INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Ozone, a highly reactive oxidant gas, is a major component of photochemical smog. As an inhaled toxicant, ozone induces its adverse effects mainly on the lung (1). Exposure to ozone, even at levels below the present National Ambient Air Quality Standard, induces airway inflammation and lung injury in both humans and animals (2). The early biochemical and cellular events that initiate ozone-induced inflammation are not fully known.

Many studies have shown that the inflammatory process associated with ozone exposure is mediator-related. This view is supported by data demonstrating increased recovery of several soluble proinflammatory mediators of inflammation in bronchoalveolar lavage (BAL) fluid from humans exposed to ozone. Among the various proinflammatory mediators, increased levels of prostaglandin E₂ (PGE₂), interleukin (IL)-6, and IL-8 have consistently been found in BAL from humans exposed to low levels of ozone (4, 5, 8).

Airway epithelial cells have been proposed to play an important role in the ozone-induced inflammatory process (11). They are among the first cells to come in contact with inhaled ozone and thus, act as a target cell for ozone-induced toxicity. Data from in vitro exposure studies have shown that airway epithelial cells respond to ozone exposure by releasing free arachidonic acid (AA) (12) and the potent proinflammatory lipid mediator platelet-activating factor (PAF) (13, 15, 16). Increased PGE₂ release in primary tracheobronchial cultures (17) and in a human bronchial epithelial cell line, BEAS-2B (14) has been reported during low level in vitro ozone exposure. Ozone exposure has also been shown to induce the release of various proinflammatory cytokines, such as IL-6 and IL-8, during in vitro exposure of cultured airway epithelial cells (18, 19). These in vitro cell culture exposure studies, along with studies examining BAL fluid from humans exposed to ozone, suggest a significant interaction between inhaled ozone and the airway epithelium (18).

Pryor and coworkers (20) have hypothesized, based on the high reactivity of ozone with unsaturated fatty acids (21), that ozone cannot penetrate deeply into the lung (24, 25). Rather, ozone rapidly reacts and is consumed as it contacts the epithelial lining fluid and airway epithelial cells. Exposure of epithelial cell membrane lipids (mainly phospholipids) to ozone will generate lipid ozonation products (LOP) whose yield can be predicted from the Criegee mechanism of ozonation (26). Because of a limited number of mono- and polyunsaturated fatty acids in the lung, only a finite number of LOP can be formed during ozone exposure (20). These LOP include hydroxy hydroperoxides and aldehydes (21, 26).

We have proposed that the LOP that are formed in the lung during ozone exposure in vivo could activate second messenger pathways such as the lipid signaling elements phospholipase A₂ (PLA₂) or phospholipase C (PLC). Activation of phospholipases would result in the generation and release of proinflammatory mediators and subsequently lung inflammation and injury. This hypothesized sequence of events has been termed the cascade theory of ozone toxicity (20).

LOP have been demonstrated in both animals and humans exposed to ozone. The LOP product, nonanal, has been isolated from lung lavage (at 150 nM concentration [31]), and from isolated lungs (300 pmole concentration [32]) of rats exposed to ozone (0.5 ppm × 2 h, and 60 nmoles ozone, respectively). Another aldehyde, 4-hydroxynonenal, also has been demonstrated to be formed during exposure to 0.25 ppm × 3 h ozone in mice (33) and to 0.4 ppm × 1 h in humans (34).

In studies designed to prove the cascade theory, we have shown that LOP, which are known to be formed in lung lavage (31, 32) during ozone exposure, activate PLA₂ and PLC in cultured human airway epithelial cells (35). The observed effects of LOP on phospholipase activation were similar to those induced during exposure of airway epithelial cells to ozone itself (16). Furthermore, different LOP have strikingly different effects (35).

We here report that specific LOP, obtained during ozonation of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) and previously shown to activate either PLA₂ or PLC (35), cause the release of mediators (PGE₂, PAF, IL-6, and IL-8) that are thought to play an important role in ozone- induced inflammation. In vitro exposure of cultured human airway epithelial cells to a specific LOP, either 1-palmitoyl-2- (9-oxononanoyl)-sn-glycero-3-phosphocholine (PC-ALD), a potent PLA₂ activator, or 1-hydroxy-1-hydroperoxynonane (HHP-C9), a PLC activator, in biologically relevant concentrations and exposures, results in the release of PAF, PGE₂, IL-6, and IL-8.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Materials

Primary human bronchial epithelial (NHBE) cells, bronchial epithelial growth medium (BEGM), keratinocyte growth medium (KGM), and Hanks' buffered saline solution (HBSS) containing 30 mM HEPES were obtained from Clonetics (San Diego, CA). The human bronchial epithelial cell line BEAS-2B was purchased from American Type Culture Collection (Rockville, MD). Excyte-Vle, a fatty acid growth supplement derived from the lipid fraction of adult bovine sera was purchased from Bayer (Kanakee, IL); ³H-lyso platelet-activating factor, ³H-lysoPAF [1-O-(³H)octadecyl-2-hydroxy-sn-glycero-3-phosphocholine] (150 Ci/mmol), was purchased from Amersham (Arlington Heights, IL); POPC, 1-O-hexadecyl- 2-acetoyl-sn-glycero-3-phosphocholine (PAF), and 1-O-hexadecyl-2-hydroxy-sn-glycero-3-phosphocholine (lysoPAF) were purchased from Avanti Polar Lipids (Alabaster, AL). Phorbol-12-myristate-13-acetate (PMA), calphostin C, sphingosine, staurosporine, calcium ionophore (A23187), U73122 {1-[6-((17b-3-methoxyestra-1,3,5 (10)-trien-17-yl)amino)hexyl]-2,5-pyrrolidinedione}, U73343, the inactive analogue of U73122, and mepacrine (quinacrine, dihydrochloride) were obtained from Calbiochem (La Jolla, CA). Recombinant human IL-6 and IL-8 enzyme-linked immunosorbent assay (ELISA) kits were purchased from Endogen (Cambridge, MA). PGE₂ enzyme-linked immunoassay (EIA) was obtained from Cayman Chemical (Ann Arbor, MI). Hard layer silica gel thin layer chromatography (TLC) plates with a preadsorbent layer were purchased from Analtech (Newark, DE). Iodine (I₂) was purchased from Sigma Chemical (St. Louis, MO). Glacial acetic acid, dimethyl sulfoxide (DMSO), high purity grade organic solvents and water, and other reagents were obtained from Fisher Scientific (Houston, TX). Ultima-Flow-AF liquid scintillation cocktail was from Packard (Meriden, CT).

Synthesis of LOP

The phosphocholine-derived aldehyde PC-ALD was produced by reducing the POPC Criegee ozonide with Zn/acetic acid (36). The phosphocholine-derived hydroxyhydroperoxide HHP-C9 was prepared in situ from the spontaneous hydrolysis of the corresponding solid stable precursor bis(1-hydroxyalkyl)peroxide. The latter is synthesized by reacting the corresponding aldehyde with H₂O₂ in dichloromethane. The solvent and excess H₂O₂ were evaporated under vacuum yielding the bis(1-hydroxyalkyl)peroxide (36). We have previously shown that PC-ALD is the most potent LOP in regard to PLA₂ activation, and that HHP-C9 is a very potent LOP in regard to PLC activation (35).

Human Bronchial Epithelial Cells

The human bronchial epithelial cell line (BEAS-2B) was maintained in KGM supplemented with 0.1 ng/ml human recombinant epidermal growth factor (hEGF), 5 µg/ml insulin (bovine), 0.5 µg/ml hydrocortisone, 15 µg/ml bovine pituitary extract (BPE), 0.15 mM calcium, 50 µg/ml gentamicin, and 50 ng/ml amphotericin B. BEAS-2B cells were plated on 24-mm collagen-coated membranes (0.45 µm pore size, 6-well format; Transwell-Col; Costar, Cambridge, MA) at a density of 3 to 5 × 10⁵ cells/4.552 cm². The cells were incubated at 37° C in a 95% air, 5% CO₂ environment with 100% humidity, with 1 ml KGM apically and 3 ml basolaterally, and the medium was changed every other day. Cells became confluent in 72 h and were then cultured for an additional 48 h to promote polarization (13). Passages 39 through 64 were used in the studies. The cell line BEAS-2B has been used extensively as a relevant model of airway epithelial cells (13, 14, 18). This cell line was derived by transformation of human bronchial epithelial cells with an Ad 12-SV40 construct (37). These cells retain characteristics of primary epithelial cells that include polygonal appearance, formation of tight junctions, polarization, and squamous differentiation under serum-containing growth conditions (38). BEAS-2B cells also release a number of mediators in response to agonists such as histamine (16) and release IL-8, IL-6, and AA in response to ozone exposure (13, 14, 18). Samet and coworkers (13) have shown that BEAS-2B cells synthesize and release PAF in response to ozone similar to primary human bronchial epithelial cells.

NHBE were utilized in the PGE₂ studies. NHBE cells were maintained in BEGM supplemented with 0.5 ng/ml hEGF, 0.15 mM calcium, 0.5 µg/ml hydrocortisone, 5 µg/ml insulin, 10 µg/ml transferrin, 0.5 µg/ml epinephrine, 6.5 ng/ml triiodothyronine, 52 µg/ml BPE, 0.1 ng/ml retinoic acid, 50 µg/ml gentamicin, and 50 ng/ml amphotericin B. NHBE cells were plated on 35-mm-diameter, 6-well format plastic plates (Costar, Cambridge, MA) at a density of 4 to 6 × 10³ cells/cm². Cells were incubated at 37° C in a 95% air, 5% CO₂ environment with 100% humidity. The growth medium was changed every other day and cells became confluent in 9 d. Passages 2 through 5 were used in the studies.

Measurement of PAF Release

Confluent BEAS-2B cells were labeled with ³H-lysoPAF (2 µCi/ml) in KGM containing 0.25% bovine serum albumin (BSA; fatty acid free) for 2 h. Under these conditions the cells incorporated an average of 37% of the label. After labeling, cells were washed twice with HBSS and were treated apically with PC-ALD at 10 µM for 1 h in the presence or absence of mepacrine (1 mM, with 30 min preincubation), a known PLA₂ inhibitor (16). In these studies, since the cells are grown on collagen-coated membranes that are soluble in organic solvents and would interfere with intracellular PAF measurements, only extracellularly released PAF was measured (13, 16). After exposure, the apical and basolateral media were collected and the apical medium was centrifuged at 1,000 × g for 10 min at 4° C to pellet the cell debris. Lipids were extracted with two volumes of chloroform:methanol (2:1) in the presence of 50 µg unlabeled PAF as carrier, followed by a second extraction with chloroform. Chloroform phases were pooled together and dried under vacuum (Speed Vac; Savant, Farmingdale, NY). Lipids were suspended in 30 µl of chloroform and separated on silica gel plates with chloroform:methanol:water:acetic acid (100:50:16:8) as the developing solvent system in a preequilibrated TLC chamber. Identity of the released PAF was confirmed by comigration with authentic unlabeled PAF standards on silica. Lipids were visualized with I₂ vapor and the PAF band scraped. Silica-bound PAF was desorbed by 1 ml of water:methanol (1:1) and the radioactivity was counted in a liquid scintillation counter (LSC 6500; Beckman Instruments, Fullerton, CA) and reported as extracellular PAF release (16).

Measurement of PGE₂ Production

We used NHBE cells in the PGE₂ studies because of reported data demonstrating the presence of cyclooxygenase activity (39), and the ability to produce PGE₂ (17). In the NHBE cells compared with the BEAS-2B, since protease digestion during isolation of cells and prolonged stimulation with agonists also may cause depletion in AA stores (40) that may alter eicosanoid production profile, we cultured the NHBE cells under fatty acid-enriched conditions as reported by others (41). Two days before reaching full confluence, NHBE cells were incubated in BEGM containing fatty acid growth supplements (Excyte-Vle, 21 µg/ml) for 48 h at 37° C in a 95% air and 5% CO₂ environment. Confluent cultures were washed twice with HBSS and then treated with vehicle (HBSS containing 30 mM HEPES) alone, vehicle containing PC-ALD at 1, 10, or 100 µM or vehicle containing calcium ionophore A23187 (10 µM) for 4 h. DMSO (0.3%) was used as a carrier for both PC-ALD and A23187. After treatment, PGE₂ levels were determined in the cell-free supernatants diluted with excess HBSS (100-fold) using EIA (Cayman Chemical, Ann Arbor, MI) following the manufacturer's instructions. The PGE₂ EIA kit has a lower sensitivity limit of 14.9 pg/ml (at 4° C), and is specific for PGE₂.

Measurement of IL-8 and IL-6 Release

To measure the effect of LOP or other agonists on IL-8 production, confluent monolayers of BEAS-2B cells, grown on Transwell-Col membranes, were exposed at 37° C to either vehicle alone, or to vehicle containing individual LOP compounds or PMA in the presence or absence of PLC or protein kinase C (PKC) inhibitors. After the various treatments, the amount of IL-8 released to the apical or basolateral compartments was quantified using a recombinant human IL-8 ELISA (Endogen) according to the manufacturer's directions. IL-8 protein concentrations in the culture supernatants were calculated from corresponding absorbances measured at 450 nm using a Bio-Tek EL311 autoplate reader (Bio-Tek, Winooski, VT) and standard calibration curves. The ELISA kit has a lower limit of detection of 5 pg/ ml, is specific for IL-8, and does not cross react with IL-6 or other interleukins.

IL-6 release into the culture supernatants following various exposure regimens was measured using a recombinant human IL-6 ELISA kit obtained from Endogen (Cambridge, MA) following the manufacturer's instructions. The ELISA kit has a lower limit of detection of 7 pg/ml, is specific for IL-6, and does not cross react with IL-8 or other interleukins.

Statistical Analysis

Data are expressed as mean ± standard error of the mean (SEM). Comparisons of PAF, PGE₂, IL-6, and IL-8 production between LOP-exposed and control groups were carried out using one-way analysis of variance (ANOVA) (42).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Induction of PAF Release by PLA₂-activating LOP

The effect of PC-ALD, a known PLA₂ activator, on PAF release was studied. Confluent BEAS-2B monolayers (prelabeled with ³H-lysoPAF) were exposed to 10 µM of PC-ALD for 1 h. As shown in Figure 1, PC-ALD exposure induced a significant 113 ± 11% increase in the amounts of ³H-PAF released apically, compared with vehicle-exposed controls (n = 6, p < 0.05). There was no significant change in the levels of ³H-PAF released basolaterally between PC-ALD-exposed cells and cells exposed to vehicle alone (data not shown).

View larger version (24K): [in this window] [in a new window]   Figure 1.   Effect of the 9-oxononanoyl (PC-ALD) derivative of ozonized POPC on PAF release in BEAS-2B cells. Confluent monolayers of BEAS-2B prelabeled apically with 3H-lysoPAF (2 µCi/ml in KGM with 0.25% BSA [fatty acid free] for 2 h) were treated with vehicle (HBSS containing 30 mM HEPES) alone or vehicle containing PC-ALD (10 µM) in the presence or absence of mepacrine (1 mM with 30 min preincubation) for 1 h and PAF released extracellularly was measured. Data (percent change from controls) of PAF released apically are shown as mean ± SEM (n = 6, PC-ALD-treated group; n = 4, PC-ALD exposure in the presence of mepacrine). Asterisk indicates significant difference from control values (p < 0.05).

To determine whether the observed increase in ³H-PAF release was due to PC-ALD-induced activation of PLA₂, BEAS-2B cells were exposed to PC-ALD (10 µM × 1 h) in the presence of the known PLA₂ inhibitor, mepacrine. As shown in Figure 1, mepacrine (1 mM with 30 min preincubation) significantly inhibited PC-ALD-induced ³H-PAF release (90 ± 6% inhibition, n = 4, p < 0.05) (Figure 1), suggesting that the PC-ALD-induced PAF release was PLA₂-dependent.

Effect of PLC-activating LOP on PAF Release

To determine whether PLC activation was involved in PAF release, confluent monolayers of BEAS-2B (prelabeled with ³H-lysoPAF [2 µCi/ml] for 2 h in KGM containing 0.25% BSA) were exposed to HHP-C9 (an LOP that specifically activates PLC) for 1 h and ³H-PAF release was measured. As shown in Figure 2, at the 10 µM concentration level, HHP-C9 exposure induced a significant 134 ± 40% increase in ³H-PAF levels in the apical compartment (n = 4, p < 0.01). There was no significant difference in the levels of ³H-PAF released basolaterally between HHP-C9-exposed and vehicle-exposed cells (mean difference, 11 ± 6%; n = 4, p = not significant [NS]).

View larger version (31K): [in this window] [in a new window]   Figure 2.   Effect of the hydroxyhydroperoxide derivative of ozonized POPC, HHP-C9, on PAF release in BEAS-2B cells. Confluent BEAS-2B cells prelabeled apically with 3H-lysoPAF (2 µCi/ml) in KGM containing 0.25% BSA (fatty acid free) for 2 h were treated apically with vehicle (HBSS containing 30 mM HEPES), or HHP-C9 (10 µM) in the presence or absence of the PLC inhibitor U73122 (5 µM with 30 min pretreatment), or the PKC inhibitor calphostin C (calph, 1 µM, light-activated; 30 min preincubation) for 1 h, and PAF released extracellularly was measured. Data (percent change from controls) of PAF released apically are presented as mean ± SEM (n =4, each group). Asterisk indicates significant difference from control values (p < 0.05).

To determine whether HHP-C9 exerted its effect through PLC activation, BEAS-2B cells were exposed for 1 h to 10 µM of HHP-C9 in the presence or absence of the known PLC inhibitor U73122 (5 µM with 30 min pretreatment). As demonstrated in Figure 2, U73122 pretreatment significantly inhibited the HHP-C9-dependent ³H-PAF release (86 ± 6% inhibition, n = 4, p < 0.05).

The role of PKC in HHP-C9-induced PAF release also was determined. As shown in Figure 2, calphostin C (1 µM), a selective PKC inhibitor, inhibited HHP-C9-dependent PAF release by 74 ± 17% (n = 4, p < 0.01) in BEAS-2B cultures.

Induction of PGE₂ Release by PLA₂-activating LOP

LOP-induced PGE₂ release was measured after 4 h of exposure to PC-ALD. Exposure of NHBE cultures (enriched with fatty acid supplements) to the calcium ionophore A23187 (10 µM) for 4 h resulted in a 620 ± 101% increase in PGE₂ release above vehicle-exposed cultures (n = 4, p < 0.01). As shown in Figure 3, exposure to PC-ALD for 4 h at 10 µM induced a significant 127 ± 21% increase in PGE₂ release (n = 6, p < 0.05). PC-ALD at either 1 µM or at 100 µM did not induce any significant increase in PGE₂ release (11 ± 5% and 9 ± 4% increase, respectively, n = 6, p = NS). Cell cultures also were exposed to PC-ALD (10 µM × 4 h) in the presence of mepacrine (1 mM, with 30 min preincubation). PC-ALD-induced PGE₂ release was significantly inhibited by mepacrine pretreatment (96 ± 1% inhibition, n = 3, p < 0.01). These data suggest that the observed PC-ALD-induced PGE₂ release was PLA₂-dependent. Furthermore, exposure of NHBE cultures (enriched with fatty acid supplements) to either PC-ALD (10 µM), or to A23187 (10 µM), for 1 h did not induce any detectable amounts of PGE₂ (data not shown), similar to control cultures.

View larger version (24K): [in this window] [in a new window]   Figure 3.   Effect of the 9-oxononanoyl derivative of ozonized POPC (PC-ALD) on PGE2 release in primary cultures of NHBE cells. Two days before reaching full confluence, NHBE cells were incubated in growth medium containing fatty acid growth supplements. Confluent cultures were washed twice with HBSS and treated with vehicle (HBSS) alone or vehicle containing PC-ALD at 1, 10, or 100 µM for 4 h. In other experiments, treatment of BEAS-2B cultures with PC-ALD (10 µM × 4 h) in the presence of mepacrine (1 mM, 30 min preincubation) was also performed. Cell-free supernatants were diluted with excess HBSS (100-fold) and PGE2 release was determined using EIA. Data (percent change from controls) are shown as mean ± SEM (n = 6, each group). Asterisk indicates significant difference from control or other treatment groups (p < 0.05).

In contrast, NHBE cells grown in standard BEGM without any fatty acid enrichment did not release any significant amounts of PGE₂ when exposed for 4 h to either PC-ALD (10 µM) or to A23187 (10 µM) (27 ± 6 pg/ml and 29 ± 8 pg/ml, respectively, n = 6, p = NS), compared with control cultures (22 ± 4 pg/ml, n = 6). Thus, under growth conditions where fatty acids are not depleted, PLA₂-stimulatory LOP can induce the production of PGE₂.

Effect of PLA₂-activating LOP on IL-8 Production

To determine whether PLA₂-activating LOP had any effect on IL-8 synthesis and release, confluent monolayers of BEAS-2B cells were exposed to PC-ALD (100 µM × 4 h) and IL-8 release was measured by ELISA. Exposure of epithelial cell cultures to PC-ALD had no significant effect on IL-8 release from either the apical or basolateral compartments compared with vehicle-exposed control cultures (average increase of 12 ± 3%, n = 3, p = NS).

Effect of PLC-activating LOP on IL-8 Production

Confluent BEAS-2B monolayers were exposed for 1 or 4 h to HHP-C9, a PLC-stimulatory LOP. After HHP-C9 exposure, IL-8 release was measured in the cell-free supernatants by ELISA. The results are shown in Figure 4. Increased apical release of IL-8 was found only after exposure to 100 µM of HHP-C9 for 4 h (101 ± 23% increase above control cultures, n = 8, p < 0.05). HHP-C9 at the 100 µM level for 1 h or at the 10 µM exposure level for 1 or 4 h did not result in IL-8 release from either the apical or basolateral compartments (average of 17 ± 6%, 12 ± 4%, and 11 ± 4% increase, respectively, n = 6, p = NS). No significant difference in the amounts of IL-8 released to the basolateral compartment was found after exposure to 100 µM HHP-C9 (data not shown).

View larger version (26K): [in this window] [in a new window]   Figure 4.   Induction of IL-8 production by HHP-C9 in BEAS-2B cells. Confluent monolayers of BEAS-2B cells were treated with vehicle (HBSS with 30 mM HEPES) alone, or vehicle containing HHP-C9 (10 or 100 µM) in the presence or absence of the PLC inhibitor U73122 (5 µM with 30 min pretreatment) for 4 h, and IL-8 production was measured using human recombinant IL-8 ELISA. Data (percent change from controls) of apically released IL-8 are shown as mean ± SEM (n = 8, each of control and HHP-C9-treated groups; n = 4, inhibitor group). Asterisk indicates significant difference from control values (p < 0.05).

Role of PLC in Regulating LOP-induced IL-8 Release

To assess whether the observed increase in IL-8 release induced by HHP-C9 exposure was dependent on PLC activation, the effect of U73122, a known PLC inhibitor, was studied. As shown in Figure 4, preincubation of the cell cultures with U73122 (5 µM with 30 min preincubation) significantly inhibited the HHP-C9-dependent increase in apical IL-8 release during a 4-h exposure (93 ± 4% inhibition, n = 4, p < 0.05). These data suggest that HHP-C9-dependent IL-8 release involves PLC-dependent activation pathways.

Role of PKC in Regulating LOP-induced IL-8 Release

In preliminary experiments, we first determined if activation of PKC induced IL-8 release in BEAS-2B cell cultures. Confluent BEAS-2B monolayers were exposed for 4 h to a known PKC activator, PMA (0.1 µM), in the presence or absence of the PKC inhibitors calphostin C (1 µM), sphingosine (1 µM), or staurosporine (0.01 µM). As shown in Figure 5, there was a significant increase in the concentrations of IL-8 released into the apical compartment after a 4-h exposure to PMA (1,064 ± 213% increase in IL-8 release compared with vehicle-exposed cells, n = 6, p < 0.05). In contrast, there was no significant difference in the amounts of IL-8 released to the basolateral compartment between PMA-treated and untreated cells (21 ± 11% increase, n = 6, p = NS, data not shown).

View larger version (38K): [in this window] [in a new window]   Figure 5.   Effect of PMA, a PKC activator on IL-8 production in BEAS-2B cells. Confluent BEAS-2B monolayers were treated with vehicle (HBSS containing 30 mM HEPES) alone or vehicle containing PMA (0.1 µM) in the presence or absence of sphingosine (sph, 1 µM), staurosporine (str, 0.01 µM), or calphostin C (calph, 1 µM, light-activated) for 4 h and IL-8 release was measured using human recombinant IL-8 enzyme-linked immunosorbent assay. Data (percent change from controls) are presented as mean ± SEM (n = 6, PMA-treated groups; n = 4, PMA-treated in the presence of inhibitors). Asterisk indicates significant difference from control or inhibitor values (p < 0.05).

BEAS-2B monolayers were treated with PMA at 0.1 µM for 4 h in the presence or absence of calphostin C (1 µM with 30 min preincubation), sphingosine (1 µM, 30 min preincubation), or staurosporine (0.01 µM, 30 min preincubation). As shown in Figure 5, both calphostin C and sphingosine significantly inhibited the PMA-dependent IL-8 release (96 ± 28% and 89 ± 7% inhibition, respectively, n = 4, p < 0.05). In contrast, PMA-induced IL-8 release was not inhibited in the presence of staurosporine (29 ± 8% inhibition, n = 4, p = NS).

Prolonged (18-h) incubation with PMA (0.1 µM) resulted in the loss of the ability of PMA to stimulate any significant increase in IL-8 production compared with resting cells (19 ± 8% increase above control values, n = 4, p = NS). These data suggest downregulation of IL-8 production as a result of PKC depletion secondary to prolonged incubation with PMA.

To determine whether the observed increase in IL-8 release induced by HHP-C9 was dependent on activation of PKC, confluent BEAS-2B monolayers were treated with 100 µM of HHP-C9 for 4 h in the presence or absence of calphostin C (1 µM), sphingosine (1 µM, 30 min preincubation), or staurosporine (0.01 µM, 30 min preincubation). As shown in Figure 6, exposure of BEAS-2B cells to 100 µM HHP-C9 resulted in a significant 101 ± 23% increase in IL-8 release (n = 8, p < 0.05). Similar to the results obtained using PMA, calphostin C and sphingosine significantly inhibited the HHP-C9-dependent IL-8 release (84 ± 12% and 89 ± 4% inhibition, respectively, n = 6, p < 0.05), whereas HHP-C9-induced IL-8 release was not inhibited in the presence of staurosporine (71 ± 21% increase over untreated cultures, n = 6, p < 0.01).

View larger version (39K): [in this window] [in a new window]   Figure 6.   Effect of PKC inhibitors, sphingosine, calphostin C, or staurosporine on IL-8 production induced by HHP-C9 in BEAS-2B cells. Confluent BEAS-2B monolayers were treated with vehicle (HBSS containing 30 mM HEPES) alone or vehicle containing HHP-C9 (100 µM) in the presence or absence of sphingosine (sph, 1 µM with 30 min pretreatment), calphostin C (calph, 1 µM, light-activated and 30 min pretreatment), or staurosporine (str, 0.01 µM with 30 min pretreatment) for 4 h and IL-8 release was determined using human recombinant IL-8 enzyme-linked immunosorbent assay. Data (percent change from controls) of IL-8 released apically are shown as mean ± SEM (n = 8, HHP-C9-treated group; n = 6, HHP-C9-treated in the presence of inhibitors). Asterisk indicates significant difference from control or inhibitor values (p < 0.05).

Effect of PLC-activating LOP on IL-6 Release

Exposure of BEAS-2B cultures to PMA (0.1 µM × 4 h) resulted in a significant 1,022 ± 247% increase in apical IL-6 release (n = 4, p < 0.01, Figure 7). There were no detectable amounts of IL-6 found in the basolateral supernatants. Similar to IL-8 inhibition data (Figure 4), calphostin C (1 µM) significantly inhibited PMA-induced IL-6 release (81 ± 22% inhibition, n = 4, p < 0.05, Figure 7), which was not inhibited by either 1 µM of sphingosine (11 ± 6% increase, n = 4, p = NS) or 0.01 µM of staurosporine (14 ± 7% increase, n = 4, p = NS).

View larger version (28K): [in this window] [in a new window]   Figure 7.   Effect of calphostin C, a PKC inhibitor, on IL-6 production induced by PMA in BEAS-2B cells. Confluent monolayers of BEAS-2B were treated wtih vehicles (HBSS containing 30 mM HEPES) alone or vehicle containing PMA (0.1 µM) in the presence or absence of calphostin C (calph, 1 µM, light-activated and 30 min pretreatment) for 4 h, and IL-6 release was determined using human recombinant IL-6 enzyme-linked immunosorbent assay. Data (percent change from controls) of apically released IL-6 are expressed as mean ± SEM (n = 4, each group). Asterisk indicates significant difference from control or inhibitor values (p < 0.01).

To study whether PLC-activating LOP had any effect on IL-6 release, confluent monolayers of BEAS-2B cells were exposed to HHP-C9 (100 µM × 4 h). As shown in Figure 8, HHP-C9 induced a significant 178 ± 23% increase in the amount of IL-6 released to the apical compartment (n = 6, p < 0.05). There was no significant increase in the amounts of IL-6 released basolaterally compared with unexposed control cultures (13 ± 6% increase, n = 6, p = NS). In the presence of the PLC inhibitor U73122 (5 µM with 30 min preincubation), IL-6 release was significantly inhibited after exposure to HHP-C9 (91 ± 11% inhibition, n = 6, p < 0.05, Figure 8).

View larger version (32K): [in this window] [in a new window]   Figure 8.   Effect of HHP-C9 on IL-6 release in BEAS-2B cells. Confluent BEAS-2B monolayers were treated with vehicle (HBSS containing 30 mM HEPES) alone or vehicle containing HHP-C9 (100 µM) in the presence or absence of the PLC inhibitor U73122 (5 µM with 30 min pretreatment) or the PKC inhibitor calphostin C (calph, 1 µM, light-activated and 30 min pretreatment) for 4 h, and IL-6 release was determined using a human recombinant IL-6 enzyme-linked immunosorbent assay. Data (percent change from controls) of apically released IL-6 are presented as mean ± SEM (n = 6, each group). Asterisk indicates significant difference from control or inhibitor values (p < 0.05).

Exposure of BEAS-2B cultures to HHP-C9 (100 µM × 4 h) in the presence of the selective PKC inhibitor, calphostin C at 1 µM resulted in a significant 71 ± 19% inhibition in HHP-C9-provoked IL-6 release (n = 4, p < 0.05) (Figure 8). In contrast, neither sphingosine (1 µM) nor staurosporine (0.01 µM) had any inhibitory effect on HHP-C9-induced IL-6 release (data not shown).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Exposure to ozone can induce rapid cellular and biochemical changes in the lung, as well as decrements in pulmonary function (11). It is well established that human subjects (both normal and asthmatics) exposed to ozone acutely develop lung inflammation as evidenced by increases in BAL fluid of several proinflammatory mediators including PGE₂, IL-8, and IL-6 (3, 4, 9, 10, 43).

The early and acute stage of inflammation that occurs during ozone exposure is suggested to be, in part, a direct effect of ozone on pulmonary cells (11). Airway epithelial cells have been implicated as effector cells, responding to ozone by releasing a variety of proinflammatory mediators. The basis for this suggestion includes data demonstrating that PGE₂, PAF, IL-8, and IL-6, many of which have been found in BAL of humans exposed to ozone, have also been shown to be induced in cultured airway epithelial cells in response to ozone exposure in vitro (13, 14, 18). The cellular and molecular mechanisms underlying ozone-induced epithelial cell-dependent lung inflammation are not fully clear.

A cascade theory to explain ozone-induced inflammation has been proposed; this theory suggests that LOP are produced as ozone enters the upper tissue layer of the air/tissue boundary, and these LOP would then activate cell signaling elements leading to mediator formation and release (20). We have already shown that LOP, at biologically relevant concentrations, activate PLA₂ and PLC in cultured human airway epithelial cells (35) similar to ozone exposure itself (16).

We believe that the LOP concentrations used in this study are physiologically relevant. LOP species are lipophilic and during incubation with BEAS-2B cultures would partition in the BEAS-2B monolayer. We used the method by Hansch and Leo (see reference 35 for details) to approximate the octanol- water partition coefficient (K_ow) for HHP-C9 (log₁₀ K_ow = 0.477 ± 0.12). The diameter of the BEAS-2B cell (20 µm) was used to compute the volume of the monolayer (2 × 10⁶ cells/ well) relative to the volume of the incubation medium (1 ml/ well). Using this approach, about 1 nmol of HHP-C9 was estimated to partition into and interact with the BEAS-2B monolayer. Human bronchial epithelial cells contain about 4 to 5 nmol of linoleic acid and 2 to 6 nmol of AA per 10⁶ cells (22). Assuming 1 to 10% of inhaled ozone reacts with unsaturated fatty acids in membrane phospholipids, LOP would be formed in the range of 0.1 to 1 nmol per 10⁶ cells. Therefore, the concentrations of LOP used in our studies should closely approximate the levels of LOP that actually would be formed in vivo during ozone exposure.

In the present study, our goal was to determine whether LOP-induced activation of PLA₂ and PLC is involved in the induction of proinflammatory mediators associated with ozone exposure. We studied the effect of two LOP, PC-ALD and HHP-C9, previously demonstrated to be potent activators of PLA₂ and PLC, respectively (35), on PAF, PGE₂, IL-6, and IL-8 release. These mediators have been shown to be released by the airway epithelium in response to ozone exposure (13, 14, 16, 18, 43). We believe that the 4-h time point was optimal to examine increases in the products studied based on the following data: (1) Jaspers and coworkers (19) have shown that ozone-induced IL-8 messenger RNA (mRNA) expression in human airway epithelial cells (A549) was maximal after 4 h of exposure to ozone; (2) in preliminary studies, exposure of the cells to the HHP-C9 for 1 h did not induce any significant increase in IL-8 release (compared with 4 h); (3) Newton and coworkers (44) have shown maximal inducible cyclooxygenase mRNA expression after 4 h of exposure of human airway epithelial cells (A549) to IL-1 and other cytokines (e.g., tumor necrosis factor-alpha [TNF-]); and (4) exposure of the cells in our studies to the PC-ALD product for 1 h did not induce any significant increase in PGE₂, whereas PGE₂ production was significantly increased after 4 h of exposure to PC-ALD.

PAF is a potent proinflammatory lipid mediator (45, 46). PAF synthesis is catalyzed, in part, by activated PLA₂ (47). Tramposch and coworkers (48) have demonstrated that specific inhibition of PLA₂ (without any effect on other lipases) inhibited PAF biosynthesis in human polymorphonuclear leukocytes. Our data demonstrating increased release of PAF (Figure 1) by PC-ALD, an LOP known to activate cytosolic PLA₂ in BEAS-2B cells (35), suggest that specific LOP can induce PAF release. The effect of PC-ALD on PAF release was significantly inhibited by mepacrine, a known PLA₂ inhibitor, further supporting the hypothesis that PLA₂ is activated by PC-ALD (Figure 1). The data in the present study also correlate with previously reported in vitro ozone exposure studies. Samet and coworkers (13) demonstrated a significant increase in PAF release in the cell line BEAS-2B, and in cultures of primary human bronchial epithelial cells during exposure to ozone. Wright and coworkers (16) also showed induction of PAF release during ozone exposure using primary guinea pig tracheal epithelial cells.

We previously demonstrated that exposure to HHP-C9 (10 µM for 30 min) activates PLC and causes significant turnover of inositol phosphates (35). In the present study, activation of PLC by the HHP-C9 also contributed to PAF release. Further evidence for LOP-induced PLC activation is shown by the data demonstrating that the HHP-C9-induced PAF release was significantly inhibited by the known PLC inhibitor U73122 (Figure 2). PKC involvement in PAF synthesis has been suggested (49). The inhibitory effect of calphostin C, a selective PKC inhibitor on HHP-C9-provoked PAF release (Figure 2), demonstrates that PKC plays a regulatory step in LOP-induced PAF production. McIntyre and coworkers (50) showed enhanced A23187-induced PAF biosynthesis by the PKC activator PMA. Activators of PKC have been reported to inhibit the acylation of AA into intracellular phospholipid pools (51), resulting in PAF precursor availability.

PLA₂ is a critical enzyme in the liberation of AA, a substrate for and thus a rate-limiting step in prostanoid synthesis, including PGE₂ (52, 53). Suppression of cytosolic-type PLA₂ (85 kD cPLA₂) inhibits prostaglandin production in human monocytes (54). Although cPLA₂ is predominantly responsible for prostaglandin biosynthesis owing to its selectivity for sn-2 arachidonoyl, specific inhibition of secretory PLA₂ (sPLA) decreases PGE₂ synthesis in renal kidney cells (55). Similar findings have been reported by Miyake and coworkers (56) in mouse macrophage. In the present study, a PLA₂-stimulatory LOP, PC-ALD, significantly induced the production of PGE₂ in NHBE cells. The LOP-induced increase in PGE₂ is similar to data reported by McKinnon and coworkers (14) using the BEAS-2B cell line exposed to 0.1 to 1.0 ppm ozone. The observed increase in PGE₂ release by airway epithelial cells in response to PC-ALD could, in part, explain the observed increase in PGE₂ concentrations in BAL from both normal (3- 5) and asthmatic subjects (7, 10) exposed to ozone in vivo.

The data in the present study supporting the concept that LOP activates PLA₂ are consistent with the data reported from studies examining the effect of ozone exposure on PLA₂ in airway epithelial cells (16, 35). In these studies, ozone- induced activation of PLA₂ was inhibited by mepacrine. In the present study, we also found that LOP can activate PLA₂ as indicated by the inhibition of PC-ALD-mediated PGE₂ release by mepacrine (Figure 3). However, PC-ALD at the 100-µM exposure level, a concentration shown previously to activate PLA₂ (35), did not induce any significant PGE₂ release. Although it is possible that this could be the result of different sampling times, inhibition of the inducible cyclooxygenase enzyme at this concentration also is possible.

It has been suggested that oxidant stress is a specific and important regulator of IL-8 gene expression serving to recruit neutrophils to sites of inflammation (57). In the present study, IL-8 production by cultured human airway epithelial cells was selectively induced by exposure to the PLC-stimulatory LOP HHP-C9 but not by exposure to the PLA₂-stimulatory LOP PC-ALD. This selectivity is confirmed by the PLC inhibitor results (Figure 4). The LOP-induced increases in IL-8 release are similar to the data of Devlin and coworkers (18), who demonstrated a significant increase in apical IL-8 release in BEAS cells exposed to 0.1 to 1.0 ppm ozone in vitro. Jaspers and coworkers (19) also demonstrated increased release of IL-8 during low-level in vitro ozone exposure using a type II-like epithelial cell line, A549. The similarity between the responses of cells toward ozone itself and the responses to HHP-C9 in the present study supports the hypothesis that LOP induce responses similar to ozone.

Published data are limited on PLC and PKC signaling pathways involved in IL-8 production in human airway epithelial cells. PKC activation has been shown to play a role in IL-8 synthesis in human keratinocytes (58). However, in other cells, IL-8 biosynthesis was shown to be regulated by a protein tyrosine kinase pathway (59, 60). In BEAS-2B cultures, LOP-induced IL-8 production involves a PKC-dependent pathway (Figures 5 and 6). The inability of staurosporine to inhibit PMA- or HHP-C9-provoked IL-8 release can be attributed to the poor selectivity of staurosporine toward PKC (61) as compared with that of calphostin C (62, 63). Furthermore, the inhibition of IL-8 release by sphingosine, but not by staurosporine, after exposure to either PMA or HHP-C9 suggests that only certain PKC isoforms may be involved in IL-8 release (61). The specific PKC isotypes involved in the LOP-induced IL-8 production are not known and remain to be characterized. However, Jaspers and coworkers (60) have shown that increased IL-8 concentrations in A549 cells exposed to 0.1 ppm (5 h) ozone were not PKC-dependent. In our studies, exposure of BEAS-2B cultures to 0.1 ppm ozone (1 h) did not induce any increase in IL-8 production (data not shown). This difference between the two studies probably results from differences between the two cell lines used. Furthermore, the ability of HHP-C9 to induce IL-8 production at 100 µM, but not at 10 µM (35), suggests that activation of PKC by LOP signaling pathways in airway epithelial cells may exhibit a concentration threshold.

Increased IL-8 gene expression by ozone exposure in airway epithelial type II-like cells, A549 has been linked to activation of nuclear transcription factor B (NF-B) (19). In preliminary studies, we have found that specific LOP activate NF-B (64) in ozone target cells.

In regard to IL-6, the increases in LOP-induced IL-6 release found in the present study are similar to the data reported during ozone exposure itself by Devlin and coworkers (18). These investigators reported increased apical release of IL-6 after exposure of BEAS-2B cells to 0.1 to 1.0 ppm ozone. Our data also show that, with the exception of the inhibitory effect of sphingosine, the PLC and PKC activation pathways regulating IL-6 biosynthesis in BEAS-2B cultures are similar to the results examining IL-8 (Figures 7 and 8).

In conclusion, the data presented here support previous reports of LOP-induced phospholipase activation. These new data strongly suggest that LOP may be responsible, in part, for the increased production of proinflammatory mediators observed during ozone exposure. Furthermore, various LOP known to be formed during ozone exposure appear to have specificity in regard to signal transduction elements in the epithelial cell. The discrete molecular mechanisms underlying the observed LOP-mediated phospholipase activation and release of proinflammatory mediators by lung cells remain to be further elucidated.