells to the cultured glial cells. For example, exposure of the DITNC cells to all three factors indicated a time-dependent increase in PLA2 release after a lag time of 2 h, whereas cultured glial cells exhibited a longer lag time of 4 h (Oka and Arita, 1991). It is apparent that induction time varies greatly among cell types. For example, a lag time of 8 h was required for rat mesangial cells (Pfeilschifter et al., 1989) and the calvaria bone forming cells showed a lag Molecular and Chemical Neuropathology

VoL 25, 1995

Cytokine Effects on Phospholipase A2 in Astrocytes

15

time of 5 d (Vadas et al., 1991). Not only did the DITNC cells have shorter lag time, but the time course for induction of PLA2 release by cytokines was completed at a relatively short time of 8-10 h. In comparision, exposure of cultured glial cells to LPS resulted in a steady release of PLA2 for 24 h (Oka and Arita, 1991). It appears that in both DITNC cells and cultured glial cells, TNFc~ was least effective in inducing PLA2 as compared to IL-I~ and LPS. Similar observation was found in rat mesangial cells (Pfeilschifter et al., 1989). These differences were attributed to using TNFc~ (from recombinant human and murine) that were not compatible with the receptor in rat cells. Using [14C]arachidonoyl-phospholipids as substrates, it can be shown that homogenates from brain, cultured glial cells, and DITNC cells exhibit endogenous PLA2 activity. However, labeled PI was the only substrate showing a concomitant release of DG and FFA. These results confirm previous observations using rat brain synaptosomes (Navidi et al., 1990) and further illustrate that cytokine induction is specific for PLA2 and not PLC. There is evidence that in some cell systems, cytokines may cause the increase in activity of cytosolic PLA2 (Lin et al., 1992; Schalkwijk et al., 1992; Gronich et al., 1993; Hayakawa et al., 1993; Wu et al., 1994). In this study, exposure of DITNC cells to LPS and IL-lfl also showed a small increase in cellular PLA2 activity (40-50%). However, more information is needed to identify the PLA2 species involved intracellularly. Cultured glial cell preparations are known to contain a heterogeneous mixture of cells, including oligodendroglial progenitor cells, microglia, and astrocytes. With increasing passages, the proportion of astrocytes in the culture also increased. Using this type of culture, previous studies have shown the ability of LPS to stimulate the release of cytokines, especially IL-lfl and TNFo~ (Lieberman et al., 1989). Furthermore, TNFc~ secretion could be primed by LPS together with interferon 7 (Chung and Benveniste, 1990). In the present study, two experiments were carried out to test for interactive effects of the cytokines and whether LPS effects on DITNC cells may be contributed by its ability to induce IL-lfl and TNFo~ synthesis. Results demonstrated synergistic responses in induction of PLA2 release when TNFo~ was added to the culture medium together with LPS or IL-lfl. On the other hand, the combination of LPS and IL-I~ gave only an additive response. Although pretreatment of DITNC cells with antibodies to IL-lfl and TNFc~ could neutralize the effects of these agents on induction of PLA2 release, the antibodies did not alter the effect of LPS. Therefore, it can be concluded that LPS action on DITNC cells was not complicated by the release of IL-lfl and TNFo~. Despite the ability of cycloheximide to block cytokine-induced PLA2 synthesis, the transcriptional regulation as well as the signal transduction pathway(s) mediated by these cytokines remain to be elucidated. Our recent study, however, demonstrated TNFoMnduced activation of DNA binding of a RelA homodimer which was preceded by phosphorylation of RelA and MAD-3 (Diehl et al., 1995). Experiments to further elucidate the transcriptional regulation of PLA2 gene expression are in progress. Molecular and Chemical Neuropathology

Vol. 25, 1995

16

Tong, Hu, and Sun

ACKNOWLEDGMENT This research project was supported in part by NS30178 from NIH.

REFERENCES Burch R. M. and Tiffany C. W. (1989) Tumor necrosis factor causes amplification of arachidonic acid metabolism in response to interleukin-1, bradykinin, and other agonist. J. Cell. Physiol. 141, 85-89. Chang J., Gilman S. C., and Jewis A. J. (1986) Interleukin-1 activates phospholipase A2 in rabbit chondrocytes: a possible signal for IL-1 action. J. Immunol. 136, 1283-1287. Chao C. C., Hu S. X., Close K., Choi C. S., Molitor T., Novick W. J., and Peterson P. K. (1992) Cytokine release from microglia: differential inhibition by pentoxifylline and dexamethasone. J. Infectious Diseases 166, 847-853. Chomczynski P. and Sacchi N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156-159. Chung I. Y. and Benveniste E. N. (1990) Tumor necrosis factor-o~ production by astrocytes, induction by lipopolysaccharide, IFN-~, and IL-I~/. J. Immunol. 144, 2999-3007. Corbin D. R. and Sun G. Y. (1978) Characterization of the enzymic transfer of arachidonoyl groups to 1-acyl-phosphoglycerides in mouse synaptosome fraction. J. Neurochem. 30, 77-82. Diehl J. A., Tong W., Sun G. Y., and Hannink M. (1995) TNFo~-dependent activation of a RelA homodimer in astrocytes: Increased phosphorylation of RelA and MAD-3 precede activation of RelA. J. Biol. Chem. 270, 2703-2707. Fontana A., Frei K., Bodmer S., and Hofer E. (1987) Immune-mediated encephalitis on the role of antigen-presenting cells in brain tissue. Immunol. Rev. 100, 185-201. Ganter S., Northoff H., Mannel D., and Gebicke-Harter P. J. (1992) Growth control of cultured microglia. J. Neurosci. Res. 33, 218-230. Hayakawa M., Ishida N., Takeuchi K., Shibamoto S., Hori T., Oku N., Ito F., and Tsujimoto M. (1993) Arachidonic acid-selective cytosolic phospholipase A 2 is crucial in the cytotoxic action of tumor necrosis factor. J. Biol. Chem. 268, 11290-11295. Ishizaki J., Ohara O., Nakamura E., Tamaki M., Ono T., Kanada A., Yoshida N., Teraoka H., Tojo H., and Okamoto M. (1989) cDNA cloning and sequence determination of rat membrane-associated phospholipase A2. Biochem. Biophys. Res. Commun. 162, 1030-1036. Kelleher J. A. and Sun G. Y. (1985) Enzymatic hydrolysis of arachidonoyl-phospholipids by rat brain synaptosomes. Neurochem. Int. 7, 825-831. Kimelber~ H. K. and Norenberg M. D. (1989) Astrocytes. Scientific American 260, 66-72. Lieberman A. P., Pitha P. M., Shin H. S., and Shin M. L. (1989) Production of tumor necrosis factor and other cytokines by astrocytes stimulated with Molecular and Chemical Neuropathology

Vol. 25, 1995

Cytokine Effects on Phospholipase A2 in Astrocytes

17

lipopolysaccharide or a neurotropic virus. Proc. Natl. Acad. Sci. USA 86, 6348-6352. Lin L. L., Lin A. Y., and Dewitt D. L. (1992) Intefleukin-1 alpha induces the accumulation of cytosolic phospholipase A2 and the release of prostaglandin E2 in human fibroblasts. J. Biol. Chem. 267, 23451-23454. Mayer R. J. and Marshall L. A. (1993) New insights on mammalian phospholipase A2 (s); comparison of arachidonoyl-selective and -nonselective enzymes. FASEB J. 7, 339-348. Merrill J. E. (1991) Effects of interleukin-1 and tumor necrosis factor-~ on astrocytes, microglia, oligodendrocytes, and glial precursors in vitro. Dev. Neurosci. 13, 130-137. Moldeus P., Hogberg J., and Orrenius S. (1978) Isolation and use of liver cells, in Methods in Enzymology (Jaboby W. B., ed.), pp. 60-71, Academic Press, New York. Murphy S. (1990) Generation of astrocyte cultures from normal and neoplastic central nervous system, in Methods in Neurosciences, vol. 2 (Conn P. M., ed.), pp. 33-47, Academic Press, New York. Navidi M., MacQuarrie, R., and Sun, G. Y. (1990) Metabolism of phosphatidylinositol in plasma membranes and synaptosomes of rat cerebral cortex: A comparison between endogenous vs exogenous substrate pools. Lipids 25, 273-277. Oka S. and Arita H. (1991) Inflammatory factors stimulate expression of group II phospholipase A 2 in rat cultured astrocytes. J. Biol. Chem. 266, 9956-9960. Radany E. H., Brenner M., Besnard F., Bigornia V., Bishop J. M., and Deschepper C. F. (1992) Directed establishment of rat brain cell lines with the phenotypic characteristics of type I astrocytes. Proc. Natl. Acad. Sci. USA 89, 6467-6471. Sawada M., Kondo N., Suzumura A., and Marunouchi T. (1989) Production of tumor necrosis factor-alpha by microglia and astrocytes in culture. Brain Res. 491, 394-397. Selmaj K., Shafit-Zagardo B., Aquino D. A., Farooq M., Raine C. S., Norton W. T., and Brosnan C. F. (1991) Tumor necrosis factor-induced proliferation of astrocytes from mature brain is associated with down-regulation of glial fibrillary acidic protein mRNA. J. Neurochem. 57, 823-830. Shinohara H., Amabe Y., Komatsubara T., Tojo H., Okamoto M~ Wakano Y., and Ishida H. (1992) Group II phospholipase A2 induced by interleukin-1/~ in cultured rat gingival fibroblasts. FEBS Lett. 304, 69-72. Sun G. Y. (1988) Preparation and analysis of acyl and alkenyl groups of glycerophospholipids from brain subcellular membranes, in Neuromethods: Lipids and Related Compounds, vol. 7 (Boulton A. A., Baker G. B., and Horrocks L. A., eds.), pp. 63-82, Humana, Totowa, NJ. Sun G. Y. and Hu Z. Y. (1995) Stimulation of phospholipase A2 expression in rat cultured astrocytes by LPS, TNFcr and IL-I~, in Progress in Brain Research, vol. 105 (Yu A. C. H., Eng L. F., McMahan K. J., Schulman H., Shooter E. M., and Stadlin A., eds.), pp. 231-238, Elsevier Science, Amsterdam, The Netherlands.

Molecular and Chemical Neuropathology

Vol. 25, 1995