Induction of inflammatory mediators in human airway epithelial cells by lipid ozonation products.

We have proposed that exposure of epithelial cell membrane lipids in the lung (mainly phospholipids) to ozone will generate lipid ozonation products (LOP), which could be responsible for the proinflammatory effects of ozone. The ozonation of phosphocholine, the principal membrane phospholipid, produces a limited number of LOP, including hydroxyhydroperoxides and aldehydes. We now report that exposure of cultured human bronchial epithelial cells to the ozonized 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) product, 1-palmitoyl-2-(9-oxononanoyl)-sn-glycero-3-phosphocholine (PC-ALD), a phospholipase A(2) (PLA(2))-stimulatory LOP, resulted in a 113 +/- 11% increase in the amounts of tritiated platelet-activating factor ((3)H-PAF) released apically. (3)H-PAF release was also induced by 1-hydroxy-1-hydroperoxynonane of ozonized POPC (HHP-C9), a phospholipase C (PLC)- stimulatory LOP (134 +/- 40% increase in (3)H-PAF). PC-ALD at 10 microM, but not HHP-C9, induced a 127 +/- 24% increase in prostaglandin E(2) (PGE(2)) release (n = 6, p < 0.05). In contrast, HHP-C9, but not PC-ALD, induced interleukin (IL)-6 release (178 +/- 23% increase, n = 6, p < 0.05) and IL-8 release (101 +/- 23% increase, n = 8, p < 0. 05). These results suggest that LOP-dependent release of proinflammatory mediators may play an important role in the early inflammatory response seen during exposure to ozone.

Beta carotene and its oxidation products have different effects on microsome mediated binding of benzo[a]pyrene to DNA.

The effects of beta-carotene (betaC) and its oxidation products on the binding of benzo[a]pyrene (BaP) metabolites to calf thymus DNA was investigated in the presence of rat liver microsomes. Mixtures of betaC oxidation products (betaCOP) as well as separated, individual betaC oxidation products were studied. One set of experiments, for example, involved the use of the mixture of betaCOP obtained after a 2-h radical-initiated oxidation. For this data set, the incorporation of unoxidized betaC into microsomal membranes caused the level of binding of BaP metabolites to DNA to decrease by 29% over that observed in the absence of betaC; however, the incorporation of the mixture of betaCOP caused the binding of BaP metabolites to DNA to increase 1.7-fold relative to controls without betaC. Two variations of this experiment were studied: (1) When no NADPH was added, betaC decreased the binding of BaP metabolites to DNA by 19%, but the mixture of betaCOP increased binding by 3.3-fold relative to that observed in the absence of betaC. (2) When NADPH was added under near-anaerobic conditions, betaC caused an almost total (94%) decrease in binding whereas betaCOP had no effect on the amount of binding relative to that observed in the absence of betaC. Both betaCOP and cumene hydroperoxide caused BaP metabolites to bind to DNA even when NADPH was omitted from the incubation mixture. Separation of the mixture of betaC oxidation products into fractions by HPLC allowed preliminary testing of individual betaC oxidation products separately; of the various fractions tested, the products tentatively identified as 11,15'-cyclo-12,15-epoxy-11,12,15,15'-tetrahydro-beta-carotene and beta-carotene-5,6-epoxide appeared to cause the largest increase in BaP-DNA binding. Microsomes from rats induced with 3-methylcholanthrene (3MC) or Aroclor 1254 produced different levels of binding in some experimental conditions. We hypothesize that, under some conditions, the incorporation of betaC into microsomal membranes can be protective against P450-catalyzed BaP binding to DNA; however, the incorporation of betaCOP facilitates the formation of BaP metabolites that bind DNA, although only certain P450 isoforms catalyze the binding process.

Lipid ozonation products activate phospholipases A2, C, and D.

Ozone exposure, in vitro, has been shown to activate phospholipases A2 (PLA2), C (PLC), and D (PLD) in airway epithelial cells. However, because of its high reactivity, ozone cannot penetrate far into the air/lung tissue interface. It has been proposed that ozone reacts with unsaturated fatty acids (UFA) in the epithelial lining fluid (ELF) and cell membranes to generate a cascade of lipid ozonation products (LOP) that mediate ozone-induced toxicity. To test this hypothesis, we exposed cultured human bronchial epithelial cells (BEAS-2B) to LOP (1-100 microM) produced from the ozonation of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphatidylcholine (POPC) and measured the activity of PLA2, PLC, and PLD. The PLA2 isoform responsible for arachidonic acid release (AA) in stimulated cultures was also characterized. Activation of PLA2, PLC, and PLD by three oxidants, hydrogen peroxide (H2O2), tert-butyl hydroperoxide (t-BOOH) and 2,2'-azobis(2-amidinopropane)dihydrochloride (AAPH) also was measured and compared to that of LOP. The derivatives of ozonized POPC at the sn-2 residue, 9-oxononanoyl (PC-ALD), 9-hydroxy-9-hydroperoxynonanoyl (PC-HHP), and 8-(-5-octyl-1,2,4-trioxolan-3-yl-) octanoyl (POPC-OZ) selectively activated PLA2 in a dose-dependent fashion. Cytosolic PLA2 (cPLA2) measured in the cytosolic fraction of stimulated cell lysates was found to be the predominant isoform responsible for AA release. PLC activation was exclusively induced by the hydroxyhydroperoxide derivatives. PC-HHP and the 9-carbon hydroxyhydroperoxide (HHP-C9) increased PLC activity. PLD activity also was induced by LOP generated from POPC. Incubation of cultures with H2O2 alone did not stimulate PLC; however, in the presence of the aldehyde, nonanal, a 62 +/- 2% increase in PLC activity was found, suggesting that the increase in activity was due to the formation of the intermediate HHP-C9. t-BOOH, and AAPH also failed to induce PLA2 activation, but did activate PLC, under conditions of exposure identical to that of LOP. Only t-BOOH activated PLD. These results suggest that biologically relevant concentrations of LOP activate PLA2, PLC, and PLD in the airway epithelial cell, a primary target to ozone exposure. The activation of these phospholipases may play a role in the development of lung inflammation during ozone exposure.

Scavenging of peroxynitrite by a phenolic/peroxidase system prevents oxidative damage to DNA.

We examined the ability of horseradish peroxidase (HRP), an analog of human myeloperoxidase, to protect DNA against oxidative damage caused by peroxynitrite in the presence of chlorogenic acid (CGA), a naturally occurring polyphenol. Chlorogenic acid inhibits the formation of single strand breaks in supercoiled pBR322 DNA by acting as a scavenger of peroxynitrite. Horseradish peroxidase markedly enhances the extent of DNA protection by catalyzing the decomposition of peroxynitrite in the presence of CGA. Horseradish peroxidase alone does not inhibit peroxynitrite-induced DNA strand breaks, indicating that CGA is required as an electron donor to regenerate the active enzyme. The apparent second order rate constant for the HRP-mediated oxidation of CGA in the presence of peroxynitrite at pH 6.9 is 3.4 x 10(7) M(-1) s(-1). This high rate suggests that CGA and other dietary polyphenols might efficiently scavenge peroxynitrite in peroxidase-containing systems in vivo.

Effects of melatonin on lipid peroxidation induced by oxygen radicals.

We here report the activity of the neurohormone melatonin (MLT) as a scavenger of free radicals in two different experimental models: (a) linoleic acid peroxidation initiated by different free radical-generating systems and (b) a multilamellar vesicle system composed of dilinoleoylphosphatidylcholine. In system (a) linoleic acid peroxidation, induced by either the water-soluble initiator 2,2'-azobis (2-amidinopropane) dihydrochloride (ABAP) or Fe2+-EDTA addition to 2.6 mM linoleic acid dispersed in SDS-phosphate buffer, was evaluated as the formation of conjugated dienes, measured spectrophotometrically at 236 nm. MLT did not reduce the rate of peroxidation induced by ABAP, but did reduce, in a concentration-dependent fashion, the rate of the reaction activated by Fe2+-EDTA. In system (b) multilamellar vesicles were used as the substrate for lipid peroxidation, initiated by Fe2+-EDTA and determined by means of malonaldehyde (MDA) and 4-hydroxyalkenal (4-HDA) content. MLT was found to be slightly more effective in system (b) than in the dispersed linoleic acid system (see a). These results show that MLT inhibits lipid damage induced by oxygen free radicals. However, MLT is only about one one-hundredth as effective an antioxidant as vitamin E in the micelles system.

Trolox inhibits peroxynitrite-mediated oxidative stress and apoptosis in rat thymocytes.

Peroxynitrite is a strong oxidant that reacts with a variety of biomolecules in vivo and in vitro. When rat thymocytes in phosphate buffer are exposed to 25 microM peroxynitrite for 10 min, DNA single strand breaks (SSB) can be detected. These SSB are repaired if the cells are incubated in fresh media at 37 degrees C for 120 min. In addition, DNA protein cross-links and apoptosis are observed 1 and 6 h, respectively, after peroxynitrite exposure. Peroxynitrite mediates the formation of thiobarbituric acid-reactive substances (TBARS) that may be responsible for the DNA-protein crosslinks (DPXL). Both TBARS and DPXL formation are lowered by posttreating the cells immediately after the 10-min exposure to peroxynitrite with Trolox, a water-soluble vitamin E analog. These results suggest that oxidative stress mediated by peroxynitrite can trigger a critical sequence of events ending in programmed cell death and that intracellular oxidation is a component of the apoptosis of thymocytes, since both oxidative processes and apoptosis can be prevented by Trolox. In addition to Trolox, we obtained partial data on three other phenolic antioxidants (3-tert-butyl-4-hydroxyanisole, butylated hydroxytoluene, and 2,6-diisopropylphenol). We find that Trolox and these three phenols similarly protect rat thymocytes from apoptosis mediated by peroxynitrite.

Competitive reactions of peroxynitrite with 2'-deoxyguanosine and 7,8-dihydro-8-oxo-2'-deoxyguanosine (8-oxodG): relevance to the formation of 8-oxodG in DNA exposed to peroxynitrite.

We have examined the formation of 7,8-dihydro-8-oxo-2'-deoxyguanosine (8-oxodG) in reactions of peroxynitrite with 2'-deoxyguanosine (dG) and calf-thymus DNA. Peroxynitrite reacts with dG at neutral pH, but this reaction does not result in the buildup of 8-oxodG. We also do not find any evidence for the formation of 8-oxodG in calf-thymus DNA upon exposure to peroxynitrite. When 8-oxodG is mixed with 1000-fold excess dG and then allowed to react with peroxynitrite, about 50% of the 8-oxodG is destroyed. The preferential reaction of 8-oxodG is also evident when dG in calf-thymus DNA is partially oxidized in an Udenfriend system and then allowed to react with peroxynitrite. We suggest that 8-oxodG is not produced in peroxynitrite-mediated oxidations of dG and DNA or that it is produced but then is rapidly consumed in further reactions with peroxynitrite. Oxidized DNA bases frequently can be more oxidation sensitive than their corresponding progenitors and, therefore, may be present at] low steady-state concentrations and not represent stable markers of oxidative stress status. The importance of the 8-oxodG/peroxynitrite reaction is discussed in relation to the formation of more stable, secondary oxidation products that might be more useful markers of DNA damage.

Effect of lipid ozonation products on liposomal membranes detected by Laurdan fluorescence.

We report here the influence of the lipid ozonation products, 1-palmitoyl-2-(9-oxononanoyl)-sn-glycero-3-phosphocholine (PC-aldehyde) and 1-palmitoyl-2[8-(5-octyl-1, 2, 4,-trioxolan-3-yl)- octanoyl]-sn-glycero-3-phosphocholine (PC-Criegee ozonide), on the phase domains of small unilamellar vesicles. (See Scheme 1 for structures.) 6-Lauroyl-2-dimethylaminonaphtalene (Laurdan) fluorescence excitation and emission spectra and generalized polarization measurements allowed us to study how lipid ozonation products affect the phase components of phospholipid membranes. A shift of excitation and emission spectra and a decrease in generalized polarization reveal the presence of a more polar environment surrounding the probe. We find that when either PC-aldehyde or PC-Criegee ozonide are incorporated into a 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) membrane, or when the POPC membrane is directly ozonated, a change in polarity of the phospholipid environment occurs that changes the properties of the bilayer. The introduction of more oxygenated and more polar phospholipids creates a more polar environment allowing the deeper penetration of water molecules into the membrane. Water penetration also is facilitated by the membrane disorder-producing effect of the ozonation products. The presence of an increased number of water molecules in the membrane effects the bilayer, altering packing order and cooperatively among fatty acyl chains as well as enhancing membrane fluidity.

Peroxynitrite causes DNA damage and oxidation of thiols in rat thymocytes [corrected].

We report here the ability of peroxynitrite to cause DNA strand breaks and to oxidize cellular thiol groups in viable rat thymocytes. Peroxynitrite was added to rat thymocytes in a phosphate buffer and DNA damage was measured by the fluorescence analysis of DNA unwinding assay. Peroxynitrite causes DNA strand breaks in a dose-dependent fashion. Four hydroxyl radical scavengers, namely mannitol, dimethyl sulfoxide, sodium benzoate, and Trolox, were tested for their ability to protect DNA from oxidative damage by peroxynitrite. Mannitol failed to protect DNA at concentrations at which it would have conferred nearly complete protection from damage by the hydroxyl radical. Strikingly, dimethyl sulfoxide and benzoate, which are more efficient hydroxyl radical scavengers than mannitol, caused an increase in DNA damage. Trolox was the only scavenger, among the four tested here, that was able to protect DNA from oxidative damage by peroxynitrite. We have previously shown that, among the scavengers tested, Trolox is the most effective scavenger of HOONO*, where HOONO* is a reactive form of HOONO that is a more selective oxidant than is the hydroxyl radical (see W. A. Pryor, X. Jin, and G. L. Squadrito, 1994, Proc. Natl. Acad. Sci. USA 91, 11173-11177). Thus, these results are consistent with our earlier observations that oxidations by peroxynitrite involve a reactive intermediate, HOONO*, rather than hydroxyl radicals. Peroxynitrite also oxidized cellular thiols in a dose-dependent fashion. Greater than 90% of the cells exposed to peroxynitrite were still viable for up to 10 min after DNA damage and thiol oxidation had occurred. In conclusion, DNA damage caused by peroxynitrite can be rationalized as caused by a powerful oxidant, HOONO*, which is formed during the decomposition of peroxynitrite to nitrite.

Peroxynitrite causes apoptosis in rat thymocytes.

Peroxynitrite is demonstrated to cause apoptosis in freshly harvested rat thymocytes suspended in phosphate buffer, as shown by the typical ladder pattern of DNA fragmentation. This evidence suggests a new function for peroxynitrite, already known to be a strong oxidant. Peroxynitrite can nick DNA, oxidize cellular thiols and initiate the apoptotic process responsible for genomic degradation.

Peroxynitrite causes DNA nicks in plasmid pBR322.

Peroxynitrite causes single-strand breaks in pBR322 supercoiled DNA as evidenced by agarose gel electrophoresis analysis. The effect of three free radical scavengers, namely mannitol, benzoate and dimethylsulfoxide, were studied. Mannitol failed to protect DNA from damage by peroxynitrite while benzoate and dimethylsulfoxide amplified the damage. These results suggest the damage caused by peroxynitrite alone is not mediated by free radicals since typical free radical scavengers fail to prevent the damage.

Effect of ozonation products on phospholipase A2 hydrolytic activity: use of bis(1-hydroxyheptyl)peroxide as a precursor of the ozonation product 1-hydroxy-1-hydroperoxyheptane.

Bis(1-hydroxyheptyl)peroxide (BisC7) upon spontaneous hydrolysis affords the difficult-to-isolate ozonation product 1-hydroxy-1-hydroperoxyheptane along with heptanal. PLA2 hydrolytic activity is enhanced when the BisC7 hydrolysis products are incorporated in the membrane of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine liposomes, suggesting they cause a pronounced alteration in the bilayer packing order. Conversely, inhibition is observed when PLA2 is incubated with BisC7 hydrolytic products prior to incubation with the liposomes, suggesting that these products are capable of reacting with and modifying the enzyme when present in solution.

Activation of phospholipase A2 in 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine liposomes containing lipid ozonation products.

The activation of phospholipase A2 (PLA2) by lipid ozonation products is reported. The principal products from the ozonation of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) are 1-palmitoyl-2-[8-[3(5-octyl-1,2,4-trioxolan-3-yl)octanoyl]--sn-gly cero-3- phosphocholine (PC-Criegee ozonide) and 1-palmitoyl-2-(9-oxononanoyl)-sn-glycero-3- phosphocholine (PC-aldehyde). (See Figure 1 for structures.) Here we test the hypothesis that these two compounds are mediators of ozone toxicity. Using POC vesicles, we find that PLA2 recognizes and hydrolyzes PC-Criegee ozonide at the same rate as that of arachidonic acid. Although the PC-aldehyde is not a substrate for PLA2, the enzymatic rate of hydrolysis by PLA2 of unaltered fatty acids incorporated into POPC is enhanced when PC-aldehyde is present in the bilayer. Thus, both PC-Criegee ozonide and PC-aldehyde alter the activity of PLA2, perhaps via an effect on membrane packing order. The capacity of PLA2 to recognize the PC-Criegee, and therefore ozone-induced damage, suggests a detoxification property of the enzyme and also a role in maintaining the structural properties of bilayer membranes that have been altered by exposure to ozone.

Enhanced interfacial catalysis and hydrolytic specificity of phospholipase A2 toward peroxidized phosphatidylcholine vesicles.

The hydrolytic action of phospholipase A2 was examined with unilamellar vesicles composed of soybean phosphatidylcholine in terms of the calcium dependency of the enzyme and substrate specificity following lipid peroxidation. Experiments were performed under conditions where enzyme:substrate ratios were low, specifically in the range of one to five enzyme molecules for every 10 vesicle particles. Accordingly, low hydrolytic activities were found where less than 15% of the phospholipids were hydrolyzed under the various conditions of time, enzyme:substrate ratios, calcium concentrations, and extent of peroxidation utilized. Vesicle peroxidation increased the Ca2+ binding potential to a degree comparable to addition of the anionic phospholipid, dioleoylphosphatidic acid (DOPA). A remarkable similarity was found between the binding profiles for Ca2+ and phospholipase A2 activity; however, enzyme activity toward oxidized vesicles was beyond the increases observed for Ca2+ binding. Under conditions where approximately 5% of the phospholipids were peroxidized the effective Ca2+ concentration required for half-maximal activity was less than one-half that required for unoxidized vesicles. Peroxidation of vesicle phospholipids markedly increased the rate and extent of hydrolysis, even in the presence of DOPA or deoxycholate. Deoxycholate is known to induce vesicle fusion such that a larger proportion of enzyme is associated with a fewer number of enlarged vesicles. Using a dual isotope technique to measure hydrolysis of oxidized vs unoxidized phospholipids and covesicle preparations to study enzyme binding and activity, a significantly greater apparent intervesicle exchange of enzyme was found after peroxidation of vesicles with more than a twofold hydrolytic specificity toward the oxidized phospholipids. We postulate that a combination of structural and Ca2+ binding affinity changes are produced in membranes following lipid peroxidation which evoke an additive effect on PLA2 activity. Although oxidized phospholipids may serve as activators of phospholipase A2 by presenting the interface in a form where Ca2+ and enzyme binding and/or specific activity are increased, an additional and important factor appears to involve membrane fusion or vesicle-vesicle interactions. This process facilitates enzyme activity through the replenishment of substrates wherein the otherwise limited interaction of enzyme and substrate is overcome by more rapid or extensive vesicle fusion which increases access to the phospholipids available in the preparation.

Peroxidation and phospholipase A2 hydrolytic susceptibility of liposomes consisting of mixed species of phosphatidylcholine and phosphatidylethanolamine.

The relationship between lipid peroxidation and phospholipase A2 (PLA2) hydrolytic activity was studied using unilamellar vesicles (liposomes) as model membranes. Hydrolytic specificity was examined using vesicles prepared with pure bovine heart phosphatidylcholine (PC), bovine heart phosphatidylethanolamine (PE), or mixtures of these phospholipids, using two preparative procedures, i.e., sonication or extrusion. Lipid peroxidation was induced by incubating vesicles with cumene hydroperoxide and hematin at 37 degrees C. Determinations of the extent of peroxidation by means of diene conjugate content derived from second derivative spectra or by polarographic measurement of oxygen consumption rates provided a basis for comparing the extent of peroxidation of each phospholipid species to their subsequent hydrolysis by PLA2 (from Crotalus adamanteus). The extent of hydrolysis was determined through the release of arachidonic acid from either PC or PE. The PE distribution among the outer vs. inner leaflet of the membrane bilayer was nearly equal in sonicated vesicles, whereas most of the phospholipid was incorporated into the inner leaflet in extruded vesicles. The proportion of PE found in the inner leaflet progressively increased as the ratio of PE to PC increased in both sonicated and extruded vesicle preparations. Lipid peroxidation had no effect on PE distribution under the conditions examined. There was a clear preference for PC peroxidation for all vesicle compositions tested and PC was preferentially hydrolyzed by PLA2. This effect is proposed to result from a perturbation of membrane structure following peroxidation with assimilation of PC into PLA2-susceptible domains whereas PE peroxidation and hydrolysis is less affected in mixed PC/PE vesicles. Lipid peroxidation imposes an additional hydrolytic susceptibility over the effects exerted through the mixing of these phospholipids which is based on structural changes rather than formation of specific substrates for PLA2.

Conjugated diene and trans fatty acids in a choline-devoid diet hepatocarcinogenic in the rat.

It has been postulated that the hepatocarcinogenicity of a choline-devoid diet in rats stems from peroxidation of liver lipids. We have investigated whether the diet contains conjugated dienes that could account directly for those detected in liver lipids of rats fed a choline-devoid diet. Analyses were performed on samples of corn oil and of a partially hydrogenated fat used to prepare semipurified choline-devoid and choline-supplemented diets, and on fat extracted from two pairs of diets, one set containing 5% corn oil and 10% partially hydrogenated fat, and the other only corn oil (15%). The analyses consisted of quantitation of conjugated dienes by UV spectrophotometry, separation of fatty acids with conjugated dienes by HPLC, and quantitation of trans fatty acids by IR spectrophotometry. Small levels of conjugated diene and trans fatty acids were present in the corn oil, but much higher amounts were found in the partially hydrogenated fat. HPLC analysis yielded distinct elution profiles for the fatty acids with conjugated dienes present in the two fats, and similar results were obtained with fat extracted from the diets. However, no differences were observed between choline-devoid and control choline-supplemented diets. The results indicate that caution must be exercised in interpreting data from UV analysis of tissue lipids of rats fed diets containing a partially hydrogenated fat.

Conjugated diene and trans fatty acids in tissue lipids of rats fed an hepatocarcinogenic choline-devoid diet.

Groups of male Fischer-344 rats were fed two pairs of semi-purified choline-devoid and choline-supplemented diets, one set containing 5% corn oil and 10% partially hydrogenated fat, and the other only corn oil (15%). Analyses were performed on lipids extracted from whole liver, liver nuclei and adipose tissue of the rats. The analyses consisted of quantitation of conjugated dienes by UV spectrophotometry, separation of fatty acids with conjugated dienes by HPLC and quantitation of trans fatty acids by IR spectrophotometry. Conjugated dienes and conjugated diene and trans fatty acids were observed in adipose tissue total lipids, at concentrations that reflected those in the diets fed. The same was true of trans fatty acids in liver lipids. However, no conjugated dienes, or fatty acids with conjugated dienes, were detected in liver lipids of rats fed the diets formulated with only corn oil. In contrast, conjugated dienes were detected in total and neutral lipids, but not in phospholipids, of whole liver and liver nuclei of rats fed the diets formulated with partially hydrogenated fat. The neutral lipids contained fatty acids with conjugated dienes that eluted with the retention time of conjugated diene fatty acids, present in the dietary partially hydrogenated fat. It is concluded that a choline-devoid diet, which is hepatocarcinogenic in the rat, does not induce a peroxidation of liver cell membrane lipids, and that not only trans fatty acids, but also fatty acids with conjugated dienes present in a partially hydrogenated fat, are absorbed and assimilated in rat tissue lipids.