TOXICOLOGY

AND APPLIED

PHARMACOLOGY

117, 19-25 ( 1992)

Effect of Ozone on Platelet Activating Factor Metabolism in Phorbol-Differentiated HL60 Cells’ JAMES M. SAMET AND MITCHELL FRIEDMAN Center for Environmental Medicine and Lung Biology, Curriculum in Toxicology and Department of Medicine, University of North Carolina at Chapel Hill. Chapel Hill, North Carolina 27599 Received April 15, 1992; accepted July 14, I992

0alkyl-2-acetyl-sn-glycero-3-phosphocholine, PAF) is a potent lipid inflammatory mediator that has a wide spectrum of biologic activity in many tissues (Hanahan and Kumar, 1987; Sturk et al., 1989; Braquet and Rola-Pleszczynski, 1987). Interestingly, some of the effects of PAF in the lung are similar to those observed following O3 inhalation (McManus and Deavers, 1989; Lippman, 1989). PAF is synthesized in response to proinflammatory stimuli in cells such as macrophages via the remodeling metabolic pathway (Albert and Snyder, 1983; Snyder, 1987). The first step in this pathway consists of the removal of the sn-2 fatty acid from 1-O-alkyl-2-acyl-sn-glycero-3-phosphocholine (alkylacyl-glycerophosphocholine, alkylacyl-GPC) by phospholipase A2 (PLA*), to produce 1-O-alkyl-2-lyso-sn-glycero3-phosphocholine (1ysoPAF). In addition to PLA2, a coenzyme A-independent transacylase has been reported to be involved in the formation of 1ysoPAF from alkylacyl-GPC via the transacylation of ethanolamine or choline-containing lysophospholipids (Uemura et al., 1991). LysoPAF is then acetylated at the sn-2 position by a specific acetyltransferase ( 1-O-alkyl-2-lyso-sn-glycero-3-phosphocholine:acetylCoA acetyltransferase) to produce PAF (Lee, 1985). The enzymatic activities of PLA2 and acetyltransferase rapidly increase

Effect of Ozone on Platelet Activating Factor Metabolism in Phorbol-Differentiated HL60 Cells. SAMET, J. M., AND FRIEDMAN, M. (1992). Toxicol, Appl. Pharmacol. 117, 19-25. The mechanisms of ozone (0,) toxicity in the lung may involve the formation of lipid inflammatory mediators. We have previously demonstrated that exposure to O3 in vitro results in increased accumulation and release of platelet activating factor (PAF) in the macrophage-like cell line HL60 differentiated with phorbol ester (dHL60). In the present study we have examined possible biochemical mechanisms responsible for the O,-induced increase in PAF levels in dHL60 cells. Specifically, we studied the effect of O3 on phospholipase AZ (PLA*), acetyltransferase, acetylhydrolase, and reacylation activities. dHL60 cells were exposed to 1.0 ppm O3 or air alone. O3 exposure was found to significantly decrease dHL60 cell acetylhydrolase activity by 36%. Additional experiments demonstrated that extracellular acetylhydrolase activity, but not intracellular acetylhydrolase activity, was inhibited by OS exposure of dHL60 cells. O9 exposure resulted in a small (13%) but statistically significant reduction in reacylation activity in dHL60 cells. In addition, a significant (22%) contribution of PLA2 activation to the 03-induced increase in PAF levels was also found. Basal and calcium ionophore-induced acetyltransferase activity was found to be unaffected by exposure of dHL60 cells to 03. These data suggest that in vitro exposure to O3 affects both synthetic and degradative pathways of PAF metabolism in dHL60 cells. o 1992 Academic

upon stimulation

of the cell with a suitable agonist or with

calcium ionophore (Suga et al., 1990; Lapetina and Crouch, 1989; Ninio et al., 1987). PAF inactivation and catabolism, essentially a reversal of the synthetic steps, is accomplished by deacetylation of the molecule by the enzyme 1-O-alkyl2-acetyl-sn-glycero-3-phosphocholine:acetylhydrolase, to form 1ysoPAF (Blank et al., 1981). Two reacylation mechanisms have been described for the conversion of 1ysoPAF back to alkylacyl-GPC; one is a nonspecific CaA-dependent mechanism and the other an arachidonic acid-specific, CoAindependent mechanism (Robinson et al., 1985; Sugiura and Waku, 1985). We have previously demonstrated that in vitro O3 exposure of relevant lung cell types, including human alveolar macrophages and the cell line HL60 differentiated into macrophage-like cells with phorbol ester (dHL60 cells), results in increased production and release of PAF by these cells (Friedman et al., 1992; Samet et al., 1992). To elucidate the

Press. Inc.

Ozone (0,) is a highly reactive oxidant molecule that is commonly found in the troposphere as a major constituent of photochemical smog (Hough and Derwent, 1990). The

lung is a principal target of its toxicity (Mehlman and Borek, 1987). Both in vivo and in vitro studies have suggested that the mechanism of toxicity following O3 inhalation lung may involve the production of lipid inflammatory mediators by cells directly exposed to O3 (Seltzer et al., 1986; Koren et al., 1989; Madden et al., 1991). Platelet activating factor (l’ Supported by United States Environmental Protection Agency Cooperative Agreement CR8 12738, United States Environmental Protection Agency Grant R8 17298, and National Institutes of Health Grants ES0495 1 and ES07 126. 19

0041-008X/92 $5.00 Copyright 0 1992 by Academic Press, Inc. All rights of remoduction in anv form reserved.

20

SAMET

AND FRIEDMAN

mechanism(s) responsible for the observed increased accumulation and release of PAF by these cells in response to 03, we have studied the effects of in vitro O3 exposure on the remodeling pathway of PAF metabolism in dHL60 cells. dHL60 cells offer the advantages of low cost, ready availability in sufficient numbers, reproducibility, and the fact that they release PAF, in response to 03, similar to primary human alveolar macrophages (Friedman et al., 1992). Specifically, we have examined PLA2, acetyltransferase, acetylhydrolase, and reacylation pathways in dHL60 cells and now report that O3 affects both synthetic and degradative pathways of PAF metabolism in these cells. METHODS Muteriuls. DMEM/F 12, gentamicinsulfate,L-glutamine,and PBSwere obtained from the UNC Lineberger Cancer Center Tissue Culture Facility (Chapel Hill, NC). Nu Serum IV was obtained from GIBCO (Grand Island, NY). Tissue culture plasticware was bought from Falcon (Oxnard, CA). PMA, unlabeled PAF and lysoPAF, acetylCoA, fatty acid-free bovine serum albumin, mepacrine (quinacrine dihydrochloride), and calcium ionophore A23 187 were purchased from Sigma (St. Louis, MO). TLC plates were bought from Merck (VWR, Durham, NC). Zero Grade air tanks were obtained from National Welders (Raleigh, NC). l-O-[3H]octadecyl-2-sn-glycero-3phosphocholine ([‘HllysoPAF; specific activity, 80-180 Ci/mmol) and [Nmethyl-‘4C]l-O-alkyl-2-acetyl-sn-3-glycerophosphocholine ([‘4C]PAF; specific activity, 50-60 mCi/mmol) were obtained from Amersham (Arlington Heights, IL). ([3H]acetyl)acetyl coenzyme A ([‘HlacetylCoA; specific activity, 2.1 Ci/mmol) was obtained from NEN Du Pant (Wilmington, DE). HPLCgrade solvents and common laboratory reagents were purchased from Fisher (Raleigh, NC). Coomassieblue protein assay reagent was purchased from BioRad (Rockville Center, NY). dZfL60 cell culture. Undifferentiated myeloid leukemic HL60 cells (American Type Culture Collection, Rockville, MD) were grown in DMEM/ F12 (1:1) supplemented with 10% Nu Serum IV, 50 mg/liter gentamicin sulfate and 2 mM L-glutamine (complete DMEM/FlZ) at a density of 0.52.0 X lo6 cells/ml. Cells were induced to differentiate toward monocytic development with 30 nM PMA at a density of 0.5-l X IO6 cells ml in 4 ml of complete DMEM/F12 in 60-mm polystyrene tissue culture dishes for approximately 72 hr. Differentiation was assessedby morphologic and enzymatic criteria, as previously described (Samet et al., 1992; Rovera et al., 1977). HL60 cells treated with PMA became growth arrested, adherent, flattened and elongated in appearance, and had long pseudopod-like processes.Acid phosphatase activity in PMA-differentiated HL60 cells at Day 3 of differentiation, as assessedwith a standard assay kit, was 96.7 + 6.3 nmol/min mg protein, compared to 20.0 + 2.7 nmol/mg protein for the undifferentiated cells (n = 2 experiments assayed in triplicate). dHL60 cells were exposed to O3 or air in 1.5 ml of PBS containing 1 mM calcium and magnesium and 1 g/liter glucose, while rocking at 3 rpm. In vitro 0, exposure. Exposures were carried out in an in vitro O3 exposure system that we have used to expose macrophages and other lung cell types while maintaining them in a viable state (Friedman et a/., 1992). The in vitro 0, exposure system consisted of two glass chambers (2.5 liters in volume) through which humidified Zero Grade air (< 1 ppm organic contaminants), with or without 03, was passed at a flow rate of 1.2 liters/min. O3 was generated by passing humidified air over a partially occluded uv pen lamp. Air exiting the chambers was dehumidified in a condenser and analyzed in a Bendix 8002 Ozone Analyzer (Bendix, Lewisburg, WV). All O3 exposures were performed at a concentration of 1.O ppm which was maintained within *lo% of the desired concentration throughout the exposure period. The chambers were placed over rocking platforms and the system was enclosed in a 37°C incubator.

Extraction and analysis of PAF and metabolites. PAF synthesis was measured by labeling dHL60 cells prior to exposure with 1.5 &i [‘H]lysoPAF in 1.5 ml of supplemented DMEM/FI2 containing 0.1% BSA for 30 min at 37°C. Lipids were extracted from cells and media using the procedure of Folch et a/. (Folch et al., 1957). Carrier PAF (50 pg/sample) and IysoPAF (50 &le) were included in the extraction solvents to improve recoveries and serve as markers on thin-layer chromatography (TLC). In some cases a few hundred dpm of [14C]PAF were also included as an internal standard. Lipids were dried under a stream of N2 gas,redissolved in 50 ~1 of chloroform, and chromatographed on silica TLC plates in a solvent system consisting of 100:50: 16:8 chloroform:methanol:acetic acid:water (Chilton et al., 1984). Lipid bands were visualized with Iz vapor, scraped, and quantified by liquid scintillation counting. We have previously demonstrated (Samet et al., 1992) that this method of analysis for PAF in dHL60 cells gives results that are similar to those obtained using a commercially available PAF radioimmunoassay kit (New England Nuclear, Wilmington, DE). The identity of PAF produced by dHL60 cells was confirmed by its sensitivity to PLAz digestion, chromatographic behavior on TLC and HPLC, and bioactivity toward human neutrophils and washed rabbit platelets (Samet et al., 1992). dHL60 cells were incubated with the PLA2 inhibitor mepacrine (1 mM) for I5 min, washed, and then challenged with O3 or A23 187 (10 FM) in PBS with Ca*+, Mg’+, and glucose. Acetyltransferase and acetylhydrolase activity assays. Acetyltransferase activity was assayed using a modification of previously published methodology (Ninio et al., 1987). The acetyltransferase activity assay was optimized in preliminary experiments for linearity over the duration of the assay (between 15 and 45 min of assay) and for the duration of stimulus of the cells with A23 187 (30-90 set). Immediately following exposure to O3 (or air alone as a control) for 5-90 min, 0.75 ml of the exposure medium was removed and replaced with the same volume of PBS or PBS containing A23 187 at 37°C (final A23187 concentration, 10 FM). After 30 set of stimulation with A23 187, the cells were rinsed once with ice-cold PBS and scraped in 600 /II of PBS (without Ca2+ or Mg’+) using a plastic cell scraper. The cells were then sonicated at 4°C for 5 set using a Fisher Model 300 sonic dismembrator fitted with a titanium microtip at a relative output of 0.6. No intact cells were visible by light microscopic examination following this treatment. Fifty microliters of the crude cell homogenate was then added to 450 ~1 of assay buffer consisting of Tyrode’s buffer (pH 7.3) containing 0.25% BSA and 190 pg of IysoPAF and preincubated for 5 min at 37°C. Twenty microliters of a substrate solution consisting of 0.5 &i [‘H]acetylCoA and 160 pegof unlabeled acetylCoA was added, and the mixture was incubated at 37°C for an additional 30 min. The reaction was stopped by immediate addition of 500 ~1 of methanol and 1.0 ml of chloroform containing 100 pg of carrier PAF. Lipids were then extracted and chromatographed as described above. The dpm of [3H]PAF detected was normalized for the protein content of the homogenate assayed in each sample. The amount of protein in 50 ~1 of crude cell homogenate averaged 5.43 + 0.32 Kg (n = 74). No differences were observed between the protein content of cell sonicates from cells exposed to vehicle alone compared to A23 187, air, or 03. Acetylhydrolase activity was measured as the apparent half-life of exogenous [14C]PAF in intact dHL60 cell cultures, cell sonicates, and conditioned exposure media (i.e., PBS conditioned by the dHL60 cells during the exposure period), using a modification of previously described methods (Stafforini el al.. 1987, 1989). Cell cultures were scraped and extracted together with the exposure medium as described above. The recovered lipids were chromatographed and the distribution of 14C label among bands corresponding to IysoPAF, PAF, and alkylacyl-GPC was determined by scintillation counting. The half-life of exogenous PAF was determined from semilog plots of PAF deacetylation over time measured by including 1.2 pM ( I X l-2 X 1O5dpm) [14C]PAF in the exposure medium added to dHL60 cultures immediately prior to exposure, or to cell sonicates or conditioned exposure media from exposed cells. Calculations of [14C]PAF half-life and acetylhydrolase activity were made using the linear range of the plotted data. Acetylhydrolase activity in cell sonicates and conditioned exposure media from dHL60 cell cultures exposed to 0, or air for 60 min were assayed separately. A 60-min exposure

EFFECT OF OZONE

TABLE 1 Effect of Ozone on Acetyltransferase Activity in Differentiated HL60 Cells” Duration of ozone exposure (min) 5 Air Air + A23 187 Ozone Ozone + A23 187

100 + 308 2 115f 300 f

15 50 30* 8 19*

100 + 326 + lOOk 334 f

45 55 31* 7 30*

100 + 348 f 90+ 354 f

90 35 26* 6 21*

100 f 12 304 +- 50* 90* 4 254 + 24*

’ Phorbol-differentiated HL60 cells were exposed to air or 1.O ppm ozone for 5-90 min prior to treatment with vehicle or 10 FM calcium ionophore A23187 for 30 sec. Cell sonicates were then assayed for acetyltransferase activity. See text for methodological details. Data are shown as acetyltransferase activity expressed as a percentage of the air-exposed control group, means + SEM, n = 4. * Significant difference from Air group, p < 0.0 1.

period was chosen because earlier experiments had shown it to produce a peak increase in PAF levels in dHL60 cells exposed to O9 (Samet et al., 1992). Conditioned exposure media was removed from cultures immediately following exposure, centrifuged at 12,000 rpm for 10 set to remove any detached cells, and the pH adjusted to 8.0 with sodium hydroxide. The cells were scraped in 1.5 ml of PBS at pH 8.0, chilled, and sonicated 5X for 30 set at a relative output of 0.4. Duplicate 700~~1 aliquots of conditioned exposure media and sonicates were taken; one was assayed in the presence of 4 mM EGTA and the other in 1 mM Ca2+ and Mg*+. Substrate [14C]PAF was then added in 10 ~1 of 0.25% BSA to prewarmed samples, vortexed, and incubated at 37°C for 15-120 min. The reactions were stopped by extraction of the lipids as described above. In assay controls that received no enzyme source, i.e., cell-free PBS, 96-98% of the radioactivity was found in the PAF band. The protein content of 1.5 ml of cell sonicate was 450500 rg, while 1.5 ml of conditioned exposure media contained