Spontaneous intermembrane transfer of various cholesterol-derived hydroperoxide species: kinetic studies with model membranes and cells.

Whereas spontaneous and protein-mediated transfer/exchange of cholesterol (Ch) between membranes has been widely studied, relatively little is known about the translocation of Ch oxidation products, particularly hydroperoxide species (ChOOHs), which can act as cytotoxic prooxidants. A major aim of the present study was to examine and compare the intermembrane transfer characteristics of several biologically relevant ChOOH isomers, including singlet oxygen-derived 5alpha-OOH, 6alpha-OOH, and 6beta-OOH and free radical-derived 7alpha-OOH and 7beta-OOH. These species were generated in [(14)C]Ch-labeled donor membranes [erythrocyte ghosts or unilamellar DMPC/Ch (1.0:0.8 mol/mol) liposomes] by means of dye-sensitized photoperoxidation. Spontaneous transfer to nonoxidized acceptor membranes (DMPC liposomes or ghosts, respectively) at 37 degrees C was monitored by thin-layer chromatography with phosphorimaging radiodetection (HPTLC-PI) or liquid chromatography with mercury cathode electrochemical detection [HPLC-EC(Hg)]. The former allowed measurement of total (unresolved) ChOOH along with parent Ch, whereas the latter allowed measurement of individual ChOOHs. Ghost membranes in which approximately 4% of the Ch had been peroxidized, giving mainly 5alpha-OOH, transferred total ChOOH and Ch to liposomes in apparent first-order fashion, the rate constant for ChOOH being approximately 65 times greater. Like Ch desorption, ChOOH desorption from donor membranes was found to be rate limiting, and rate varied inversely with size when liposomal donors were used. For individual ChOOHs, rate constant magnitude (7alpha/7beta-OOH > 5alpha-OOH > 6alpha-OOH > 6beta-OOH) correlated inversely with reverse-phase HPLC retention time, suggesting that faster transfer reflects greater hydrophilicity. Liposome-borne ChOOHs exhibited the same order of toxicity toward COH-BR1 cells, which are deficient in ability to detoxify these peroxides. The prospect of disseminating oxidative cell injury via translocation of ChOOHs and other lipid hydroperoxides is readily apparent from these findings.

Hyperresistance to cholesterol hydroperoxide-induced peroxidative injury and apoptotic death in a tumor cell line that overexpresses glutathione peroxidase isotype-4.

The selenoenzyme phospholipid hydroperoxide glutathione peroxidase (PHGPX; GPX4) plays a key role in eukaryotic defense against potentially lethal peroxidative injury and also regulation of physiological peroxide tone. In this work we focused on the cytoprotective antiperoxidant effects of GPX4, using a breast tumor epithelial cell line that over-expresses the enzyme. Wild-type COH-BR1 cells, which exhibit little (if any) GPX4 activity, were transfected with a construct encoding the mitochondrion-targeted long (L) form of the enzyme. Several transfectant clones were selected which expressed relatively large amounts of GPX4, as determined by both Northern and Western analysis. Enzyme activity ranged from 15-fold to 190-fold greater than that of wild-type or null-transfected cells. The functional ramifications of GPX4 overexpression were tested by challenging cells with photochemically generated cholesterol hydroperoxides (ChOOHs) in liposomal form. Compared with vector controls, overexpressing clones were found to be substantially more resistant to ChOOH-induced killing, as determined by annexin-V (early apoptotic) and thiazolyl blue (mitochondrial dehydrogenase) reactivity. Concomitantly, the clones exhibited a striking hyper-resistance to free radical-mediated lipid peroxidation, as assessed by labeling cell membranes with [(14)C]cholesterol and measuring a family of radiolabeled oxidation products (ChOX). L-form GPX4's antiperoxidant and cytoprotective effects could reflect its ability to detoxify ChOOHs as they enter cells and/or cell-derived lipid hydroperoxides arising from ChOOH one-electron turnover.

Photosensitized oxidation of membrane lipids: reaction pathways, cytotoxic effects, and cytoprotective mechanisms.

Unsaturated lipids in cell membranes, including phospholipids and cholesterol, are well-known targets of oxidative modification, which can be induced by a variety of stresses, including ultraviolet A (UVA)- and visible light-induced photodynamic stress. Photodynamic lipid peroxidation has been associated with pathological conditions such as skin phototoxicity and carcinogenesis, as well as therapeutic treatments such as antitumor photodynamic therapy (PDT). Lipid hydroperoxides (LOOHs), including cholesterol hydroperoxides (ChOOHs), are important non-radical intermediates of the peroxidative process which can (i) serve as in situ reporters of type I vs. type II chemistry; (ii) undergo one-electron or two-electron reductive turnover which determines whether peroxidative injury is respectively intensified or suppressed; and (iii) mediate signaling cascades which either fortify antioxidant defenses of cells or evoke apoptotic death if oxidative pressure is too great. The purpose of this article is to review current understanding of photodynamic (UVA- or visible light-induced) lipid peroxidation with a special focus on LOOH generation and reactivity. Future goals in this area, many of which depend on continued development of state-of-the-art analytical techniques, will also be discussed.

Inhibition of free radical-mediated cholesterol peroxidation by diazeniumdiolate-derived nitric oxide: effect of release rate on mechanism of action in a membrane system.

Nitric oxide ((*)NO) flux in relation to antiperoxidant action has been studied, using large unilamellar liposomes (LUVs) as target membranes. LUVs consisting of an oxidizable phosphatidylcholine (PC), [(14)C]cholesterol (Ch) as a reaction probe, and 5alpha-hydroperoxycholesterol (5alpha-OOH) as a nonregenerable primer underwent chain peroxidation when exposed to a lipophilic iron chelate [Fe(HQ)(3), 1 microM] and ascorbate (AH(-), 1 mM) at 37 degrees C. Reaction progress was monitored by (i) HPLC with reductive-mode electrochemical detection to assess the decay of 5alpha-OOH and the formation and/or turnover of free radical-derived 7alpha- and 7beta-hydroperoxycholesterol (7alphabeta-OOH) and (ii) HPTLC with phosphorimaging to track all major (14)C-labeled oxidation products (ChOX), including 7alphabeta-OOH, 7alpha-OH, 7beta-OH, and 5,6-epoxide. Three diazeniumdiolate (*)NO donors with different half-lives were tested for their ability to interfere with peroxidation: MANO ( approximately 1 min), PANO (15 min), and SPNO (38 min). At starting concentrations of < or =200 microM, none of the donors slowed 5alpha-OOH exponential decay, ruling out any interference with redox-active iron. However, SPNO and to a greater extent PANO (but not the decomposed donors) decreased both the initial rate and steady state of 7alphabeta-OOH accumulation in a strong dose-dependent fashion. In contrast, MANO completely inhibited 7alphabeta-OOH formation over the first 5 min of reaction, but thereafter, the peroxide accumulated rapidly, albeit more slowly than without MANO and independently of the MANO dose. The latter response diminished with increasing Fe(HQ)(3) concentration, coincident with more rapid 5alpha-OOH loss. The same general trends with MANO, PANO, and SPNO were observed when the entire population of [(14)C]ChOX species was monitored. These effects are attributed to interception of Ch- and PC-derived free radicals by (*)NO, high-flux (*)NO from MANO acting mainly on 5alpha-OOH-derived radicals (chain prevention), low-flux (*)NO from SPNO mainly on downstream radicals (chain termination), and intermediate-flux (*)NO from PANO by a combination of these mechanisms. Thus, delivery rate can be an important determinant of how (*)NO inhibits peroxide-induced lipid peroxidation.

Dissemination of peroxidative stress via intermembrane transfer of lipid hydroperoxides: model studies with cholesterol hydroperoxides.

Lipid hydroperoxides (LOOHs) can be generated in cells when cholesterol (Ch) and other unsaturated lipids in cell membranes are degraded under conditions of oxidative stress. If LOOHs escape reductive detoxification by glutathione-dependent selenoperoxidases, they may undergo iron-catalyzed one-electron reduction to free radical species, thus triggering peroxidative chain reactions which exacerbate oxidative membrane damage. LOOHs are more polar than parent lipids and much longer-lived than free radical precursors or products. Accordingly, intermembrane transfer of LOOHs (analogous to that of unoxidized precursors) might be possible, and this could jeopardize acceptor membranes. We have investigated this possibility, using photoperoxidized [(14)C]Ch-labeled erythrocyte ghosts as cholesterol hydroperoxide (ChOOH) donors and unilamellar liposomes [e.g., dimyristoyl-phosphatidylcholine/Ch, 9:1 mol/mol] as acceptors. ChOOH material consisted mainly of 5alpha-hydroperoxide, a singlet oxygen adduct. Time-dependent transfer of ChOOH versus Ch at 37 degrees C was determined, using high-performance liquid and thin-layer chromatographic methods to analyze liposomal extracts for these species. A typical experiment in which the starting ChOOH/Ch mol ratio in ghosts was approximately 0.05 showed that the initial transfer rate of ChOOH was approximately 16 times greater than that of parent Ch. Using [(14)C]Ch as a reporter in liposome acceptors, we found that transfer-acquired ChOOHs, when exposed to a lipophilic iron chelate and ascorbate, could trigger strong peroxidative chain reactions, as detected by accumulation of [(14)C]Ch oxidation products. These findings support the hypothesis that intermembrane transfer of ChOOHs can contribute to their prooxidant membrane damaging and cytotoxic potential.

Cholesterol as a singlet oxygen detector in biological systems.

In cells under oxidative attack, membrane Ch, through the formation of its signature hydroperoxide and diol products, can serve as a unique detector in situ, allowing discrimination between 1O2 and free radical intermediacy. Of the two techniques described for analyzing Ch oxidation products, TLC with color development suffices for preliminary, mainly qualitative product screening, whereas a high-performance approach such as HPLC-EC(Hg) is advised when maximum resolution and sensitivity of quantitation are necessary. By using these strategies, one can monitor the formation of 1O2, for example, in a biologically relevant milieu (membrane), thus avoiding the difficulties associated with external detection, e.g., by physical means. These approaches would be valuable for assessing reaction mechanisms for various oxidative agents of biomedical importance, including environmental phototoxins and the rapidly emerging family of phototherapeutic drugs. Although photodynamic stress has been emphasized, the methods described should have broad applicability in the elucidation of oxidative mechanisms.

Nitric oxide inhibition of free radical-mediated cholesterol peroxidation in liposomal membranes.

The ability of nitric oxide ((*)NO) to inhibit propagative lipid peroxidation was investigated using unilamellar liposomes (LUVs) constituted with egg phosphatidylcholine (PC) or 1-palmitoyl-2-oleoylphosphatidylcholine (POPC), [(14)C]cholesterol (Ch), and a nonregenerable singlet oxygen-derived primer, 5alpha-hydroperoxycholesterol (5alpha-OOH). Exposing LUVs to ascorbate and a lipophilic iron chelate at 37 degrees C resulted in an exponential decay of 5alpha-OOH and accumulation of free radical-derived 7alpha- and 7beta-hydroperoxycholesterol (7alphabeta-OOH), as detected by high-performance liquid chromatography with electrochemical detection. Thiobarbituric acid-reactive species (TBARS) were generated concurrently in egg PC-containing LUVs. Including the (*)NO donor spermine NONOate (SPNO, 5-50 microM) or S-nitroso-N-acetyl-D,L-penicillamine (SNAP, 50-100 microM) in the reaction mixture had no effect on 5alpha-OOH decay (suggesting that iron was not redox-inhibited) but slowed TBARS and 7alphabeta-OOH accumulation in a strongly dose-dependent fashion. Decomposed SPNO or SNAP had no such effects, implying that (*)NO was the responsible agent. Accumulation of several [(14)C]Ch oxidation products, detected by high-performance thin-layer chromatography with phosphorimaging, was similarly diminished by active SPNO or SNAP. Concomitantly, a new band referred to as RCh.4 appeared, the radioactivity of which increased as a function of incubation time and (*)NO donor concentration. RCh.4 material was also generated via direct iron/ascorbate reduction of 7alpha-OOH in the presence of (*)NO, consistent with 7alpha-nitrite (7alpha-ONO) identity. However, various other lines of evidence suggest that RCh.4 is not 7alpha-ONO, but rather 5alpha-hydroxycholesterol (5alpha-OH) generated by reduction of 5alpha-ONO arising from 7alpha-ONO rearrangement. 5alpha-OH was only detected when (*)NO was present in the reaction system, thus providing indirect evidence for the existence of nitrosated Ch intermediates arising from (*)NO chain-breaking activity.

Singlet oxygen adducts of cholesterol: photogeneration and reductive turnover in membrane systems.

Identification of signature products provides a powerful means for establishing whether singlet molecular oxygen (1O2) is a reactive intermediate in a photodynamic process. This approach is particularly attractive for biological systems in which direct physical measurement is difficult because of the short lifetime of 1O2. Among the many possible reporter molecules in a target system, cholesterol (Ch) has the advantage of affording a limited number of readily distinguishable oxidation products, among which are the hydroperoxides 3 beta-hydroxy-5 alpha-cholest-6-ene-5-hydroperoxide (5 alpha-OOH), 3 beta-hydroxycholest-4-ene-6 alpha-hydroperoxide (6 alpha-OOH) and 3 beta-hydroxycholest-4-ene-6 beta-hydroperoxide (6 beta-OOH) that derive specifically from 1O2 addition. The purpose of this study was to compare these species in terms of (1) rates of accumulation in photodynamically treated liposomal membranes; (2) susceptibility to iron-mediated 1 e- reduction that triggers chain peroxidative damage; (3) susceptibility to selenoperoxidase (phospholipid hydroperoxide glutathione peroxidase [PHGPX])-mediated 2 e- reduction that protects against such damage and (4) relative toxicity to mammalian cells. Our results indicate that 5 alpha-OOH is photogenerated at a much greater initial rate than 6 alpha-OOH or 6 beta-OOH. Although liposomal 5 alpha-OOH, 6 alpha-OOH, and 6 beta-OOH exhibit similar first-order decay kinetics during iron/ascorbate treatment, the former decays much more slowly during GSH/PHGPX treatment, and is more toxic to L1210 cells. These and related findings suggest that 5 alpha-OOH is potentially the most damaging ChOOH to arise in photodynamically treated cells.

Low-density lipoprotein susceptibility to oxidation and cytotoxicity to endothelium in sickle cell anemia.

Patients with sickle-cell anemia exhibit pro-oxidative metabolic perturbations. We hypothesize that because of chronic oxidative stress, plasma low-density lipoprotein (LDL) from patients with sickle-cell anemia is more susceptible to oxidation. To test this hypothesis, LDL susceptibility to copper-mediated oxidation was measured in 24 patients with sickle-cell anemia and 48 control subjects. Sickle-cell LDL was more susceptible to oxidation than control LDL, measured by a 22% shorter mean lag time between LDL exposure to CuSO4 and conjugated diene formation (97 vs 124 minutes; P = .023). LDL vitamin E, iron, heme, and cholesterol ester hydroperoxide (CEOOH) levels were also measured. LDL vitamin E levels were significantly lower in patients with sickle-cell anemia compared with control subjects (1.8 vs 2.9 mol/mol LDL; P = .025), but there was no correlation with lag time. Pro-oxidant heme and iron levels were the same in sickle-cell and control LDL. LDL CEOOHs were not significantly different in sickle and control LDL (3.1 vs 1.2 mmol/mol of LDL unesterified cholesterol, P = .15), but LDL CEOOH levels were inversely correlated with lag times in patients with sickle-cell anemia (r2 = 0.38; P = .018). The cytotoxicity of partially oxidized LDL to porcine aortic endothelial cells was inversely correlated with lag times (r2 = 0.48; P = .001). These preliminary data suggest that increased LDL susceptibility to oxidation could be a marker of oxidant stress and vasculopathy in patients with sickle-cell anemia.

Radiolabeled cholesterol as a reporter for assessing one-electron turnover of lipid hydroperoxides.

A novel approach for assessing the peroxidative chain initiation potency of lipid hydroperoxides has been developed, which involves use of 14C-labeled cholesterol (Ch) as a "reporter" lipid. Unilamellar liposomes containing 1-palmitoyl-2-oleoyl-phosphatidylcholine, [14C]Ch, and 3beta-hydroxy-5alpha-cholest-6-ene-5-hydroperoxide (5alpha-OOH) or 3beta-hydroxycholest-5-ene-7alpha-hydroperoxide (7alpha-OOH) [100:75:5, mol/mol] were used as a test system. Liposomes incubated in the presence of ascorbate and a lipophilic iron complex were analyzed for radiolabeled oxidation products/intermediates (ChOX) by means of silica gel high-performance thin layer chromatography with phosphorimaging detection. The following ChOX were detected and quantified: 7alpha-OOH, 7beta-OOH, 7alpha-OH, 7beta-OH, and 5, 6-epoxide. Total ChOX yield increased in essentially the same time- and [iron]-dependent fashion for initiating 5alpha-OOH and 7alpha-OOH. The initial rate of [14C]7alphabeta-OH formation was greatly diminished when GSH and ebselen (a selenoperoxidase mimetic) were present, consistent with the attenuation of one-electron peroxide turnover. [14C]Ch-labeled L1210 cells also accumulated ChOX when incubated with 5alpha-OOH-containing liposomes. The rate of accumulation was substantially greater for Se-deficient than Se-sufficient cells, indicating that peroxide-induced chain reactions were modulated by selenoperoxidase action. These results illustrate the advantages of the new approach for highly sensitive in situ monitoring of cellular peroxidative damage.

Protoporphyrin IX-sensitized photoinactivation of 5-aminolevulinate-treated leukemia cells: effects of exogenous iron.

Photodynamic therapy with 5-aminolevulinic acid (ALA) is based on metabolism of ALA to a photosensitizing agent, protoporphyrin IX (PpIX), in tumor cells. Photosensitivity of target cells may be influenced by mitochondrial iron levels because ferrochelatase-catalyzed insertion of Fe2+ into PpIX converts it to heme, a nonsensitizer. To investigate this prospect, we exposed L1210 cells (approximately 10(6)/mL in 1% serum-containing medium) to a lipophilic iron chelate, ferric-8-hydroxyquinoline (Fe[HQ]2, 0.5 microM), prior to treating with ALA (0.2 mM, 4 h) and irradiating with broadband visible light. When Fe(HQ)2 was added to cells immediately or 1 h before ALA, the initial rate of photokilling, as measured by thiazolyl blue (mitochondrial dehydrogenase) assay, was markedly less than that of non-iron controls. The HPLC analysis of cell extracts indicated that ALA-induced PpIX was at least 50% lower after this Fe(HQ)2 treatment, presumably explaining the drop in photolethality. By contrast, cells treated with ALA and light 20 h after being exposed to Fe(HQ)2 contained the same amount of PpIX as non-iron controls and were photoinactivated at nearly the same rate. The 20 h delayed cells contained approximately 12 times more immunodetectable ferritin heavy subunit than controls or 1 h counterparts, which could account for the disappearance of iron's antisensitization effects in the former. Consistent with this idea, the short-term effects of Fe(HQ)2 on ALA-induced sensitization were found to be blunted significantly in ferritin-enriched cells. The Fe(HQ)2 produced strikingly different results when cells were sensitized with exogenous PpIX, stimulating photokilling after short-term contact but inhibiting it after long-term contact while having no significant effect on the level of cell-associated PpIX in either case. Thus, iron can have diverse effects on PpIX-mediated photokilling, depending on contact time with cells and whether the porphyrin is metabolically derived or applied as such.

Lipid hydroperoxide analysis by high-performance liquid chromatography with mercury cathode electrochemical detection.

In addition to the applications described, HPLC-EC(Hg) can be used for determining LOOHs in lipoproteins and for monitoring LOOH detoxification in cells. As it continues to be developed and refined, this approach should prove to be valuable not only for ultrasensitive determination of lipid-derived peroxides, but protein- and nucleic acid-derived peroxided as well.

Delayed hyperresistance of endothelial cells to photodynamic inactivation after contact with hemin.

Hemin (ferriprotoporphyrin IX), the oxidized prosthetic group of hemoglobin, is a potential source of prooxidant iron in heavily vascularized tumors. We have evaluated hemin's effects on photodynamic inactivation of bovine artery endothelial cells, using a partially purified oligomeric fraction of hematoporphyrin derivative (HPD-A) as the sensitizing agent. Confluent cells in 5% serum/RPMI medium showed a progressive loss of thiazolyl blue (MTT)-detectable viability when irradiated with broadband visible light in the presence of HPD-A. Cells pretreated with desferrioxamine (DFO) were substantially less sensitive to photokilling, implying that non-heme iron plays a role in cytotoxic activity. Hemin (10-20 microM) had remarkably different effects on photokilling, depending on the time interval between adding it to cells and exposing them to photodynamic action. For example, cells were more sensitive when photostressed immediately after 1 h hemin treatment and washing but much more resistant when photostressed 23 h later. Similar responses were observed when cells were challenged with glucose oxidase. Immunoblot analysis following hemin treatment revealed a progressive induction of the heavy (H) subunit of ferritin that paralleled the development of hyperresistance. After incubation with saturating levels of the synthetic iron donor [55Fe]ferric-8-hydroxyquinoline, hemin-stimulated cells contained about four times more immunoprecipitable ferritin 55Fe than controls. This is consistent with the notion that sequestration of toxic iron as a result of induction of H-chain-enriched ferritin is a key factor in hyperresistance. Inflammatory injury in tumor vasculatures could expose endothelial and neoplastic cells to chronic hemoglobin-derived iron. Consequent upregulation of ferritin could impact negatively on the efficacy of photodynamic therapy and other oxidant-based cancer therapies.

Lipid hydroperoxide generation, turnover, and effector action in biological systems.

Lipid peroxidation is a well known example of oxidative damage in cell membranes, lipoproteins, and other lipid-containing structures. Peroxidative modification of unsaturated phospholipids, glycolipids, and cholesterol can occur in reactions triggered by i) free radical species such as oxyl radicals, peroxyl radicals, and hydroxyl radicals derived from iron-mediated reduction of hydrogen peroxide or ii) non-radical species such as singlet oxygen, ozone, and peroxynitrite generated by the reaction of superoxide with nitric oxide. Lipid hydroperoxides (LOOHs) are prominent non-radical intermediates of lipid peroxidation whose identification can often provide valuable mechanistic information, e.g., whether a primary reaction is mediated by singlet oxygen or oxyradicals. Certain cholesterol-derived hydroperoxides (ChOOHs) have been used very effectively in this regard, both in model systems and cells. Being more polar than parent lipids, LOOHs perturb membrane structure/function and can be deleterious to cells on this basis alone. However, LOOHs can also participate in redox reactions, the nature and magnitude of which often determines whether peroxidative injury is exacerbated or prevented. Exacerbation may reflect iron-catalyzed one-electron reduction of LOOHs, resulting in free radical-mediated chain peroxidation, whereas prevention may reflect selenoperoxidase-catalyzed two-electron reduction of LOOHs to relatively non-toxic alcohols. LOOH partitioning between these two pathways in an oxidatively stressed cell is still poorly understood, but recent cell studies involving various ChOOHs have begun to shed light on this important question. An aspect of related interest that is under intensive investigation is lipid peroxidation/LOOH-mediated stress signaling, which may evoke a variety of cellular responses, ranging from induction of antioxidant enzymes to apoptotic death. Ongoing exploration of these processes will have important bearing on our understanding of disease states associated with peroxidative stress.

Hemin-enhanced resistance of human leukemia cells to oxidative killing: antisense determination of ferritin involvement.

Human HL-60 cells exhibited a strong hyperresistance to the lethal effects of photodynamic activity (singlet oxygen) or glucose oxidase activity (hydrogen peroxide) 16-20 h after being exposed to hemin (ferriprotoporphyrin IX). Hyperresistance was accompanied by the overproduction of immunodetectable ferritin, predominantly the heavy (H) subunit, which exhibits ferroxidase activity. Cells that had been enriched in apoferritin via pinocytotic uptake showed similar hyperresistance to both types of oxidative challenge. On the other hand, preincubating cells with hemin in the presence of a phosphorothioate-linked antisense oligodeoxynucleotide against H-ferritin mRNA resulted in a strong diminution in both hyperresistance and H-ferritin induction. No effects were seen when a scrambled order oligodeoxynucleotide of the same base composition was used, confirming that the antisense oligomer had specifically inhibited H-ferritin translation. These results indicate that induced ferritin played a crucial role in the observed cytological responses. Enhanced oxidant resistance is attributed to the ability of this ferritin to rapidly sequester and incapacitate redox-active iron.

Elevated ferritin production, iron containment, and oxidant resistance in hemin-treated leukemia cells.

Hemin (ferriprotoporphyrin IX), the oxidized prosthetic group of hemoglobin, is a source of potentially cytotoxic iron, but in chronic low doses can induce cytoprotection against iron-stimulated oxidative stress. The latter property of hemin has been examined, using murine L1210 cells and three different oxidant generating systems: (i) glucose/glucose oxidase, (ii) near-ultraviolet irradiation, and (iii) dye-mediated photodynamic action. Cells treated with the lipophilic iron donor ferric-8-hydroxyquinoline, Fe(HQ)2 (1 microM, 30 min) were found to be more sensitive to oxidative killing than nontreated controls. However, cells challenged after long-term (20-24 h) exposure to hemin (10 microM) were substantially more resistant than controls and were sensitized far less by Fe(HQ)2. Immunoblot analyses of 24-h hemin-treated cells indicated that the ferritin heavy (H) subunit was elevated 12- to 15-fold, whereas the light (L) subunit was essentially unchanged. Experiments carried out with 55Fe(HQ)2 showed that iron uptake capacity of cells was greatly enhanced after hemin treatment. More specifically, hemin-stimulated cells were found to contain approximately 9 times more immunoprecipitable ferritin iron after incubation with saturating levels (4-5 microM) of 55Fe(HQ)2 and approximately 3 times more iron per ferritin molecule compared with nonstimulated controls. The nonferritin iron content of the latter was estimated to be approximately 40 times greater than that of the former following low-level (0.5 microM) 55Fe(HQ)2 treatment. These results are consistent with the idea that induced ferritin, enriched in H-chain, sequesters redox active iron rapidly and copiously, thereby enhancing cellular resistance to oxidants.

Lipid peroxidation in photodynamically stressed mammalian cells: use of cholesterol hydroperoxides as mechanistic reporters.

Photodynamic action of merocyanine 540, an antileukemic sensitizing dye, on murine L1210 cells results in the formation of lipid hydroperoxides and loss of cell viability. High-performance liquid chromatography with mercury cathode electrochemical detection was used for determining lipid oxidation products, including the following cholesterol-derived hydroperoxides: 5 alpha-OOH, 6 alpha-OOH, 6 beta-OOH, and unresolved 7 alpha, 7 beta-OOH. Among these species, 5 alpha-, 6 alpha-, and 6 beta-OOH (singlet oxygen adducts) were predominant in the early stages of photooxidation, whereas 7 alpha- and 7 beta-OOH (products of free radical reactions) became so after prolonged irradiation or during dark incubation after exposure to a light dose. These mechanistic changes were studied in a unique way by monitoring shifts in the peroxide ratio, i.e., 7-OOH/5 alpha-OOH, or 7-OOH/6-OOH. When cells (10(7)/ml) were exposed to a visible light fluence of 0.6 J/cm2 in the presence of 10 microM merocyanine 540, 7-OOH/5 alpha-OOH increased by approximately 100% after 2 h of dark incubation at 37 degrees C. The increase was much larger (approximately 250%) when cells were photooxidized after treatment with 1 microM ferric-8-hydroxyquinoline, a lipophilic iron donor, whereas no increase was observed when cells were pretreated with 100 microM desferrioxamine, an avid iron chelator/redox inhibitor. Correspondingly, postirradiation formation of thiobarbituric acid-reactive material was markedly enhanced by ferric-8-hydroxyquinoline and suppressed by desferrioxamine, as was the extent of cell killing. When added to cells after a light dose, chain-breaking antioxidants such as butylated hydroxytoluene and alpha-tocopherol strongly protected against cell killing and slowed the increase in 7-OOH/5 alpha-OOH ratio. It is apparent from these results that (1) the 7-OOH/5 alpha-OOH or 7-OOH/6-OOH ratio can be used as a highly sensitive index of singlet oxygen vs. free radical dominance in photodynamically stressed cells; and (2) that postirradiation chain peroxidation plays an important role in photodynamically initiated cell killing.

Hyperresistance of leukemia cells to photodynamic inactivation after long-term exposure to hemin.

Merocyanine 540 (MC540)-mediated photodynamic action is a novel approach for purging tumor cells from autologous remission bone marrow explants. The purpose of this study was to evaluate the effects of hemin (ferriprotoporphyrin IX), a potential source of pro-oxidant iron in bone marrow, on in vitro photodynamic inactivation of leukemia cells. Murine L1210 cells exhibited a progressive loss of clonogenicity when irradiated with broad-band visible light in the presence of MC540. Hemin had strikingly different effects on photokilling, depending on its contact time with cells, eliciting a sizable decrease in resistance after short-term (30-min) contact but a marked increase in resistance after long-term (24-h) contact. Similar trends were observed when cells were challenged with glucose/glucose oxidase, indicating that the responses apply to more than one type of oxidative stress. Immunoblot analyses revealed that the levels of inducible heme oxygenase (HO-1) and ferritin heavy (H) chain were substantially elevated 24 h after hemin addition. HO-1 increased relatively rapidly and maximized within 4 h after adding hemin, whereas H-ferritin increased more slowly in parallel with the development of hyperresistance, maximizing after 24-36 h. Desferrioxamine, an avid iron chelator, had no effect on HO-1 induction but inhibited both ferritin induction and the increase in cell resistance, suggesting that HO-mediated release of iron from hemin was necessary for triggering these responses. Spleen apoferritin was taken up by L1210 cells and strongly inhibited photokilling, further implicating ferritin involvement in hyperresistance. Photokilling was accompanied by free radical-mediated lipid peroxidation (thiobarbituric acid reactivity), which could be suppressed substantially by 24-h hemin preincubation. A plausible explanation for the long-term effects of hemin is that excess H-ferritin generated as a result of iron-regulatory protein deactivation sequesters toxic iron, which might otherwise catalyze damaging lipid peroxidation. Chronic oxidative release of hemin from bone marrow erythroid cells could compromise the efficacy of photopurging by making tumor cells more tolerant to photooxidative insult.

Enzymatic reducibility in relation to cytotoxicity for various cholesterol hydroperoxides.

Phospholipid hydroperoxide glutathione peroxidase (PHGPX) is a selenoenzyme that can catalyze the direct reduction of various membrane lipid hydroperoxides and by so doing could play a vital role in cytoprotection against peroxidative damage. The activity of purified testicular PHGPX on several photochemically-generated cholesterol hydroperoxide (ChOOH) species was investigated, using high-performance liquid chromatography with electrochemical detection for peroxide analysis and thinlayer chromatography with 14C-radiodetection for diol product analysis. The following ChOOH isomers were monitored: 5 alpha-OOH, 6 alpha-OOH, 6 beta-OOH (singlet oxygen adducts), and unresolved 7 alpha,7 beta-OOH (derived from 5 alpha-OOH rearrangement). Apparent first-order rate constants for GSH/PHGPX-induced peroxide loss (or diol accumulation) in Triton X-100 micelles, unilamellar liposomes, or erythrocyte ghost membranes increased in the following order: 5 alpha-OOH < 6 alpha-OOH approximately equal to 7 alpha,7 beta-OOH < 6beta-OOH. A similar trend was observed when the peroxides were incubated with Triton Iysates of Se-replete L1210 or K562 cells, implicating PHGPX in these reactions. Consistent with this, there was little or no ChOOH reduction if GSH was omitted or if lysates from Se-deprived cells were used. Liposomal 5 alpha-OOH was found to be much more cytotoxic than equimolar liposomal 6 beta-OOH, producing a 50% loss of L1210 clonogenicity at approximately 1/5 the concentration of the latter. Faster uptake of 5 alpha-OOH was ruled out as the basis for greater cytotoxicity, suggesting that relatively inefficient metabolism by the GSH/PHGPX system might be the reason. As supporting evidence, it was found that cells accumulate the diol reduction product of 5 alpha-OOH more slowly than that of 6 beta-OOH during incubation with the respective peroxides. Slow detoxification coupled with rapid formation makes 5 alpha-OOH potentially the most damaging ChOOH to arise in cells exposed to singlet oxygen.

Role of hydrogen peroxide in the cytotoxic effects of UVA/B radiation on mammalian cells.

Effects of selenium (Se) deficiency on the sensitivity of murine leukemia L1210 cells to broad band UVA/B radiation (310-400 nm) have been investigated. Cells rendered glutathione peroxidase (GPX) deficient by shortterm (2-3 week) growth in 1%, serum/RPMI medium without added Se [L.Se(-) cells] were found to be much less resistant to clonally assessed UVA/B lethality than Se-supplemented controls [L.Se(+) cells]. By contrast, long-term ( > 20 week) Se-deprived [L'.Se(-)] cells whose catalase (CAT) activity was elevated > 100-fold were far more resistant to UVA/B than L.Se(+) cells. Similar trends were observed for cells irradiated in 1% serum/RPMI or Hank's medium. Whereas the CAT inhibitor 3-amino-1,2,4-triazole had no effect on L.Se(+) photosensitivity, it produced a large increase in L'.Se(-) photosensitivity. These findings are consistent with H2O2 intermediacy in photokilling and suggest that L1210 cells depend mainly on GPX for protection against this species but switch to overexpressed CAT after chronic Se deprivation. In agreement with this, steady-state H2O2 levels measured by H2O2 electrode during UVA/B exposure were higher in L.Se(-) than L.Se(+) suspensions but much lower (barely detectable) in L'.Se(-) suspensions. Cytotoxic effects of UVA/B and variations thereof resulting from Se manipulation could be mimicked by treating cells with glucose oxidase in the presence of D-glucose, providing further support for H2O2 involvement. Whether UVA/B-generated H2O2 is directly cytotoxic or gives rise to a more damaging species such as hydroxyl radical (HO) is presently unknown.