Mitochondrial free radical production induces lipid peroxidation during myohemoglobinuria

Formation of keto and hydroxy compounds of linoleic acid in submitochondrial particles of bovine heart

To observe lipid peroxidation of additive-free submitochondrial particles, we incubated submitochondrial particles in the absence of exogenous irons and t-butyl hydroperoxide. After the incubation, the phospholipids were hydrolyzed by phopholipase A2, and the fatty acid constituents were analyzed by high-performance liquid chromatography, gas chromatography-mass spectrometry, and liquid chromatography-mass spectrometry. Contrary to a commonly accepted theory, lipid peroxidation in the submitochondrial particles did not need the addition of NADH. In the phospholipid constituent fatty acids of the oxidized submitochondrial particles, derivatives of hydroperoxides of linoleic acid such as keto, hydroxy, trihydroxy, and hydroxyepoxy compounds were generated. Lipid peroxidation in the submitochondrial particles was not inhibited by the addition of catalase, superoxide dismutase, hydroxyl radical scavengers, or ethylenediaminetetraacetic acid, but was inhibited by the addition of KCN, antimycin-A, NADH, ubiquinol, deferoxamine mesylate, ascorbic acid, and alpha-tocopherol. The cardiolipin-cytochrome c lipid peroxidation system could mimic the lipid peroxidation of the submitochondrial particles, in terms of linoleic acid products and the inhibitory patterns of radical scavengers and electron transfer chain inhibitors. Thus, lipid peroxidation in the submitochondrial particles seems to be due to phospholipid-hemoprotein lipid peroxidation systems such as the cardiolipin-cytochrome c system.

Activation of JNK in the remote myocardium after large myocardial infarction in rats

A large myocardial infarction (MI) causes a chronic hemodynamic load on the uninjured remote myocardium (RM). This may lead to oxidative stress, activation of stress-induced cell signaling and increase in myocyte apoptosis. MI was produced in 6 rats (INF) while 4 rats underwent sham operation (CON). At four weeks, there was 128% increase in right ventricular hypertrophy in the hearts from INF vs. CON. Western blot analysis showed 3.8 fold increase in JNK phosphorylation within the RM from INF vs. CON, confirmed by a 4.2 fold increase in JNK kinase activity. There was a 52% increase in TBARS within the RM from INF vs. CON, suggesting increased lipid peroxidation. Furthermore, there was a twofold increase in myocyte apoptosis within the RM in INF vs. CON. We conclude that the RM from INF is associated with activation of JNK, increased oxidative stress and enhanced myocyte apoptosis.

Mitochondrial respiratory enzymes are a major target of iron toxicity in rat heart cells

Our previous studies in iron-loaded rat heart cells showed that in vitro iron loading results in peroxidative injury, manifested in a marked decrease in rate and amplitude of heart cell contractility and rhythmicity, which is correctable by treatment with deferoxamine (DF). In the present studies we explored the role of mitochondrial damage in myocardial iron toxicity. Iron loading by 24-hour incubation with 0.36 mmol/L ferric ammonium citrate resulted in a decrease in the activity of nicotinamide adenine dinucleotide (NADH)-cytochrome c oxidoreductase (complex I+III) to 35.3%+/-11.2% of the value in untreated controls; of succinate-cytochrome c oxidoreductase (complex II+III) to 57.4%+/-3.1%; and of succinate dehydrogenase to 63.5%+/-12.6% (p < 0.001 in all cases). The decrease in activity of other mitochondrial enzymes, including NADH-ferricyanide reductase, succinate ubiquinone oxidoreductase (complex II), cytochrome c oxidase (complex IV), and ubiquinol cytochrome c oxidoreductase (complex III), was less impressive and ranged from 71.5%+/-15.8% to 91.5%+/-14.6% of controls. That the observed loss of respiratory enzyme activity was a specific effect of iron toxicity was clearly demonstrated by the complete restoration of enzyme activities by in vitro iron chelation therapy. Sequential treatment with iron and doxorubicin caused a loss of complex I+III and complex II+III activity that was greater than that seen with either agent alone but was only partially correctable by DF treatment. Alterations in cellular adenosine triphosphate measurements paralleled very closely the changes observed in respiratory complex activity. These findings demonstrate for the first time the impairment of cardiac mitochondrial respiratory enzyme activity caused by iron loading at conditions formerly shown to produce severe abnormalities in contractility and rhythmicity.

Altered sphingomyelinase and ceramide expression in the setting of ischemic and nephrotoxic acute renal failure

Diverse physical and chemical stimuli can activate sphingomyelinases (SMases), resulting in sphingomyelin (SM) hydrolysis with ceramide release. Since ceramide can profoundly impact a host of homeostatic mechanisms, the concept of a "SM (or SMase) signaling pathway" has emerged. We recently documented that ceramide levels fall abruptly during renal ischemia, and then rebound to twice normal values during early reperfusion (30 to 90 min) Therefore, the present study assessed whether these ceramide changes are paralleled, and hence potentially mediated, by comparable changes in SMase activity. Mice were subjected to 45 minutes of renal ischemia +/- 30 minutes, 90 minutes, or 24 hours of reperfusion. Renal cortices (or isolated proximal tubules) were then assayed for SMase activity (acidic, neutral forms). To characterize whether early post-ischemic ceramide increments are a relatively persistent event, ceramide was assayed following a 24-hour reperfusion period. Finally, to assess whether the observed perturbations were unique to post-ischemic injury, SMase and ceramide were quantified in the setting of glycerol-induced myohemoglobinuria and anti-glomerular basement membrane (alpha GBM) antibody-induced acute renal failure (ARF). Ischemia induced abrupt declines (approximately 50%) in both acidic and neutral SMase activities, and these persisted in an unremitting fashion throughout 24 hours of reperfusion. Nevertheless, increased ceramide expression (2x normal) resulted. Myohemoglobinuria also suppressed acidic/neutral SMases, and again, "paradoxical" ceramide increments were observed. Finally, alpha GBM nephritis increased ceramide levels, but in this instance, a correlate was increased SMase activity. These results suggest that: (1) ceramide is an acute renal "stress rectant" increasing in response to diverse renal insults; (2) this response may occur independently of the classic SM pathway, since the ceramide increments can seemingly be dissociated from increased SMase activity; and (3) given the well documented impact of ceramide and the SM(ase) pathway on apoptosis, cell proliferation, differentiation, and tissue inflammation, the present results have potentially broad ranging implications for the induction and evolution of diverse forms of ARF.

Decreased expression of mitochondrial-derived H2O2 and hydroxyl radical in cytoresistant proximal tubules

Increased production of reactive oxygen metabolites (ROM) can contribute to the initiation phase of nephrotoxic and ischemic acute renal failure (ARF). However, whether altered ROM expression also exists during the maintenance phase of ARF has not been adequately assessed. Since diverse forms of tubular injury can initiate a "cytoresistant state," this study tested whether a down-regulation of ROM expression might develop in the aftermath of acute tubular damage, potentially limiting renal susceptibility to further attack. To test this hypothesis, rats were subjected to either mild myohemoglobinuria (glycerol injection) or bilateral ureteral obstruction and 24 hours later, cytoresistant proximal tubular segments (PTS) were isolated to assess ROM expression. PTS from sham operated rats were used to establish normal values. Both sets of cytoresistant PTS manifested approximately 75% reductions in H2O2 levels, as assessed by the phenol red/horseradish peroxidase technique (P < 0.01 to 0.001). A 40% reduction in hydroxyl radical (.OH) levels was also observed (salicylate trap method), thereby substantiating decreased oxidant stress in cytoresistant PTS. Catalase, glutathione peroxidase, and free iron levels were comparable in control and cytoresistant PTS, suggesting that decreased H2O2 production (such as by mitochondria) was the cause of the decreased oxidant stress. To test this latter hypothesis, H2O2 expression by control and cytoresistant PTS was assessed in the presence of respiratory chain inhibitors. Although site 1 and site 3 inhibition markedly suppressed H2O2 production in control PTS, they had no impact on H2O2 production in cytoresistant PTS, implying that production at these sites was already maximally suppressed. Correlates of the decreased mitochondrial H2O2 production were improvements in cell energetics (increased ATP/ADP ratios with Na ionophore treatment) and approximately 40 to 90% increases in PTS/renal cortical glutathione content. We conclude that: (1) proximal tubule H2O2/.OH expression can be downregulated during the maintenance phase of ARF; (2) this seemingly reflects a decrease in mitochondrial ROM generation; and (3) the associated improvements in glutathione content and/or cellular energetics could conceivably contribute to a post-injury cytoresistant state.

Altered ceramide and sphingosine expression during the induction phase of ischemic acute renal failure

Recent evidence indicates that a "sphingomyelin signaling pathway" exists: in response to heterogeneous influences, sphingomyelin is hydrolyzed, liberating ceramide, and subsequently its sphingoid base, sphingosine. Ceramide and sphingosine can influence diverse cellular processes, including cell differentiation, proliferation, protein trafficking, and apoptosis. Each of these processes have important implications for post-ischemic acute renal failure (ARF). However, sphingosine and ceramide expression during the induction of ischemic/reperfusion injury have not been previously assessed. To this end, CD-1 mice were subjected to 45 minutes of unilateral renal ischemia +/- reperfusion, followed by cortical sphingosine, ceramide, and sphingomyelin assessments. Contralateral kidneys served as controls. Ischemia caused approximately 50% sphingosine and ceramide decrements. During reperfusion, sphingosine rebounded to normal values. Conversely, ceramide rose to, and was maintained at, supranormal levels (approximately 175% of controls). Subsequent studies performed with hypoxic or oxygenated isolated proximal tubules suggested that these changes: (1) had a multifactorial basis; (2) were partially simulated by enhanced PLA2 activity; (3) and were dissociated from alterations in net sphingomyelin content. To assess the potential pathogenic relevance of the documented ceramide increments, cultured human proximal tubule (HK-2) cells were subjected to ATP depletion/Ca2+ ionophore- or PLA2-induced attack with or without exogenous C2 ceramide loading. Ceramide worsened both forms of injury without exerting an independent lethal effect. Conversely, ceramide markedly attenuated arachidonic acid cytotoxicity. This occurred without any decrease in arachidonate uptake, suggesting a direct cytoprotective effect.

IN CONCLUSION

(1) sphingosine and ceramide fluxes are hallmarks of early ischemic/reperfusion injury; (2) these changes occur via divergent metabolic pathways; and (3) that ceramide increments can affect divergent injury pathways, and that sphingosine and ceramide have potent cell signaling effects, suggest that the currently documented sphingosine/ ceramide fluxes could have important implications for the induction phase and evolution of post-ischemic ARF.

Oxidant mechanisms in toxic acute renal failure

Over the last decade, there has been accumulating evidence for a role of reactive oxygen metabolites in the pathogenesis of a variety of renal diseases, including gentamicin, glycerol, and cyclosporine A models of toxic acute renal failure. Gentamicin has been shown in both in vitro and in vivo studies to enhance the generation of reactive oxygen metabolites. Iron is important in models of tissue injury, presumably because it is capable of catalyzing free radical formation. Gentamicin has been shown to cause release of iron from renal cortical mitochondria. Scavengers of reactive oxygen metabolites as well as iron chelators provide protection in gentamicin-induced nephrotoxicity. In glycerol-induced acute renal failure, an animal model of rhabdomyolysis, there is enhanced generation of hydrogen peroxide, and scavengers of reactive oxygen metabolites and iron chelators provide protection. Although the dogma is that the myoglobin is the source of iron, the results of recent studies suggest that cytochrome P-450 may be an important source of iron in this model. In addition, there are marked alterations in antioxidant defenses, such as glutathione, as well as changes in heme oxygenase. Cyclosporine A has been shown to enhance the generation of hydrogen peroxide in vitro and lipid peroxidation in vitro and in vivo. Antioxidants have been shown to be protective in cyclosporine A nephrotoxicity. This collective body of evidence suggests an important role for reactive oxygen metabolites in toxic acute renal failure and may provide therapeutic opportunities of preventing or treating acute renal failure in humans.

Myoglobin toxicity in proximal human kidney cells: roles of Fe, Ca2+, H2O2, and terminal mitochondrial electron transport

The purpose of this study was to gain direct insights into mechanisms by which myoglobin induces proximal tubular cell death. To avoid confounding systemic and hemodynamic influences, an in vitro model of myoglobin cytotoxicity was employed. Human proximal tubular (HK-2) cells were incubated with 10 mg/ml myoglobin, and after 24 hours the lethal cell injury was assessed (vital dye uptake; LDH release). The roles played by heme oxygenase (HO), cytochrome p450, free iron, intracellular Ca2+, nitric oxide, H2O2, hydroxyl radical (-OH), and mitochondrial electron transport were assessed. HO inhibition (Sn protoporphyrin) conferred almost complete protection against myoglobin cytotoxicity (92% vs. 22% cell viability). This benefit was fully reproduced by iron chelation therapy (deferoxamine). Conversely, divergent cytochrome p450 inhibitors (cimetidine, aminobenzotriazole, troleandomycin) were without effect Catalase induced dose dependent cytoprotection, virtually complete, at a 5000 U/ml dose. Conversely, -OH scavengers (benzoate, DMTU, mannitol), xanthine oxidase inhibition (oxypurinol), superoxide dismutase, and manipulators of nitric oxide expression (L-NAME, L-arginine) were without effect. Intracellular (but not extracellular) calcium chelation (BAPTA-AM) caused approximately 50% reductions in myoglobin-induced cell death. The ability of Ca2+ (plus iron) to drive H2O2 production (phenol red assay) suggests one potential mechanism. Blockade of site 2 (antimycin) and site 3 (azide), but not site 1 (rotenone), mitochondrial electron transport significantly reduced myoglobin cytotoxicity. Inhibition of Na, K-ATPase driven respiration (ouabain) produced a similar protective effect. We conclude that: (1) HO-generated iron release initiates myoglobin toxicity in HK-2 cells; (2) myoglobin, rather than cytochrome p450, appears to be the more likely source of toxic iron release; (3) H2O2 generation, perhaps facilitated by intracellular Ca2+/iron, appears to play a critical role; and (4) cellular respiration/terminal mitochondrial electron transport ultimately helps mediate myoglobin's cytotoxic effect. Formation of poorly characterized toxic iron/H2O2-based reactive intermediates at this site seems likely to be involved.

Hydrogen peroxide activates glibenclamide-sensitive K+ channels in LLC-PK1 cells

Oxidant-induced damage has been implicated in the pathogenesis of several forms of cellular injury. The present study employed patch-clamp methods to determine if oxidant stress leads to activation of plasma membrane K+ channels in the renal epithelial LLC-PK1 cell line. Exposure of cells to H2O2 (0.1 to 5 mM) induced a rapid (within 5-10 min), dose-dependent membrane hyperpolarization. Perforated patch whole cell voltage-clamp studies were performed to determine the ion selectivity of the currents underlying this H2O2-induced cellular hyperpolarization. H2O2 (5 mM) produced a sixfold increase in the whole cell conductance. The reversal potential of the H2O2-induced current was consistent with a K+-selective conductance. This current was blocked almost completely by 5 mM barium and 500 microM glibenclamide but only partially by 15 mM tetraethylammonium. Exposure of LLC-PK1 cells to 5 mM H2O2 reduced cell ATP content by 70%. To evaluate more directly the role of ATP depletion in the activation of K+ channels, conventional whole cell patch-clamp studies were performed. Inclusion of ATP in the pipette solution prevented H2O2-induced activation of the K+ conductance. These findings indicate that H2O2 activates an ATP-sensitive, Ca2+-independent K+ conductance in LLC-PK1 cells.