Demonstration of an extracellular matrix in multicellular tumor spheroids.

Multicellular spheroids of a human glioma cell line (U-118 MG) and a human thyroid cancer cell line ( HTh -7) were analyzed for the presence of extracellular matrix (ECM) using light microscopy, transmission electron microscopy, and indirect immunofluorescence staining for fibronectin, laminin, and collagen. These studies were supplemented by analyses of glycosaminoglycans using autoradiography or chemical methods after metabolic labeling with [35S]sulfate or [3H]glucosamine in conjunction with various extraction procedures. The results showed that both types of spheroids contained an ECM composed of fibronectin, laminin, collagen, and glycosaminoglycans. The organization of the ECM in the spheroids seemed to be similar to that of tumors in vivo. These findings help justify the use of the spheroid system as an in vitro model for the study of biological phenomena of human tumors in vivo. Furthermore, it is concluded that the formation of an ECM in vitro is not confined to normal cells but can be promoted in transformed cells using appropriate culture conditions.

Phospholipid and enzyme arrangements of rat liver rough microsomal subfractions from control and methylcholanthrene-treated animals.

Rough microsomes from rat liver of both control and methylcholanthrene-treated animals were subfractionated on a discontinuous sucrose gradient into three fractions according the their sedimentation velocity. The slowly sedimenting vesicles were enriched in electron transport enzymes, while those in the pellet showed higher phosphatase and ATPase activities. Methylcholanthrene treatment introduced typical changes in enzyme composition, mainly an increase of the cytochrome P-448. The individual phospholipids exhibited an identical distribution pattern in the three subfractions and no change occurred after induction with methylcholanthrene treatment. Nearest neighbour analysis of phosphatidylethanolamine with dinitrodifluorobenzene revealed a similar pattern in the enzymatically different subfraction, that is, no cross-linking with phosphatidylserine occurred. One-third of the phosphatidylethanolamine was in monomer and dimer form and about two-thirds was protein linked. When membrane and enzyme synthesis was induced, cross-linking to proteins were substantially decreased. The experiments indicate that the phospholipids are distributed in a homogenous fashion in the lateral plane of the rough microsomal membrane and do not support the possibility that phosphatidylethanolamine is specifically associated with cytochrome P-450.

Effects of dietary di(2-ethylhexyl)phthalate on the structure and function of rat hepatocytes.

The effect of the plasticizer di(2-ethylhexyl)phthalate on the intracellular membranes of hepatocytes was investigated. Supplementation of the diet with 2% plasticizer resulted in the appearance of a large number of peroxisomes, and the number of mitochondria was also greatly increased. No significant change in the amount or appearance of the endoplasmic reticulum was detected. The oxidation of palmitoyl-CoA in peroxisomes and the activities of carnitine-acyltransferases are increased to a great extent in both mitochondria and peroxisomes. Intact respiratory control and oxidative phosphorylation indicated that mitochondrial integrity was maintained during the induction. In microsomes, cytochrome P-450 and NADPH-cytochrome c reductase are elevated. The increased incorporation of glycerol into phospholipids indicated an increased rate of synthesis. The induction of peroxisomal and mitochondrial membranes and enzymes, but not of the membranes of the endoplasmic reticulum, by phthalate esters is an unusual and valuable induction pattern not seen with other inducers.

Relationship between the endoplasmic reticulum-Golgi membrane system and ubiquinone biosynthesis.

The involvement of the various segments of the endoplasmic reticulum (ER)-Golgi system in ubiquinone biosynthesis in rat liver was investigated using subcellular fractionation. In addition to preparing rough (R) and smooth microsomes and three different Golgi fractions, a procedure was developed to isolate a smooth vesicle fraction, designated as smooth II (SII) microsomes. The electron micrographs, chemical composition, distribution of marker enzymes, pattern of glycosidases and glycosyltransferases and participation in cholesterol transport suggest that the vesicle components of this latter fraction are intermediary between the endoplasmic reticulum and Golgi system. Both R and smooth I (SI), but not SII microsomes nor Golgi vesicles demonstrate trans-prenyltransferase activity, which synthesizes the side-chain of ubiquinone from geranyl pyrophosphate (GPP). The subsequent enzyme, which transfers solanesyl pyrophosphate (sol-PP) to 4-hydroxybenzoate, is absent from R and SI microsomes, but present in SII microsomes and exhibits high levels of activity in all of the Golgi fractions. Thus, ubiquinone is synthesized sequentially in the ER-Golgi system and thereafter translocated from this compartment to other cellular membranes.

An approach to the assessment of membrane stability of cultured cells.

A simple method for assessing the combined stability of the plasma and lysosomal membranes of cultured cells is described. Monolayers of normal, human glial cells were incubated in situ in an isotonic, buffered sucrose solution (pH 5.0) containing the acid phosphatase (AP) enzyme substrate p-nitrophenyl phosphate (PNPP). The rate of appearance, in the solution, of the reaction product p-nitrophenol (PNP) was measured spectrophotometrically, curves then plotted, and fitted by computer. "Lag time" (LT) was calculated, and an index of membrane lability constructed, termed "fragility index" (FI). Transmission electron microscopy (TEM), "vital" staining of the cells with fluorescein diacetate (FDA) and Evans Blue (EB), and use of a Gomori-type cytochemical technique, indicate that the data reflects the combined stability of lysosomal and plasma membranes. The latter playing the more critical role. Cell cultures pre-incubated with various membrane labilizing or stabilizing agents were compared. Control, 0.3 M sucrose, and normal saline treated cells demonstrated similar stability. Distilled water decreased AP latency (increased fragility), and the magnitude of this effect was time dependent. Cells fixed in glutaraldehyde (GA) retained much of their osmotic reactivity, as confirmed by distilled water treatment. Oxygen derived free radicals caused pronounced fragility, while dexamethasone, a membrane stabilizing agent, decreased membrane fragility. Triton X-100 abolished latency completely, and total AP activity was very rapidly recovered outside the cells in the surrounding incubation medium. These results suggest this technique yields a measure of membrane stability which is sensitive enough to differentiate between known stabilizers and labilizers of membranes. Hence, this may prove an easy and useful aid for the assessment of how various substances and environments modulate the lysosomal and plasma membrane stability of cultured cells.

The influence of cell co-operation, nutrients and surface area on cell division.

Two opposite views have been proposed to explain the decline of the growth potential in cell populations with a limited life span: 1 variations in the probability of cycling and in cycling times or 2 a progressive increase in the nondividing cell fraction. Human brain-derived cells were studied with respect to their proliferative potential under the influence of different growth conditions, using haptotactic palladium islands on agarose. The results emphasize the need for cell co-operation, surface area and nutrients for cell division. These parameters also influence the final cell density. The results illustrate the multiple factors that can vary the probability of initiating the division cycle and stress the uncertainty of defining the irreversible non-dividing state.

Cellular injury induced by oxidative stress is mediated through lysosomal damage.

Cultured primary hepatocytes pretreated (protected) with the iron chelator deferoxamine or the antioxidant N,N'-diphenyl-p-phenylenediamine (DPPD) were resistant to the toxicity of 5 microM naphthazarin (5,8-dihydroxy-1,4-naphthoquinone) during a 180-min exposure. Hepatocytes exposed to naphthazarin without any protection were abruptly depleted of intracellular reduced glutathione, and the level of cytosolic Ca2+ was rapidly increased. This was followed by lipid peroxidation, measured as accumulation of malondialdehyde (MDA) and 4-hydroxyalkenals (4-HNA) intra- and extracellularly; decrease in ATP levels; destabilization of lysosomes; and finally cell death. The stability of the lysosomal membranes was evaluated by determining retention of the lysosomotropic weak base acridine orange (AO). Naphthazarin exposure caused leakage of protons from the acidic compartment, as indicated by relocalization of AO to the cytosol. Protection of the cell cultures with deferoxamine or DPPD prevented destabilization of lysosomes and cell killing. It also reduced the loss of ATP but did not prevent the depletion of glutathione or the increase in Ca2+. In cells subjected to naphthazarin exposure, DPPD protection also completely inhibited lipid peroxidation, whereas deferoxamine pretreatment only slightly reduced the intracellular accumulation of MDA and 4-HNA but completely prevented cell rupture and the leakage of these lipid peroxidation products to the medium that took place in large amounts from unprotected cells exposed to naphthazarin. Deferoxamine is taken up by endocytosis and is thus transported to the acidic vacuolar apparatus, whereas the lipophilic DPPD is rapidly distributed throughout the cells. Inhibiting endocytosis during deferoxamine pretreatment, by incubating at +4 degrees C or by preexposure to a mixture of the endocytosis-inhibitors cytochalasin B and monensin, abolished the protective effect of deferoxamine. The findings suggest that naphthazarin-induced cell killing is not caused directly by either thiol oxidation or an increase in cytosolic free Ca2+, but rather is preceded by lysosomal destabilization, which may be prevented either by inhibition of cellular peroxidation in general or by prevention of iron-catalyzed oxidative reactions, and involves peroxidation of cellular membranes, energy depletion, and leakage of lysosomal content. DPPD would protect against cell killing by preventing lipid peroxidation of cellular membranes in general, whereas deferoxamine seems to allow a limited general cellular peroxidation but specifically prevents peroxidation and fragmentation of lysosomal membranes by chelating intralysosomal iron and, consequently, leakage of destructive lysosomal contents with ensuing cell rupture and death. Thus, a certain degree of cellular peroxidation does not appear to be lethal as long as lysosomal membranes are protected, placing lysosomes into a category of cellular loci minora resistentia.

One-electron transfer reactions of diquat radical to different reduction intermediates of oxygen. Formation of hydroxyl radical and electronically excited states.

The one-electron transfer activation of DQ++ by microsomal fractions comprises an aerobic phase and an anaerobic phase. The aerobic phase is characterized by O2 consumption, formation of electronically excited states with main emission below 600 nm, and H2O2 formation. The anaerobic phase is characterized by H2O2 consumption, DQ+ accumulation, HO. formation, and also electronically excited state formation with main emission beyond 600 nm. Superoxide dismutase abolishes the photoemission during the aerobic phase, whereas it has no effect on the photoemission originating during the anaerobic phase. The hydroxylation products of the aromatic compound salicylate, mainly 2,3- and 2,5-dihydroxybenzoic acids--indicative of the occurrence of HO.-, were detected by h.p.l.c. with oxidative electrochemical detection during the anaerobic phase, but not during the aerobic phase. Neither H2O2 consumption nor HO. are prevented by desferrioxamine. These experimental observations are interpreted on the grounds of two main electron-transfer reactions of DQ.+: under aerobic conditions, two one-electron transfer steps to molecular O2 and O2.- to yield H2O2. Under anaerobic conditions, one-electron transfer step to contaminating iron or any ferrioxamine formed to a ferrous complex which can support a Fenton-like reduction of H2O2 with formation of HO.. The toxicological relevance for the occurrence of such reactions is also discussed in terms of the formation of electronically excited states.

Visualization of iron in cultured macrophages: a cytochemical light and electron microscopic study using autometallography.

The objective of this study was to develop a sensitive cytochemical method for the visualization of iron, both at light microscopical (LM) and at electron microscopical (EM) levels, in glutaraldehyde-fixed cultured cells with reasonable morphological preservation. The method is based on autometallography (also called the sulfide silver method or the Timm technique). Gold, silver, and various metal sulfides have previously been shown to act as catalysts for cellular silver deposition from a physical developer (autometallography). In our modification of this cytochemistry, a high pH is used during the initial sulfidation step to guarantee adequate levels of sulfide ions to generate enough Fe(II or III) sulfide. Since this procedure may cause severe cellular distortion, we initially stabilize the cultured cells by a glutaraldehyde fixation. We have compared our new high pH, high S2- LM and EM variety of autometallography with other modifications of this technique that have previously been used for LM and EM demonstration of easily sulfidated heavy metals, such as zinc. Cultured mouse macrophages were examined for the localization of reactive metals following endocytosis of ferritin or inorganic Fe(III) iron. Ag-precipitates, presumed to indicate the presence of iron, were predominantly found within secondary lysosomes of the acidic vacuolar apparatus. The relation of the Ag-precipitates to iron was proven by the fact that iron-exposed cells showed a much reduced amount of silver precipitates after subsequent exposure to deferoxamine a potent iron chelator. Moreover, control macrophages neither exposed to iron nor to ferritin showed only a low normal lysosomal content--and a few extralysosomal sites--of reactive substances, believed to be iron.

Influence of intracellular glutathione concentration of lipofuscin accumulation in cultured neonatal rat cardiac myocytes.

The objective of this study was to investigate the relationship between intracellular glutathione (GSH) and lipofuscinogenesis in an established model system of cultured postmitotic neonatal rat cardiac myocytes exposed to moderate oxidative stress with respect to culture conditions (cells grown at 21% oxygen pressure). Intracellular glutathione was depleted by exposing cell cultures to buthionine sulfoximine (BSO), a specific inhibitor of gamma-glutamylcysteine synthetase; this caused an increase in lipofuscin-specific autofluorescence, reflecting lipofuscin accumulation. Cell cultures exposed to 100 microM BSO exhibited the following reactions, as compared to control cells: the intracellular glutathione level decreased 78%, 86%, and 89%, and lipofuscin-specific autofluorescence increased 26%, 28%, and 77% after 5, 8, and 14 days of treatment, respectively. Compared to the glutathione levels found in newly excised heart tissue, concentrations in cultured cardiac myocytes were lower during the first few days after culture establishment, probably a result of the cell preparation technique used. Due to this effect, differences between BSO-treated cells and control cells in regard to glutathione concentration and lipofuscin accumulation were more pronounced during the remainder of the 2-week culture period. Lipid peroxidation measured as thiobarbituric acid reactive substances was not increased in BSO-treated cells. These results are in agreement with the oxidative-stress theory of lipofuscinogenesis previously proposed from the authors' laboratory. The authors show that reduced GSH level leads to a simultaneous increase in accumulation of lipofuscin in cardiac myocytes, possibly by increasing the level of cytosolic hydrogen peroxide.

Delayed oxidant-induced cell death involves activation of phospholipase A2.

Short-term (1 h) exposure of cells to a low steady-state concentration of H(2)O(2) causes no immediate cell death but apoptosis occurs several hours later. This delayed cell death may arise from activation of phospholipases, in particular phospholipase A2 (PLA2), which may destabilize lysosomal and mitochondrial membranes. Indeed, the secretory PLA2 (sPLA2) inhibitor 4-bromophenacyl bromide diminishes both delayed lysosomal rupture and apoptosis. Furthermore, sPLA2 activation by mellitin, or direct micro-injection of sPLA2, causes lysosomal rupture and apoptosis. Finally, B-cell leukemia/lymphoma 2 (Bcl-2) over-expression prevents oxidant-induced activation of PLA2, delayed lysosomal destabilization and apoptosis. This supports a causal association between PLA2 activation and delayed oxidant-induced cell death and suggests that Bcl-2 may suppress apoptosis by preventing PLA2 activation.

Bcl-2 phosphorylation is required for inhibition of oxidative stress-induced lysosomal leak and ensuing apoptosis.

B-cell leukemia/lymphoma 2 (Bcl-2) blocks oxidant-induced apoptosis at least partly by stabilizing lysosomes. Here we report that phosphorylation of Bcl-2 may be required for these protective effects. J774 cells overexpressing wild-type Bcl-2 resist oxidant-induced lysosomal leak as well as apoptosis, and this protection is amplified by pretreatment with phorbol 12-myristate 13-acetate (which promotes protein kinase C (PKC)-dependent phosphorylation of Bcl-2). In contrast, cells overexpressing the Bcl-2 mutant S70A (which cannot be phosphorylated) are not protected in either circumstance. Transfection with Bcl-2(S70E), a constitutively active Bcl-2 mutant which does not require phosphorylation, is protective independent of PKC activation. In contrast, C(2)-ceramide, a putative protein phosphatase 2A activator, abolishes the protective effects of wild-type Bcl-2 overexpression but does not diminish protection afforded by Bcl-2(S70E). Additional results suggest that, perhaps as a consequence of lysosomal stabilization, Bcl-2 may prevent activation of phospholipase A2, an event potentially important in the ultimate initiation of apoptosis.

Protection against oxidant-mediated lysosomal rupture: a new anti-apoptotic activity of Bcl-2?

Bcl-2 antagonizes apoptosis through mechanisms which are not completely understood. We have proposed that apoptosis is initiated by minor lysosomal destabilization followed some time later by secondary massive lysosomal rupture. In J774 cells over-expressing Bcl-2, early oxidant-induced lysosomal destabilization is unaffected but secondary lysosomal rupture and apoptosis are suppressed, despite the fact that wild-type and Bcl-2 over-expressing cells degrade hydrogen peroxide at similar rates. It may be that Bcl-2 directly blocks the effects of released lysosomal enzymes and/or prevents downstream activation of unknown cytosolic pro-enzymes by released lysosomal hydrolases, suggesting a new and heretofore unknown activity of Bcl-2.

Relationship between mitochondrial superoxide and hydrogen peroxide production and longevity of mammalian species.

The objective of this study was to examine the possible involvement of oxygen free radicals in the aging process. Rates of mitochondrial O2.- and H2O2 production and oxygen consumption in the kidney and the heart were compared among seven different mammalian species namely, mouse, hamster, rat, guinea pig, rabbit, pig, and cow, whose maximum life span potential (MLSP) varies from 3.5 to 30 years. The rates of mitochondrial O2.- and H2O2 generation were inversely correlated to MLSP, and directly related to specific metabolic rate and state 4 mitochondrial respiration. Results of this study indicate that under identical conditions, mitochondria from shorter-lived species produce relatively higher amounts of reactive oxygen species than those from the longer-lived species, and, thus, support the free radical hypothesis of aging.

Effect of ethanol on lipofuscin accumulation in cultured rat cardiac myocytes.

The objective of this study was to determine the effect of ethanol on in vitro life span, rate of contraction and lipofuscin content of neonatal rat cardiac myocytes. Lipofuscin was quantified by microspectrofluorometry. The effects of 0, 3.1, 6.5, and 12.5 mM ethanol on myocytes, kept under an ambient oxygen concentration of 20% and 40%, were studied. Exposure to low concentrations of ethanol resulted in a decrease in the amount of lipofuscin whereas exposure to high concentration of ethanol caused an increase in the level of lipofuscin. The length of cell survival in controls and 3.1 mM ethanol exposed myocytes was similar under 20% oxygen, but was longer in the latter group under 40% oxygen, as compared to controls. The total number of contractions in 3.1 mM ethanol-exposed myocytes were, respectively, 4% and 8% higher under 20% and 40% oxygen atmosphere than in control cells.

Fenton reagents may not initiate lipid peroxidation in an emulsified linoleic acid model system.

This study includes two parts. First, the Fe2+ autooxidation and chelation processes in the presence of the chelators ethylenediaminetetraacetic acid (EDTA) and diethylenetriamine pentaacetic acid (DTPA) were studied by measuring UV light absorbance alterations. Competition for Fe3+ between chelators and water or phosphate buffer (PB) ions was confirmed. The addition of EDTA or DTPA to Fe3+ in water or PB only slowly turned the water/PB-Fe3+ complexes to EDTA-Fe3+ or DTPA-Fe3+ complexes. In the second part of this study, the initiation mechanisms of Tween 20 emulsified linoleic acid peroxidation under stimulation by chelator-Fe-O2 complexes were studied by measuring changes in UV light absorbance following diene conjugation. Fe3+ in the presence of EDTA or DTPA did not stimulate diene conjugation. Fe2+ (0.10 mM) and EDTA (0.11 mM) stimulated diene conjugation of the linoleic acid emulsion, but only after apparent Fe2+ autooxidation. Fe2+ and DTPA, as well as premixed DTPA-Fe2+ complex, resulted in very fast diene conjugation in a wide range of concentrations. A nonlinear, mainly square root relation between Fe2+ concentration and peroxidation rate was noted. Superoxide dismutase (SOD), catalase, and mannitol did not prevent the lipid peroxidation. H2O2 substantially decreased the DTPA-Fe2+ stimulated, otherwise rapid, diene conjugation but slightly enhanced the slower one stimulated by EDTA-Fe2+. Without ambient oxygen, Fenton reagents did not result in .H abstraction-related diene conjugation. The findings suggest that .OH resulting from Fenton reagents may not be the main cause for the initiation of peroxidation in this model system. Furthermore, a study with different combinations of Fe2+ and Fe3+ did not support the Fe2+/Fe3+ (1:1) optimum ratio hypothesis. We therefore conclude that perferryl ions or chelator-Fe-O2 complexes may be responsible for the first-chain initiation of lipid peroxidation, at least in this model system.

Lipofuscin-formation in cultured retinal pigment epithelial cells is related to their melanin content.

Age-related macular degeneration (AMD), the leading cause of blindness in the developed world, is accompanied by degeneration of the retinal pigment epithelial (RPE) cells. There is an inverse correlation between the melanin content of the eye and the incidence of AMD. Lipofuscin (LF)-accumulation in RPE cells accompanies the process of aging, and may also be related to AMD. This study was designed to evaluate the effect of melanin/melanosomes on the rate of LF formation in cultured rabbit and bovine RPE cells subjected to oxidative stress (40% normobaric O(2)) and daily supplementation with photoreceptor outer segments for 4 weeks. The LF content was measured at 0, 2, and 4 weeks in RPE cells from pigmented and albino rabbits, as well as in pigment-rich and pigment-poor bovine cells. Albino rabbit and pigment-poor bovine cells accumulated significantly higher amounts of LF than pigmented rabbit cells and pigment-rich bovine RPE cells after both 2 and 4 weeks of exposure. Autometallography of melanin-containing cells, without previous exposure to ammonium sulfide, showed a positive outcome, indicating either the occurrence of pre-existing iron-sulphur clusters or an extremely high intrinsic reducing capacity. These results suggest that melanin acts as an efficient antioxidant, perhaps by interacting with transition metals.

Effect of ambient oxygen concentration on lipofuscin accumulation in cultured rat heart myocytes--a novel in vitro model of lipofuscinogenesis.

The objective of this study was to elucidate the factors involved in the accumulation of lipofuscin in post-mitotic cells. The hypothesis that oxidative stress accelerates the rate of lipofuscin accumulation was tested by examining the effects of 5%, 20%, and 40% ambient oxygen concentration on lipofuscin content in cultured rat cardiac myocytes. Lipofuscin was quantified by microspectrofluorometry at 7 and 12 days of in vitro age. Lipofuscin-emitted yellow autofluorescence increased in direct relationship to ambient oxygen concentration with age. Transmission electron microscopic examination of the cells after 3, 8, and 12 days in culture indicated a progressive time and oxygen dependent increase in the frequency and size of lipofuscin organelles. The results are interpreted to suggest that oxidative stress is one of the causal factors in the accumulation of lipofuscin.

Photo-oxidative disruption of lysosomal membranes causes apoptosis of cultured human fibroblasts.

Acridine orange (AO) is a lysosomotropic weak base, a metachromatic fluorochrome, and a photosensitizer, as well. Living cells that are exposed for a short period of time to this compound at low concentration, and under ordinary culture conditions, accumulate the drug within their acidic vacuolar compartment, giving rise to a mainly red, granular fluoresence upon excitation with blue light. When AO-loaded cells are irradiated with intense blue light, AO soon starts to leak from late endosomes and lysosomes, partially shifting the fluorescence to a green, nuclear and diffuse cytosolic, one. This AO-relocalization is a consequence of photo-oxidation of the lysosomal membranes, which initially results in disruption of their proton-gradients and later, in leakage into the cytosol of a host of hydrolytic enzymes--as was here demonstrated by immunocytochemistry--which are capable of causing cellular damage. Most fibroblasts survived minor photo-oxidation, with a period of reparative autophagocytosis. Severe photo-oxidation, which resulted in severe lysosomal damage, caused cellular necrosis; whereas moderate stress, resulting in only partial lysosomal leakiness lead to apoptosis with TUNEL-positive nuclei and shrunken cytoplasm. The findings of the present study show that photo-oxidative damage to the membranes that surround the acidic vacuolar compartment, is an event that results in release of proteolytic and DNA-fragmenting enzymes into the cytosol, which may induce either necrosis, apoptosis, or reparable sublethal damage, depending on the magnitude of lysosomal rupture. Furthermore, the results strongly suggest that proteases and endonucleases of lysosomal origin may induce apoptosis if relocalized from the acidic vacuolar compartment into the cytosol.

Exposure of cells to nonlethal concentrations of hydrogen peroxide induces degeneration-repair mechanisms involving lysosomal destabilization.

The cytotoxicity of hydrogen peroxide is, at least partly, mediated by the induction of intralysosomal iron-catalyzed oxidative reactions with damage to lysosomal membranes and leakage of destructive contents. We hypothesize that minor such leakage may be nonlethal, and the ensuing cellular degeneration repairable. Consequently, we investigated, using a model system of cultured J-774 cells, the effects of hydrogen peroxide in moderate concentrations on cellular viability, lysosomal membrane integrity, morphology, and ATP and reduced glutathione concentrations. These parameters were initially estimated directly after a 30 min exposure to a bolus dose of hydrogen peroxide in phosphate buffered saline at 37 degrees C, and then again following subsequent recovery periods of different lengths under ordinary culture conditions. All cells survived an exposure to 250 microM hydrogen peroxide for 30 min, whereas 350 and 500 microM exposure was lethal to a small fraction of cells. The oxidative stress caused early, time- and dose-dependent, partial relocalization of the lysosomotropic weak base acridine orange from the lysosomal compartment to the cytosol. This phenomenon is known to parallel leakage of damaging lysosomal contents such as hydrolytic enzymes. There were also signs of cellular damage in the form of surface blebbing and increased autophagocytosis, more marked with the higher doses of hydrogen peroxide. Also found was a rapid depletion of ATP and GSH. These alterations were all reversible, as long as cells were exposed to nonlethal amounts of hydrogen peroxide. Based on these and previous findings, we suggest that lysosomes are less stable organelles than has hitherto been assumed. Restricted lysosomal leakage might be a common event, for example, during sublethal oxidative stress, causing reversible, degenerative alterations, which are repaired by autophagocytosis.