Peroxynitrite triggers a delayed resistance of coronary endothelial cells against ischemia-reperfusion injury

Experiments were designed to test whether nitric oxide (NO) and peroxynitrite trigger delayed coronary endothelial protection induced by preconditioning (PC) in rats. Prolonged ischemia reperfusion markedly reduced the response of isolated coronary arteries to acetylcholine, and this was prevented by PC performed 24 h earlier. The NO synthase (NOS) inhibitor N(G)-nitro-L-arginine methyl ester (L-NAME) administered during PC abolished its delayed endothelial protective effect, whereas the inducible NOS inhibitor N-(3(aminomethyl)benzyl)acetaminide had no effect. Delayed endothelial PC was also abolished by the peroxynitrite scavengers selenomethionine or uric acid given during PC. In parallel, the NO/peroxynitrite donor S-morpholinosydnonimine and authentic peroxynitrite, administered 24 h before prolonged ischemia-reperfusion mimicked endothelial PC, whereas the NO donor S-nitroso-N-acetylpencillamine had no effect. This suggests that peroxynitrite is an essential trigger of the delayed coronary endothelial protection induced by PC in rat hearts.     

Superoxide- and nitric oxide-derived species mediate endothelial dysfunction, endothelial glycocalyx disruption, and enhanced neutrophil adhesion in the post-ischemic guinea-pig heart.

The study was aimed at testing the hypothesis that a toxic product of the reaction between superoxide (O(2)(-)) and nitric oxide (NO) mediates, not only endothelial dysfunction, but also endothelium-glycocalyx disruption, and increased neutrophil (PMN) accumulation in the heart subjected to ischemia/reperfusion (IR) injury. Accordingly, we studied if scavengers of either O(2)(-) or NO, or a compound that was reported to attenuate cardiac production of peroxynitrite, would prevent endothelial injury and subsequent PNM adhesion in IR heart. Langendorff-perfused guinea-pig hearts were subjected to 30 min ischemia/35 min reperfusion, and infusion of PMN between 15 and 25 min of the reperfusion. Coronary flow responses to acetylcholine (ACh) and sodium nitroprusside (SNP) were used as measures of endothelium-dependent and -independent vascular function, respectively. PMN adhesion and endothelium glycocalyx ultrastructure were assessed in histological preparations. IR impaired the ACh, but not SNP, response by approximately 60%, caused endothelium-glycocalyx disruption, and approximately nine-fold increase in PMN adhesion. These alterations were prevented by superoxide dismutase (150 U/ml), NO synthase inhibitor, L-NAME (10 microM), NO scavenger, oxyhemoglobin (25 microM), and NO donor, SNAP (1 microM), and were not affected by catalase (600 u/ml). The glycocalyx-protective effect of these interventions preceded their effect on PMN adhesion. The data imply that PMN adhesion in IR guinea-pig heart is a process secondary to functional and/or structural changes in coronary endothelium, and that a toxic product of the reaction between superoxide and NO mediates these endothelial changes.     

Deficient iNOS in inflammatory bowel disease intestinal microvascular endothelial cells results in increased leukocyte adhesion

Microvascular endothelial cells play a key role in inflammation by undergoing activation and recruiting circulating immune cells into tissues and foci of inflammation, an early and rate-limiting step in the inflammatory process. We have previously [Binion et al., Gastroenterology112:1898-1907, 1997] shown that human intestinal microvascular endothelial cells (HIMEC) isolated from surgically resected inflammatory bowel disease (IBD) patient tissue demonstrate significantly increased leukocyte binding in vitro compared to normal HIMEC. Our studies [Binion et al., Am. J. Physiol.275 (Gastrointest. Liver Physiol. 38):G592-G603, 1998] have also demonstrated that nitric oxide (NO) production by inducible nitric oxide synthase (iNOS) normally plays a key role in downregulating HIMEC activation and leukocyte adhesion. Using primary cultures of HIMEC derived from normal and IBD patient tissues, we sought to determine whether alterations in iNOS-derived NO production underlies leukocyte hyperadhesion in IBD. Both nonselective (N(G)-monomethyl-L-arginine) and specific (N-Iminoethyl-L-lysine) inhibitors of iNOS significantly increased leukocyte binding by normal HIMEC activated with cytokines and lipopolysaccharide (LPS), but had no effect on leukocyte adhesion by similarly activated IBD HIMEC. When compared to normal HIMEC, IBD endothelial cells had significantly decreased levels of iNOS mRNA, protein, and NO production following activation. Addition of exogenous NO by co-culture with normal HIMEC or by pharmacologic delivery with the long-acting NO donor detaNONOate restored a normal leukocyte binding pattern in the IBD HIMEC. These data suggest that loss of iNOS expression is a feature of chronically inflamed microvascular endothelial cells, which leads to enhanced leukocyte binding, potentially contributing to chronic, destructive inflammation in IBD.     

Peroxynitrite activates NF-E2-related factor 2/antioxidant response element through the pathway of phosphatidylinositol 3-kinase: the role of nitric oxide synthase in rat glutathione S-transferase A2 induction.

The protective adaptive response to electrophiles and reactive oxygen species is mediated by the induction of phase II detoxifying genes through antioxidant response elements (AREs). Our previous study showed that sulfur amino acid deprivation (SAAD) produces peroxides and induces rat glutathione S-transferase A2 (rGSTA2) through NF-E2-related factor 2 (Nrf2)/ARE activation via the pathway of phosphatidylinositol 3-kinase (PI3-kinase). The current study was designed to investigate the role of peroxynitrite in Nrf2/ARE activation and rGSTA2 induction. L-Arginine deficiency or N(G)-nitro-L-arginine methyl ester (L-NAME) reduced peroxide production induced by SAAD in H4IIE cells. Northern and Western blot analyses revealed that the levels of rGSTA2 mRNA and protein were significantly increased 24h after incubation of the cells in SAAD medium, which was inhibited by L-arginine deficiency or L-NAME. Subcellular fractionation and gel shift analyses revealed that SAAD increased the level of nuclear Nrf2 and activated ARE, which were also blocked by L-arginine deficiency or L-NAME. Whereas the exogenous NO donor S-nitroso-N-acetyl-penicillamine (SNAP) alone failed to significantly induce rGSTA2, SNAP enhanced SAAD-inducible rGSTA2 expression, verifying the notion that peroxynitrite derived from NO contributes to rGSTA2 induction. 3-Morpholinosydnonimine (SIN-1), which decomposes and yields peroxynitrite, increased the rGSTA2 mRNA and protein levels in a dose-dependent manner. SIN-1 increased the level of nuclear Nrf2 and activated Nrf2/ARE, which was supershifted by anti-Nrf2 and anti-Maf antibodies. SIN-1 increased the activity of PI3-kinase, as monitored by phosphorylation of Akt. SIN-1-inducible rGSTA2 expression was inhibited by PI3-kinase inhibitors. These results provide evidence that peroxynitrite plays an essential role in nuclear translocation of Nrf2 and ARE activation through the pathway of PI3-kinase and that nitric oxide synthase is involved in the induction of rGSTA2.     

Endogenous reactive oxygen species are essential for proliferation of neural stem/progenitor cells

It is widely thought that accumulation of reactive oxygen species (ROS) causes injury to cells. In this study, we investigated the effect of endogenous ROS on the proliferation of neural stem/progenitor cells derived from the hippocampus of embryonic mice. The cells were treated with free radical-scavenging agents [3-methyl-1-phenyl-2-pyrazolin-5-one (edaravone) or 4-hydroxy-2,2,6,6-tetramethylpiperidine 1-oxyl (tempol)], an NADPH oxidase inhibitor (apocynin), catalase, a nitric oxide synthase inhibitor [N^ω-nitro-l-arginine methyl ester hydrochloride (L-NAME)] or a peroxynitrite generator (SIN-1) during the culture period. Edaravone and tempol had the ability to decrease endogenous ROS in the cells exposed for periods from 1 to 24 h, with attenuation of the proliferation activity of the cells during culture. Apocynin and L-NAME were also effective in attenuating cell proliferation but not cellular damage. Conversely, SIN-1 was capable of promoting the proliferation activity. However, catalase had no effect on the proliferation activity of the cells during culture. Furthermore, tempol significantly decreased the level of NFκB p65, phospho-cyclic AMP response element-binding protein, and β-catenin within the nucleus of the cells. These data suggest that endogenous ROS and nitric oxide are essential for the proliferation of embryonic neural stem/progenitor cells.     

Cathepsin B is a differentiation-resistant target for nitroxyl (HNO) in THP-1 monocyte/macrophages

We previously showed that the one-electron reduction product of nitric oxide (NO), nitroxyl (HNO), irreversibly inhibits the proteolytic activity of the model cysteine protease papain. This result led us to investigate the differential effects of the nitrogen oxides, such as nitroxyl (HNO), NO, and in situ-generated peroxynitrite on cysteine modification-sensitive cellular proteolytic enzymes. We used Angeli's salt, diethylaminenonoate (DEA/NO), and 3-morpholinosydnoniminehydrochloride (SIN-1), as donors of HNO, NO, and peroxynitrite, respectively. In this study we evaluated their inhibitory activities on the lysosomal mammalian papain homologue cathepsin B and on the cytosolic 26S proteasome in THP-1 monocyte/macrophages after LPS activation or TPA differentiation. HNO-generating Angeli's salt caused a concentration-dependent (62 +/- 4% at 316 muM) inhibition of the 26S proteasome activity, resulting in accumulation of protein-bound polyubiquitinylated proteins in LPS-activated cells, whereas neither DEA/NO nor SIN-1 showed any effect. Angeli's salt, but not DEA/NO or SIN-1, also caused (94 +/- 2% at 316 muM) inhibition of lysosomal cathepsin B activity in LPS-activated cells. Induction of macrophage differentiation did not significantly alter the inhibitory effect of HNO on lysosomal cathepsin B activity, but protected the proteasome from HNO-induced inhibition. The protection awarded by macrophage differentiation was associated with induction of the GSH synthesis rate-limiting enzyme gamma-glutamylcysteine synthetase, as well as with increased intracellular GSH. In conclusion, HNO abrogates both lysosomal and cytosolic proteolysis in THP-1 cells. Macrophage differentiation, associated with upregulation of antioxidant defenses such as increased cellular GSH, does not protect the lysosomal cysteine protease cathepsin B from inhibition.     

Reactive nitrogen species control apoptosis and autophagy in K562 cells: implication of TAp73α induction in controlling autophagy

The biological outcome of nitric oxide (NO) and reactive nitrogen species (RNS) in regulating pro survival and pro death autophagic pathways still demand further investigation. In the present study, we investigated the effect of nitrosative stress in K562 cells using NO donor compound DETA-NONOate, peroxynitrite, and SIN-1. Exposure to NO, peroxynitrite, and SIN-1 caused decrease in K562 cell survival. NO induced autophagy but not apoptosis or necrosis in K562 cells. In contrast, peroxynitrite and SIN-1 treatment induced apoptosis in K562 cells. Surprisingly, inhibition of autophagic response using 3-methyladenine led to the induction of apoptosis in K562 cells. Increase in 5’adenosine monophosphate-activated protein kinase (AMPK) phosphorylation was only observed in the presence of NO donor indicated that AMPK was crucial to induce autophagy in K562 cells. We for the first time discovered a novel role of p73 in autophagy induction under nitrosative stress in K562 cells. TAp73α was only induced upon exposure to NO but not in the presence of peroxynitrite. Reduced glutathione (GSH)/oxidised glutathione (GSSG) ratio remained unaltered upon NO exposure. Our data suggest a complex network of interaction and cross regulations between NO and p73. These data open a new path for therapies based on the abilities of RNS to induce autophagy-mediated cell death.     

Nitric oxide mediates the mitogenic effects of insulin and vascular endothelial growth factor but not of leptin in endothelial cells

The regulation of vascular wall homeostasis by nitric oxide (NO) generated by endothelium is being intensively studied. In the present paper, the involvement of NO in the vascular endothelial growth factor (VEGF), insulin or leptin-stimulated proliferation of human endothelial cells (HUVEC) was measured by [3H]thymidine or bromodeoxyuridine incorporation. VEGF and insulin, but not leptin, increased NO generation in HUVEC, as detected with ISO-NO electrode. Proliferation of HUVEC induced by leptin was not changed or was higher in the presence of N(omega)-nitro-L-arginine methyl ester (L-NAME) a nitric oxide synthase (NOS) inhibitor. In contrast, L-NAME blunted the proproliferative effect of VEGF and insulin. Furthermore, we demonstrated that, in human arterial smooth muscle cells (hASMC) transfected with endothelial NOS (eNOS) gene, the generation of biologically active VEGF protein was NO-dependent. Inhibition of NO generation by L-NAME decreased the synthesis of VEGF protein and attenuated HUVEC proliferation induced by conditioned media from transfected hASMC. Endothelium-derived NO seems to participate in VEGF and insulin, but not leptin, mitogenic activity. Additionally, the small amounts of NO released from endothelial cells, as mimicked by eNOS transfection into hASMC, may activate generation of VEGF in sub-endothelial smooth muscle cells, leading to increased synthesis of VEGF protein necessary for turnover and restitution of endothelial cells.     

C-peptide inhibits leukocyte-endothelium interaction in the microcirculation during acute endothelial dysfunction.

C-peptide is a cleavage product that comes from processing proinsulin to insulin that induces nitric oxide (NO) -mediated vasodilation. NO modulates leukocyte-endothelium interaction. We hypothesized that C-peptide might inhibit leukocyte-endothelium interaction via increased release of endothelial NO. Using intravital microscopy of the rat mesentery, we measured leukocyte-endothelium interactions after administration of C-peptide to the rat. Superfusion of the rat mesentery with either thrombin or L-NAME consistently and significantly increased the number of rolling, adhering, and transmigrated leukocytes. C-peptide significantly attenuated either thrombin- or L-NAME-induced leukocyte-endothelium interactions in rat mesenteric venules. A control scrambled sequence of C-peptide characterized by the same amino acid composition in a randomized sequence failed to inhibit leukocyte-endothelium interactions. These effects of C-peptide were associated with decreased surface expression of the cell adhesion molecules P-selectin and ICAM-1 on the microvascular endothelium. Endothelial nitric oxide synthase (eNOS) mRNA levels were increased in rats injected with C-peptide. This enhanced eNOS expression was associated with a marked increase in basal NO release from the aorta of C-peptide-treated rats. We conclude that C-peptide is a potent inhibitor of leukocyte-endothelium interaction and that this effect is specifically related to inhibition of endothelial cell adhesion molecules via maintenance of NO release from the vascular endothelium.     

Xanthine oxidase converts nitric oxide to nitroxyl that inactivates the enzyme

Xanthine oxidase (XO) was found to convert nitric oxide (NO* ) released from spermine-NONOate to nitroxyl (HNO), the one-electron reduction product of NO*, in the presence of its substrate hypoxanthine under anaerobic conditions. Under these conditions, XO lost its activity. Upon aerobic incubation of XO with its substrate, neither conversion of NO* to HNO nor inactivation of the enzyme was observed. Angeli's salt (an HNO generator) or synthetic peroxynitrite inactivated XO at low concentrations, whereas high concentrations of diethylamine-NONOate (an NO* donor) and SIN-1 (which generates peroxynitrite by releasing both NO* and superoxide) were required to inactivate XO. These results suggest that HNO generated by XO under anaerobic conditions inactivates XO. As both XO and NO* synthase are activated and/or induced in ischemia-reperfusion injury, HNO formed by XO may contribute to pathogenesis by exerting its potent oxidation activity against a variety of biological compounds.     

Effect of centrally administered nitric oxide modulators in Brewer's yeast-induced nociception in rats

Possible modulation of Brewer's yeast-induced nociception by centrally (icv) administered nitric oxide (NO) modulators, viz., NO synthase (NOS) inhibitors, NO precursor, donors, scavengers and co-administration of NO donor (SIN-1) with NOS inhibitor (L-NAME) and NO scavenger (Hb) was investigated in rats. Administration of NOS inhibitors and NO scavenger Hb increased the pain threshold capacity significantly, whereas NO donors SIN-1, SNP and NO precursor L-arginine were found to be hyperalgesic. D-arginine, the inactive isomer of L-arginine and methylene blue, inhibitor of soluble guanylate cyclase failed to alter the nociceptive behaviour in rats. Co-administration of SIN-1 with L-NAME and Hb found to increase the nociceptive threshold. The results indicate, that centrally administered NO modulators alter the nociceptive transmission induced by Brewer's yeast in rats.     

肾动脉内注射L-精氨酸抑制肾神经传入纤维的自发活动

研究旨在应用记录肾传人神经多单位和单位放电的方法,观察肾动脉内注射L—精氨酸对麻醉家兔肾神经传人纤维自发放电活动的影响。结果表明：(1）肾动脉内注射L—精氨酸(0．05、0．24和0．48mmol/kg)可呈剂量依赖性地抑制肾传人纤维的活动,而动脉血压不变;(2)静脉内预先注射一氧化氮合酶抑制剂L—NAME(0．11mmol/kg),可完全阻断L—精氨酸对肾传人纤维的抑制;(3)肾动脉注射一氧化氮(N0)供体SIN—1(3．75μmol/kg)也可抑制肾传入神经的活动。以上结果提示：肾动脉内应用N0前体L—精氨酸和N0供体SIN—1均可抑制肾传入纤维的自发活动。     

Nitric oxide-an endothelial cell survival factor

Due to its unique position in the vessel wall, the endothelium acts as a barrier and thereby controls adhesion, aggregation and invasion of immune competent cells. Apoptosis of endothelial cells may critically disturb the integrity of the endothelial monolayer and contribute to the initiation of proinflammatory events. Endothelial cell apoptosis is counteracted by nitric oxide synthesised by the endothelium nitric oxide synthase (eNOS). Thus, nitric oxide inhibits endothelial cell apoptosis induced by proinflammatory cytokines and proatherosclerotic factors including reactive oxygen species and angiotensin II. The apoptosis-suppression may contribute to the profound anti-inflammatory and anti-atherosclerotic effects of endothelial-derived NO. Furthermore, the support of endothelial cell survival by NO may further play a central role for the pro-angiogenic effects of NO.     

Nitric oxide augments oridonin-induced efferocytosis by human histocytic lymphoma U937 cells via autophagy and the NF-κB-COX-2-IL-1β pathway

Abstract We previously demonstrated that oridonin-induced autophagy enhanced efferocytosis (phagocytosis of apoptotic cells) by macrophage-like U937 cells through activation of the inflammatory pathways. In this study, exposure of U937 cells to 2.5 μM oridonin caused up-regulation of inducible nitric oxide synthase (iNOS) expression and continuous endogenous generation of nitric oxide (NO), which was reversed by pre-treatment with the inhibitors of nitric oxide synthase 1400 W (dihydrochloride) or L-NAME (hydrochloride). NO donor sodium nitroprusside (SNP) and efferocytosis irritant lipopolysaccharide (LPS) could also exert NO generation and iNOS expression. Moreover, oridonin-induced stimulation of efferocytosis was significantly suppressed by 1400 W or L-NAME. In addition, 1400 W or L-NAME impaired oridonin-induced autophagy. Inhibition of autophagy with 3-methyladenine (3MA) or Beclin-1 siRNA attenuated the uptake of apoptotic cells with a slight increase in the production of NO. The pro-inflammatory cytokine interleukin-1β (IL-1β) has been reported to be involved in oridonin-induced efferocytosis in U937 cells and interact with NO to contribute to inflammatory responses. 1400 W or L-NAME blocked the secretion of IL-1β and the activation of NF-κB and COX-2. Provision of SNP or LPS in place of oridonin resulted in the similar enhancement of efferocytosis, autophagy, the release of IL-1β and the expression of signal protein. NO augmented the oridonin-induced efferocytosis by mediating autophagy and activating the NF-κB-COX-2-IL-1β pathway. Inhibition of NF-κB or COX-2 in turn decreased the production of NO and the expression of iNOS. There exists a positive feedback loop between NO generation and NF-κB-COX-2-IL-1β pathway.     

Evidence against tetrahydrobiopterin depletion of vascular tissue exposed to nitric oxide/superoxide or nitroglycerin

Several cardiovascular disorders, including atherosclerosis and tolerance to the antianginal drug nitroglycerin (GTN), may be associated with the generation of superoxide anions, which react with nitric oxide (NO) to yield peroxynitrite. According to a widely held view, oxidation of tetrahydrobiopterin (BH₄) by peroxynitrite causes uncoupling of endothelial NO synthase (eNOS), resulting in reduced NO bioavailability and endothelial dysfunction under conditions of oxidative stress. In this study we determined the levels of reduced biopterins and endothelial function in cultured cells exposed to peroxynitrite and GTN as well as in blood vessels isolated from GTN-tolerant guinea pigs and rats. BH₄ was rapidly oxidized by peroxynitrite and 3-morpholino sydnonimine (SIN-1) in buffer, but this was prevented by glutathione and not observed in endothelial cells exposed to SIN-1 or GTN. Prolonged treatment of the cells with 0.1 mM GTN caused slow N^G-nitro-l-arginine-sensitive formation of reactive oxygen species without affecting eNOS activity. Endothelial function and BH₄/BH₂ levels were identical in blood vessels of control and GTN-tolerant animals. Our results suggest that peroxynitrite-triggered BH₄ oxidation does not occur in endothelial cells or GTN-exposed blood vessels. GTN seems to trigger minor eNOS uncoupling that is unrelated to BH₄ depletion and without observable consequence on eNOS function.     

Nitric oxide modulates endotoxin-induced platelet-endothelial cell adhesion in intestinal venules

Although platelets have been implicated in the pathogenesis of vascular diseases, little is known about factors that regulate interactions between platelets and the vessel wall under physiological conditions. The objectives of this study were to 1) define the contribution of nitric oxide (NO) to endotoxin (lipopolysaccharide, LPS)-induced platelet-endothelial cell (P/E) adhesion in murine intestinal venules and 2) determine whether the antiadhesive action of NO is mediated by soluble guanylate cyclase (sGC). Adhesive interactions between platelets and endothelial cells were monitored by intravital microscopy. LPS administration into control wild-type mice (WT) resulted in a >15-fold increase in P/E adhesion. Similar responses were observed using endothelial NO synthase (eNOS)-deficient platelets. However, treatment with the NO donor diethylenetriamine-nitric oxide (DETA-NO) attenuated the P/E adhesion response to LPS, whereas the NO synthase inhibitor N(G)-nitro-L-arginine methyl ester or eNOS deficiency resulted in an exacerbation. P/E adhesion response did not differ between LPS-treated WT and inducible NOS-deficient mice. Inhibition of sGC abolished the attenuating effects of DETA-NO, whereas the sGC activator 3-(5'-hydroxymethyl-2'-furyl)-1-benzyl indazole (YC-1) reduced LPS-induced P/E adhesion. These findings indicate that 1) eNOS-derived NO attenuates endotoxin-induced P/E adhesion and 2) sGC is responsible for the antiadhesive action of NO.     

Free radicals alter maximal diaphragmatic mitochondrial oxygen consumption in endotoxin-induced sepsis

Recent studies indicate that sepsis is associated with enhanced generation of several free radical species (nitric oxide, superoxide, hydrogen peroxide) in skeletal muscle. While studies suggest that free radical generation causes uncoupling of oxidative phosphorylation in sepsis, no previous report has examined the role of free radicals in modulating skeletal muscle oxygen consumption during State 3 respiration or inhibiting the electron transport chain in sepsis. The purpose of the present study was to examine the effects of endotoxin-induced sepsis on State 3 diaphragm mitochondrial oxygen utilization and to determine if inhibitors/scavengers of various free radical species would protect against these effects. We also examined mitochondrial protein electrophoretic patterns to determine if observed endotoxin-related physiological derangements were accompanied by overt alterations in protein composition. Studies were performed on: (a) control animals, (b) endotoxin-treated animals, (c) animals given endotoxin plus PEG-SOD, a superoxide scavenger, (d) animals given endotoxin plus L-NAME, a nitric oxide synthase inhibitor, (e) animals given only PEG-SOD or L-NAME, (f) animals given endotoxin plus D-NAME, and (g) animals given endotoxin plus denatured PEG-SOD. We found: (a) no alteration in maximal State 3 mitochondrial oxygen consumption rate at 24 h after endotoxin administration, but (b) a significant reduction in oxygen consumption rate at 48 h after endotoxin, (c) no effect of endotoxin to induce uncoupling of oxidative phosphorylation, (d) either PEG-SOD or L-NAME (but neither denatured PEG-SOD nor D-NAME) prevented endotoxin-mediated reductions in State 3 respiration rates, (e) some mitochondrial proteins underwent tyrosine nitrosylation at 24 h after endotoxin administration, and (f) SDS-page electrophoresis of mitochondria from endotoxin-treated animals revealed a selective depletion of several proteins at 48 h after endotoxin administration (but not at 24 h); (g) administration of L-NAME or PEG-SOD prevented this protein depletion. These data provide the first evidence that endotoxin-induced reductions in State 3 mitochondrial oxygen consumption are free radical-mediated.