- Email: [email protected]
Substrate and Site Specificity of Hydrogen Peroxide Generation in Mouse Mitochondria
ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS
Vol. 350, No. 1, February 1, pp. 118–126, 1998 Article No. BB970489
Substrate and Site Specificity of Hydrogen Peroxide Generation in Mouse Mitochondria1 Linda K. Kwong and Rajindar S. Sohal2 Department of Biological Sciences, Southern Methodist University, Dallas, Teaxs 75275
Received September 15, 1997, and in revised form October 20, 1997
The objective of this study was to elucidate the mechanisms of mitochondrial H2O2 generation in mouse organs by determining the nature of their differences in substrate utilization, inhibitor sensitivity, and the site specificity affecting H2O2 production. Mitochondria were isolated from heart, brain, and kidney and the rate of H2O2 generation was measured using the FADH-linked substrates succinate and a-glycerophosphate as well as the NADH-linked substrates pyruvate/malate, b-hydroxybutyrate, and glutamate. Respiratory inhibitors, antimycin and rotenone, were added singly and sequentially to each substrate-supported H2O2 generation reaction mixture to determine the mitochondrial site(s) of generation and the optimal condition(s) for maximal rates of generation. Succinate supported the highest rate of mitochondrial H2O2 generation. Moreover, it was the preferred substrate for the heart mitochondria. a-Glycerophosphate is a poor substrate for H2O2 generation in heart mitochondria. Inhibitor studies showed that heart mitochondria were the most sensitive and responsive to antimycin, while brain was the most sensitive to rotenone. A surprising finding was that NADH-linked substratesupported H2O2 generation in kidney mitochondria was not responsive to rotenone. The contribution from each of the three sites (ubiquinone, NADH dehydrogenase, and a-glycerophosphate dehydrogenase) of mitochondrial H2O2 generation to the total was both substrate and organ dependent. Results indicate that assay conditions must be considered before comparisons of sites and rates of mitochondrial H2O2 generation among different organs can be made. q 1998 Academic Press
1 This work was supported by Grants R01AG7657 and R01AG13563 from the National Institute of Health/National Institute on Aging. 2 To whom correspondence and reprint requests should be addressed at Department of Biological Sciences, Southern Methodist University, 220 Fondren Science Building, Dallas, TX 75275. Fax: 214/768-3955. E-mail: [email protected].
Key Words: mitochondria; reactive oxygen species; free radicals; H2O2 generation; antimycin; rotenone.
A widely recognized attribute of aerobic metabolism in eukaryotic cells is the production of reactive oxygen species (ROS3) by mitochondria. One-electron autooxidation of reduced mitochondrial respiratory components is believed to lead to production of superoxide anion radical (O2r0) (1) which is dismutated by superoxide dismutase to form hydrogen peroxide (H2O2), a progenitor of the highly reactive hydroxyl free radical, in a stoichiometric ratio of Ç2 (2). Two main sites of O2r0 and H2O2 generation in the inner mitochondrial membrane, namely ubiquinone at complex III (3) and NADH dehydrogenase at complex I (4), have been identified [see review by Turrens (5)]. However, there appears to be variation among different tissues and species in the sites of ROS generation. The major site of succinate-supported H2O2 generation in the bovine heart submitochondrial particles is at complex III, the ubiquinone site (3); however, in the rat heart (6) and brain (7) mitochondria, quantitatively more H2O2 is generated at complex I than at complex III via reverse electron flow. In the housefly mitochondria, NADH dehydrogenase is not a source of H2O2 (8). Furthermore, in guinea pig cerebral cortex mitochondria, a third site of H2O2 generation, autooxidation of a-glycerophosphate dehydrogenase has also been proposed (9). Thus, comparisons of the locations of ROS generation may be biased by experimental conditions such as substrates and inhibitors used as well as by the source of the samples. Recently, the validity of in vivo mitochondrial O2r0 3 Abbreviations used: PHPA, p-hydroxyphenylacetate; HRP, horseradish peroxidase; ROS, reactive oxygen species; Or0 2 , superoxide; AA, antimycin; rot, rotenone; mln, malonate; SMPs, submitochondrial particles; BSA, bovine serum albumin; SOD, superoxide dismutase; CoQ, coenzyme Q.
118
AID
0003-9861/98 $25.00 Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
arca
H2O2 GENERATION BY MITOCHONDRIA
and H2O2 production was questioned (10). The authors suggested that the production of O2r0 was thermodynamically unfavorable and that it is unreasonable to expect an increase in mitochondrial O2r0 and H2O2 production even under most pathological conditions. However, the current view of oxidative stress is that the mitochondria accumulated oxidative damages as the organism aged which resulted in mitochondrial dysfunction (11). An age-dependent increase in the production of mitochondrial O2r0 and H2O2 has been reported (8, 12). In addition, H2O2 generation was found to be increased in the in vitro oxidatively damaged mitochondria (13). Thus, additional information on the mechanism and the sites of H2O2 production may help to clarify this controversy. The objective of the present study is to validate the generation of H2O2 in mitochondria from mouse heart, brain, and kidney and to compare the relative amounts generated at the different mitochondrial sites using various NADH-linked and FADHlinked substrates in conjunction with the respiratory inhibitor rotenone and antimycin. MATERIALS AND METHODS Materials. All chemicals were of analytical grade and purchased from Sigma Chemical Co. (St. Louis, MO). Concentrated stock solutions of antimycin in ethanol and rotenone in chloroform were stored at 0207C until use. Diluted working solutions were made daily in ethanol for use in each experiment. Stock solution of horseradish peroxidase (EC 1.11.1.7, type VI) was made in H2O2 assay buffer at 100 U/ml. Aliquots were stored at 0207C until use. Animals. C57BL/6NNia male mice were obtained from the National Institute on Aging/National Center for Toxicological Research Project on Caloric Restriction Colonies (Jefferson, AR) and were housed individually at the University of North Texas Health Science Center at Fort Worth. The mice were fed ad libitum with NIH-31 diet (Purina Mills, Inc., Richmond, IN) and maintained on a 12-h light, 12-h dark cycle with the light cycle beginning at 06:00. Detailed description of animal husbandry is described elsewhere (12). For this study, 13- to 15-month-old mice were transported a short distance to the Southern Methodist University Vivarium. After an adaptation period of at least 1 week where the mice were maintained under similar conditions as described above, they were killed by carbon dioxide asphyxiation. Brain, heart, and kidney were rapidly removed, rinsed free of blood, and placed in ice-cold mitochondrial isolation buffer. Male Sprague–Dawley rats were bred at the Southern Methodist University Vivarium and maintained under the Federal Guidelines for Ethical Animal Experimentation. Rats were killed and their hearts removed as described above. Isolation of mitochondria and preparation of submitochondrial particles. Procedures for the isolation of mitochondria were conducted at 47C. Mitochondria were isolated from the brain by the method of Sims (14). The brain was finely minced and homogenized in 10 vol of isolation medium containing 320 mM sucrose, 1 mM K2EDTA, and 10 mM Tris–HCl, pH 7.1, using a ground-glass homogenizer. The homogenate was centrifuged at 1330g for 3 min. The pellet was rehomogenized in 5 vol of isolation medium and centrifuged as before. The combined supernatants were centrifuged at 21,200g for 10 min. The pellet was resuspended in isolation medium containing 15% Percoll (10 ml/g of starting material). The suspension was separated on a discontinuous gradient of Percoll: 3 ml 15%, 3.5
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
119
ml 23%, and 3.5 ml 40%. Separation was achieved by centrifugation at 30,700g for 5 min. Mitochondria banded at the interface of 23 and 40% Percoll (third band). The material was removed, diluted 1:4 with isolation medium, and centrifuged at 16,700g for 10 min. The supernatant was removed to within a few millimeters of the loose pellet. The pellet was gently resuspended in the remaining buffer and fatty-acid-free bovine serum albumin (BSA, 10 mg/ml; 0.5 ml/ brain) was added. The mixture was diluted with isolation medium (3 ml/brain) and centrifuged at 6900g for 10 min. The final pellet was rinsed and resuspended in isolation medium using a Teflon-toglass homogenizer. Mitochondria were isolated from the heart using a modified procedure of Rouslin (15). The ventricles were finely minced and homogenized in 10 vol of isolation medium containing 180 mM KCl, 10 mM EGTA–Tris base, pH 7.2, 0.5% fatty-acid-free BSA, and 10 mM Mops–KOH using a Polytron homogenizer (10 s, Brinkmann Instructions, Westbury, NY). The homogenate was centrifuged at 1000g for 10 min. The resulting supernatant was centrifuged at 17,500g for 10 min. The final pellet was rinsed free of any light or loosely packed damaged mitochondria and resuspended in buffer containing 250 mM sucrose, 1 mM EGTA, 10 mM Mops–KOH, pH 7.2. Submitochondrial particles (SMPs) were prepared as described previously (16). Kidney mitochondria were isolated according to the method of Lash and Sall (17). The cortices of the kidneys were minced and homogenized in 10 vol of isolation buffer containing 220 mM sucrose, 5 mM KH2PO4 , 5 mM MgCl2 , 20 mM KCl, 20 mM triethanolamine–HCl, 0.1 mM phenylmethylsulfonyl fluoride, and 2 mM EGTA, pH 7.4, using a Polytron homogenizer. The homogenate was centrifuged at 600g for 10 min. The pellet was resuspended in the original volume of buffer using a ground-glass homogenizer and centrifuged at 600g for 10 min. The combined supernatants were recentrifuged at 15,000g for 5 min. The pellet was carefully rinsed and resuspended in the isolation buffer without EGTA. Measurement of the rates of H2O2 generation and state 4 respiration. The rate of H2O2 released from the mitochondria was measured fluorometrically by the method of Hyslop and Sklar (18). The reaction mixture (3 ml) contained 75–150 mg of mitochondrial protein, 500 mg p-hydroxyphenylacetate (PHPA), 4 U HRP, and 7 mM substrate in 154 mM KCl, 10 mM KH2PO4 , 3 mM MgCl2 , and 0.1 mM EGTA (pH 7.4). The rate of H2O2 released was determined by following the increase in fluorescence at 377C using a Perkin–Elmer LS-5 spectrofluorometer equipped with a thermal-controlled and magnetic stirring sample compartment (Ex: 320 nm; Em: 400 nm). The specific standard curve for each substrate and inhibitor was determined by measuring the fluorescence of known amounts of H2O2 in assay containing the substrate or inhibitor with no mitochondrial protein. The rate of mitochondrial state 4 (substrate in excess but no exogenous ADP) respiration was measured polarographically using a Clark-type electrode and a YSI Model 5300 biological oxygen monitor (YSI Inc., Yellow Springs, OH) at 307C. The incubation mixture consisted of buffer (same as that used for H2O2 generation) and 80–400 mg of mitochondrial protein. The rate of oxygen consumption was measured after the addition of substrate and after the addition of inhibitor(s). Mitochondrial protein concentration was estimated by the method of Harrington (19) using bovine serum albumin as standard. Measurement of Or0 The rate of Or0 2 generation. 2 generation was measured in SMPs as the SOD-inhibitable reduction of acetylated ferricytochrome c (20). The SOD-inhibitable reaction mixture (1 ml) contained 15–30 mg submitochondrial protein, 7.8 mM acetylated ferricytochrome c, 1.2 mM antimycin, 5 mM rotenone, 7 mM succinate, and 100 U SOD in 100 mM potassium phosphate, pH 7.4. The sample mixture was similar to the SOD-inhibitable except no SOD was added and the inhibitors were added as indicated in the text. The reduction of acetylated ferricytochrome c was monitored kinetically at 550 nm and 377C.
arca
120
KWONG AND SOHAL
Effect of respiratory substrates and inhibitors on the quantitation of H2O2 . The milieu of dye-coupled fluorescent reactions may have an effect on the quantitation of the fluorescence. Thus, the effect of different mitochondrial respiratory substrates and inhibitors on PHPA–HRP fluorescent detection of H2O2 was assessed. As shown in Fig. 2, antimycin and rotenone had minimal effect on the fluorescent response of the H2O2 standards. However, succinate slightly lowered the fluorescent response, and pyruvate/malate appeared to significantly quench the fluorescence. Therefore, different H2O2 standard curves were established for each substrate used in this study.
FIG. 1. Catalase inhibition of PHPA–HRP quantitation of hydrogen peroxide generated in the mouse heart mitochondria. H2O2 generation in the mouse heart mitochondria (200 mg) was monitored as described under Materials and Methods. Succinate (7 mM) was added to initiate the reaction. Catalase (200 U) was added as indicated.
Respiratory substrate selectivity of H2O2 generation by mitochondria from different organs. Mitochondria metabolize a variety of respiratory substrates to generate NADH and FADH2 . It is conceivable that in vivo the mitochondria from different organs have selective substrate preference. Therefore, the abilities of various NADH and FADH-linked substrates to support in vitro H2O2 generation in mitochondria isolated from heart, brain, and kidney were compared (Table I). In the mouse heart mitochondria, succinate supported the highest rate of H2O2 generation with a-glycerophosphate supporting the lowest, Ç5% of the rate measured with succinate. The pyruvate/malate-, glutamate-, and b-hydroxybutyrate-supported rates were Ç15–30% that with succinate. In the brain and kidney mitochondria, succinate also supported the highest rate. However, in the presence of the other four substrates brain
RESULTS
Specificity of the dye-coupled fluorescent assay for H2O2 . The rates of extramitochondrial release of H2O2 , i.e., H2O2 generation, are often measured by assays coupled to horseradish peroxidase formation of compound II, which reacts with an electron donor that either gains or loses fluorescent capacity (21). In this study, the rate of H2O2 generation is monitored by the increase in fluorescence due to the accumulation of the oxidized form of PHPA (18). However, other reaction(s) in the mitochondrial sample may also oxidize PHPA and contribute to the measured fluorescence. Therefore, catalase, which catalyzes the reduction of H2O2 without the coupled oxidation of PHPA, is added to the reaction to determine if the increase in fluorescence is H2O2 dependent. As seen in Fig. 1, the rate of succinatesupported increase in fluorescence was lessened with the addition of catalase. Further addition of catalase continued to diminish the rate. Thus, the measured fluorescent response is dependent on the availability of H2O2 .
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
FIG. 2. Effects of respiratory substrates and inhibitors on the PHPA–HRP quantitation of hydrogen peroxide. Reaction condition was as described under Materials and Methods. (j) No addition; (n) 0.5 mM antimycin; (l) 1 mM rotenone; (h) 7 mM succinate; and (.) 5 mM/1 mM pyruvate/malate.
arca
H2O2 GENERATION BY MITOCHONDRIA TABLE I
Substrate Specificity of Hydrogen Release from Mouse Heart, Brain, and Kidney Mitochondria Extramitochondrial H2O2 releasea Substrate Succinate (7 mM) Pyruvate/malate (5 mM/1 mM) a-Glycerophosphate (7 mM) b-Hydroxybutyrate (7 mM) Glutamate (7 mM)
Heart
Brain
Kidney
317 { 43 (n Å 7) 51 { 18 (n Å 6) 18 { 15 (n Å 2) 71 { 34 (n Å 5) 98 { 16 (n Å 4)
387 { 37 (n Å 7) 149 { 22 (n Å 9) 148 { 29 (n Å 5) 118 { 39 (n Å 5) 157 { 42 (n Å 4)
222 { 26 (n Å 6) 171 { 37 (n Å 3) 151 { 20 (n Å 6) 145 { 16 (n Å 2) 180 { 11 (n Å 4)
Note. Extramitochondrial H2O2 release was measured as described under Materials and Methods. The reaction was initiated by the addition of specific substrate. Values represent means { SE. a Rate is expressed as pmol H2O2/mg protein/min.
and kidney mitochondria generated H2O2 at a rate near 50% of that with succinate. Effect of respiratory inhibitors on H2O2 generation in mitochondria from different organs. The rate of H2O2 generation by isolated mitochondria depends on their metabolic state (22). The highest rate was observed when the components of the electron transport in the mitochondria were highly reduced (state 4). Thus, respiratory inhibitors which block electron flow through the mitochondria and produce a full reduction of all components located on the substrate side of the inhibition site would increase the rate H2O2 generation at a site(s) between the substrate and inhibitor. The effects of rotenone, an inhibitor at complex I, and antimycin, an inhibitor at complex III, on the rate of H2O2 generation in mitochondria from heart, brain, and kidney were compared. In the succinate-supported H2O2 generation, heart mitochondria were the most responsive to antimycin (Fig. 3A) and were also the most sensitive. At 0.25 mM antimycin, the rate of H2O2 generation in the heart mitochondria was maximally stimulated, greater than 4-fold that of control. However, at the same concentration of antimycin, the rates in the brain and kidney mitochondria increased approximately 2and 1.4-fold that of control, respectively. The rate in the brain and kidney mitochondria reached maximal at 0.75 mM, approximately 4- and 2-fold that of control, respectively. When rotenone was added to the reaction mixture containing pyruvate/malate, a NADH-linked substrate, the rate of H2O2 generation in brain mitochondria was the most sensitive and responsive (Fig. 3B). At 0.25 mM rotenone, it was maximally stimulated to approximately sevenfold that of control. Although the pyr-
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
121
uvate/malate mixture was a marginal substrate for H2O2 generation in heart mitochondria (see Table I), the addition of rotenone (0.5 mM) maximally stimulated the rate to approximately sixfold that of control. However, in the kidney mitochondria, the rate of H2O2 generation was not stimulated by the addition of rotenone. Furthermore, concentrations above 1 mM rotenone decreased the rate of H2O2 generation by Ç50%. Similar concentration effects of rotenone were observed when another NADH-linked substrate, b-hydroxybutyrate, was used (data not shown). Effects of sequential addition of inhibitors on H2O2 generation. To better understand the characteristics of the two main sites of H2O2 generation in the mammalian mitochondria, rotenone and antimycin were added sequentially to FADH-linked and NADH-linked substrate-supported reactions. FADH-linked substrates succinate and a-glycerophosphate were used to measure H2O2 generation at the ubiquinone site. Antimycin was added to block electron flow between cytochrome b and ubiquinone, while rotenone was added to prevent backflow of electrons to the NADH dehydrogenase site. The addition of antimycin differentially increased the rates of H2O2 generation to 4.0-fold (heart), 3.2-fold (brain), and 1.8-fold (kidney) that of the controls (Table II). Further addition of rotenone decreased the rates, but the levels were still higher than those measured with succinate alone. When rotenone was added before antimycin, the rates decreased to approximately 50% of succinate controls. Subsequent addition of antimycin increased the rates to levels similar to those measured with reversed addition of the inhibitors. These results suggest that with succinate as substrate H2O2 generation occurred at both the ubiquinone site and the NADH dehydrogenase site in the heart, brain, and kidney mitochondria. In a-glycerophosphate-supported H2O2 generation, the addition of antimycin greatly increased the rate of generation in mitochondria from all three organs (Table II). Sequential addition of rotenone to the antimycin-supplemented reaction further increased the rates slightly. On the other hand, incubation with rotenone only slightly stimulated the rates above those of substrate control, but further addition of antimycin greatly increased the rates. This suggests that for this substrate the majority of H2O2 was produced at the ubiquinone site. Thus, the main site of a-glycerophosphate-supported H2O2 generation in the mouse heart, brain, and kidney mitochondria was different from that of succinate. NADH-linked substrates pyruvate/malate, b-hydroxybutyrate, and glutamate were used to measure H2O2 generation at both the NADH dehydrogenase site and the ubiquinone site. Again, rotenone and antimycin were added singly and sequentially to estimate the pro-
arca
122
KWONG AND SOHAL
FIG. 3. Effect of respiratory inhibitors on hydrogen peroxide generation. Reaction condition was as described under Materials and Methods. (A) Effect of antimycin on succinate (7 mM)-supported H2O2 generation in the heart (j), brain (h), and (.) kidney mitochondria. (B) Effect of rotenone on pyruvate/malate (5 mM/1 mM)-supported H2O2 generation in the heart (j), brain (h), and (.) kidney mitochondria.
portion of H2O2 generated at each of the sites. As shown in Table III, rotenone when added to pyruvate/malatesupported H2O2 generation increased the rate in the heart and brain mitochondria approximately six- and eightfold that of controls, respectively. Further addition of antimycin did not change the rate substantially. Addition of antimycin before rotenone also greatly increased the rate. Subsequent addition of rotenone had a slight stimulatory effect on the rate in the brain mitochondria. In contrast, it had a slight inhibitory effect on the rate in the heart mitochondria. Surprisingly, in the kidney mitochondria rotenone had little effect on H2O2 generation, while antimycin had a stimulatory effect. However, the order of addition appeared to have an influence on H2O2 generation. Rotenone obliterated the stimulatory effect of antimycin, while antimycin increased the rate of rotenone-supplemented H2O2 generation. In addition, the observed antimycin-stimulated increases in the mitochondria of all organs were not sensitive to malonate, a potent competitive inhibitor of succinate dehydrogenase. This suggests that succinate dehydrogenase was not involved in pyruvate/ malate-supported H2O2 generation. In the b-hydroxybutyrate-supported H2O2 generation, rotenone did not appear to have an effect on heart mitochondria, but had a small stimulatory effect on brain and kidney mitochondria. Furthermore, subsequent addition of antimycin decreased the rate in heart and brain mitochondria, but not in kidney. However, addition of antimycin before rotenone greatly increased the rates in the mitochondria of all three organs and these increases were not sensitive to malonate, but were to rotenone. Similar results were obtained from
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
the glutamate-supported system with the inhibitors except that the antimycin-stimulated increases were sensitive to malonate as well as to rotenone. Maximum potential rate of H2O2 generation in heart, brain, and kidney mitochondria. The generation of
TABLE II
Effects of Rotenone and Antimycin on FADH-Linked Substrate-Supported Hydrogen Peroxide Release from Heart, Brain, and Kidney Mitochondria Relative extramitochondrial H2O2 release Substrates Succinate Succinate / AA Succinate / AA / rot Succinate / rot Succinate / rot / AA a-Glycerophosphate a-Glycerophosphate / a-Glycerophosphate / a-Glycerophosphate / a-Glycerophosphate /
AA AA / rot rot rot / AA
Heart
Brain
Kidney
1 4.0 1.9 0.8 2.0 1 4.1 5.1 2.8 5.5
1 3.2 1.4 0.5 1.9 1 7.2 7.6 1.2 8.8
1 1.8 1.3 0.4 1.2 1 5.9 6.7 1.2 5.2
Note. Relative extramitochondrial H2O2 release is expressed as a ratio of the rate measured with inhibitor(s) to that of substrate alone. The initial rate of H2O2 generation in the mitochondrial reaction mixture was measured in the presence of substrate, and then the inhibitor(s) was added in the order indicated and the resulting rate was measured. Concentrations of substrates and inhibitors were as followed: 7 mM succinate, 7 mM a-glycerophosphate, 0.25 mM (heart) and 0.75 mM (brain and kidney) antimycin (AA), and 0.5 mM (heart and kidney) and 1 mM (brain) rotenone (rot).
arca
H2O2 GENERATION BY MITOCHONDRIA TABLE III
Effects of Rotenone and Antimycin on NADH-Linked Substrate-Supported Hydrogen Peroxide Release from Heart, Brain, and Kidney Mitochondria Relative extramitochondrial H2O2 release Substrates
Heart
Brain
Kidney
Pyruvate/malate Pyruvate/malate / rot Pyruvate/malate / rot / AA Pyruvate/malate / AA Pyruvate/malate / AA / rot Pyruvate/malate / AA / mln b-Hydroxybutyrate b-Hydroxybutyrate / rot b-Hydroxybutyrate / rot / AA b-Hydroxybutyrate / AA b-Hydroxybutyrate / AA / rot b-Hydroxybutyrate / AA / mln Glutamate Glutamate / rot Glutamate / rot / AA Glutamate / AA Glutamate / AA / rot Glutamate / AA / mln
1 6.2 6.6 11.0 8.5 11.0 1 1.0 0.3 6.6 1 5.4 1 1.2 1.7 4.8 1.6 1.4
1 8.0 7.5 5.4 6.7 5.4 1 2.2 1.1 5.6 2.2 5.1 1 1.0 0.6 6.7 1.0 3.1
1 0.9 3.3 2.5 .8 2.5 1 1.4 2.4 2.8 1.3 2.8 1 0.7 0.4 3.8 0.8 2.2
Note. Relative extramitochondrial H2O2 release was measured as described in the note to Table 2. Concentrations of substrates and inhibitors were as followed: 5 mM/1 mM pyruvate/malate, 7 mM bhydroxybutyrate, 7 mM glutamate, 0.5 mM (brain and kidney) and 0.75 mM (heart) antimycin (AA), 1 mM rotenone (rot), and 5 mM malonate (mln).
H2O2 by mitochondria is considered to have much physiological significance. It is thought to be associated with oxidative damage in the aging process as well as pathogenesis of many diseases. Although no direct comparison of in vitro and in vivo rates can be made, the in vitro rate of H2O2 production in the mitochondria may lend insights to the in vivo potential of oxidative stress (damage). Thus, the maximal rates of H2O2 production under optimal conditions in heart, brain, and kidney mitochondria were compared. Heart and brain mitochondria had the highest rate of H2O2 generation with the supplement of succinate and antimycin, 1161 { 119 and 1382 { 218 (mean { SD) pmol/min/mg protein, respectively. Brain mitochondria also had a similar high rate of generation with the addition of a-glycerophosphate, antimycin, and rotenone (Table IV). Kidney mitochondria had the lowest maximal rate, Ç50% that of heart and brain. The rate of O atom consumption was also measured in mitochondria from these organs under similar assay conditions (Table IV). The rates did not correlate directly with the rate of H2O2 release, implying that varying degrees of free radical leakage occurred in the mitochondria during electron flow. Thus, in vivo the heart and brain mitochondria may
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
123
produce more ROS than the kidney and therefore may be at higher risk of oxidative damage. However, the rate of production may not be related to the rate of oxygen consumption. Furthermore, the maximal rate of H2O2 generation in mouse heart mitochondria isolated from 23-monthold animals was Ç14% higher than that from 4-monthold animals (1393 { 122 pmol/min/mg protein versus 1220 { 111 pmol/min/mg protein; mean { SD; P õ 0.05). Similarly, in a previous study an increase of Ç50% in the rate of succinate-supported heart mitochondrial H2O2 generation was observed in 23-monthold mice when compared to 9-month-old animals (16). This suggests that heart mitochondria from older animals may be at higher risk of oxidative stress than those from younger animals. Mitochondrial state 4 respiration in heart, brain, and kidney. It is well established that mitochondrial O2r0 and H2O2 generation occurs by the autooxidation of certain components of the electron transport chain under condition of state 4 respiration (22). To determine whether a relationship existed between oxygen consumption and H2O2 generation, the rates of state 4 oxygen consumption in heart, brain, and kidney mitochondria were measured with selected substrates. As seen in Table V, succinate-supplemented heart and brain mitochondria consumed oxygen most efficiently. Moreover, varying rates of oxygen consumption were observed in the different substrate and mitochondria combinations. However, when compared to the rates of H2O2 generation (Table I), no clear relationship between state 4 oxygen consumption and H2O2 generation was evident. Superoxide and hydrogen peroxide generation in mouse and rat heart SMPs. The inhibitory effect of rotenone on the mitochondrial rate of succinate-supported H2O2 generation suggests that O2r0 was generated at complex I as well as at complex III. Thus, the rate of succinate-supported generation of O2r0 and H2O2 was studied in the mouse and rat heart SMPs using the inhibitors antimycin and rotenone (Table VI). The ratio of O2r0 to H2O2 was approximately 2 which is in good agreement with the previously reported values (21). Contrary to the results observed with mitochondria, rotenone did not appear to suppress the antimycin-stimulated rate of O2r0 and H2O2 generation in mouse SMPs. A similar finding was obtained with rat SMPs. Thus, the results indicate that the major site of succinate-supported O2r0 and H2O2 generation in mouse and rat heart SMPs is located in the ubiquinone site, consistent with previous findings (3). DISCUSSION
The present study demonstrates that mitochondria isolated from mouse heart, brain, and kidney have se-
arca
124
KWONG AND SOHAL TABLE IV
Rates of Hydrogen Peroxide Release and Oxygen Consumption in Heart, Brain, and Kidney Mitochondria in the Presence of Respiratory Inhibitors
Organ
Substrate (mM)/inhibitor (mM)
Heart Kidney Brain Brain
Succinate (7)/antimycin (0.5) a-Glycerophosphate (7)/antimycin (0.75) Succinate (7)/antimycin (0.75) a-Glycerophosphate (7)/antimycin (0.5)/rotenone (1)
Rate of H2O2 release (pmol/min/mg protein) 1161 618 1382 1470
{ { { {
119 52 218 132
Rate of O consumption (nmol/min/mg protein) 11.4 11.9 11.7 16.9
{ { { {
1.9 3.0 6.6 3.6
Note. Rates of H2O2 release and O atom consumption were measured as described under Materials and Methods. The reaction was initiated with the addition of the specific substrate which was followed by the addition of the inhibitor. Values represent means { SD.
lective substrate and inhibitor preferences for in vitro H2O2 generation and that the apparent sites of H2O2 generation are both substrate and organ specific. Of those used in this study, succinate was the most effective substrate for mitochondrial H2O2 generation in all three organs. Brain and kidney mitochondria also utilized the other substrates, but the rates were much lower than those supported by succinate. However, heart mitochondria appeared to have a high preference for succinate as the substrate. Mitochondria from the different organs also varied in their sensitivity and responsiveness to the respiratory inhibitors rotenone and antimycin. Antimycin had the greatest effect on heart mitochondria, while rotenone had the greatest effect on brain. In addition, our results also suggest that the sites of mitochondrial H2O2 generation are dissimilar in the heart, brain, and kidney; that the relative amounts of H2O2 generated at the different sites are dependent on the substrate used; and that there are varying amounts of free radical leakage during electron flow. The higher rates of H2O2 generation measured in the heart and brain mitochondria when compared to kidney in this study may be due to the interplay of several factors. Superoxide is produced from the autooxidation of respiratory components during electron transport. It is believed that O2r0 is formed on the inner
TABLE V
State 4 Respiration in Mouse Heart, Brain, and Kidney Mitochondria
Substrates
Organ
Succinate
Heart Brain Brain Kidney Kidney Heart
a-Glycerophosphate Pyruvate/malate b-Hydroxybutyrate
AID
ABB 0489
/
6b49$$$281
Rate of O consumption (nmol/min/mg protein) 118 115 27.8 17.1 25.7 60.4
{ { { { { {
6.4 16 8.5 4.4 3.4 8.1
01-02-98 19:25:43
side of the inner mitochondrial membrane because if it was formed on the outer side, it would be eliminated by cytochrome c which is located in the intermembrane space (5). However, in the rat heart mitochondria, the presence of cytochrome c was required for the succinate-supported generation of H2O2 (3). Superoxide is released to the mitochondrial matrix where it is dismutated via superoxide dismutase to form H2O2 . H2O2 is either reduced to water by mitochondrial glutathione peroxidase or diffused out of the mitochondria. Thus, the measured extramitochondrial release of H2O2 is the net balance of these complex reactions as well as the rate of O2r0 generation. Furthermore, the in vitro measured rate of H2O2 generation may not be directly comparable to the in vivo rate. Nonetheless, the present results indicate that the heart and brain may potentially be more at risk for oxidative stress than the kidney.
TABLE VI
Superoxide and Hydrogen Peroxide Generation in Mouse and Rat Heart Submitochondrial Particles Superoxide generation (nmol/min/mg protein) Mouse AA AA / rot Rot Rot / AA Rat AA AA / rot Rot Rot / AA
H2O2 generation (nmol/min/mg protein)
2.96 3.46 0.97 3.43
{ { { {
0.42 0.53 0.13 0.42
1.79 { 0.36 2.13 { 0.13 Not detectable 1.55 { 0.44
3.87 4.28 1.41 4.34
{ { { {
0.27 0.56 0.12 0.56
2.64 { 0.14 2.36 { 0.18 Not detectable 2.68 { 0.24
Note. Values are means { SD. The reactions were essentially as described under Materials and Methods. Succinate was the substrate in both reactions. SOD (300 U/reaction) was included in the H2O2 generation reaction to catalyze the formation of H2O2 from Or0 2 . The inhibitors were added in sequence separated by a 5-min incubation interval.
arca
H2O2 GENERATION BY MITOCHONDRIA
Succinate has been used extensively in studies of H2O2 generation in mammalian mitochondria (9, 23– 26). It is used generally in combination with antimycin (24, 26, 27) and occasionally with rotenone (24) to determine the site(s) of H2O2 generation. It is commonly accepted that antimycin has a stimulatory effect on the rate of generation. The present study of mouse heart, brain, and kidney mitochondria confirms that antimycin does potentiate the rate of succinate-supported H2O2 generation. Furthermore, the extent of potentiation varies in the different organs. Since antimycin is purported to bind to cytochrome b, the magnitude of its potentiation may be related to the cytochrome b content of the mitochondria. An important finding of this study is that rotenone inhibited the rate of succinate-supported H2O2 generation in mouse heart, brain, and kidney mitochondria with or without the prior addition of antimycin. This inhibitory effect of rotenone on succinate-supported H2O2 generation has been reported in rat heart (24) and brain (7) mitochondria. Thus, the present study supports the findings that there may be a backflow of electrons from coenzyme Q to complex I and that succinate-supported mitochondrial H2O2 generation occurs not only at complex III, but also at a site upstream from the rotenone block, complex I. However, contrary to the earlier findings that more H2O2 is generated at complex I than at complex III, the present results show that a quantitatively similar rate of H2O2 generation occurs at both sites. Furthermore, the results from the study on SMPs (Table VI) clearly indicate that much of the O2r0 and H2O2 is formed at complex III, consistent with the earlier finding that the flavin of NADH dehydrogenase in SMPs does not undergo reduction in the presence of succinate (28). Rotenone is believed to bind at the Fe–S center of NADH dehydrogenase (29), but appears to have a dual binding site (30) whose subunit location still has not been established. Its interaction with mitochondria may be substantially different from that with SMPs due to the presence of mitochondrial components absent in SMPs. Thus, it can be concluded that succinate-supported H2O2 generation in the mitochondria occurs at both complex I and complex III. The in vivo significance of this reverse flow of electrons is not known. On the other hand, a-glycerophosphate-supported H2O2 generation occurs mainly at the ubiquinone site. Contrary to other studies (7, 9), the rate of H2O2 generation was measurable in brain and kidney mitochondria in the absence of inhibitor; however, it was extremely low in heart mitochondria. Addition of antimycin greatly stimulated the rates. As seen previously, rotenone also stimulated the a-glycerophosphate-supported rate of H2O2 generation (13). It was postulated that autooxidation of a-glycerophosphate dehydrogenase contributed to H2O2 generation in guinea pig cere-
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
125
bral cortex mitochondria (9). Hence, the observed increase in this study may be due to a rotenone-dependent autooxidation of the flavoprotein dehydrogenase in the a-glycerophosphate shuttle. The results from the sequential addition of antimycin and rotenone study confirm that complex III is the major site of a-glycerophosphate-supported H2O2 generation with minor contribution from the autooxidation of a-glycerophosphate flavoprotein dehydrogenase. No reverse flow of electrons to complex I was evident. Thus, it appears that although the electrons from a-glycerophosphate enter the respiratory chain at the level of CoQ similarly to succinate, the paths must be different such that reverse flow of electrons is unlikely. Another interesting, but puzzling finding of this study was that whereas complex I is the main site for pyruvate/malate-supported H2O2 generation in the mouse heart and brain mitochondria, this is not so for b-hydroxybutyrate, another NADH-linked substrate. Its main site appears to be at complex III, and not at complex I (Table III), since rotenone had a minimal effect on its supported rate of H2O2 generation, while antimycin greatly potentiated it. It seems unlikely that the interaction of rotenone and NADH dehydrogenase with NADH is dependent on the initial substrate. It is possible that a metabolite(s) of b-hydroxybutyrate may enter the electron transport chain via the CoQ site. However, our data do not support this possibility. Thus, further studies are needed to understand this phenomenon. Another NADH substrate, glutamate, proved to be a poor substrate for the study of H2O2 generation at complex I because its metabolite(s) also enters the respiratory chain at the CoQ site via succinate dehydrogenase as evident by malonate inhibition of antimycin-stimulated H2O2 generation. An unanticipated finding was the lack of effect of rotenone on NADH-linked substrate-supported H2O2 generation in the kidney mitochondria. Although rotenone inhibited the pyruvate/malate-supported state 4 oxygen consumption by more than 80% in the kidney mitochondria, a measurable amount of oxygen was still consumed (data not shown). However, no increase in H2O2 generation was evident. As mentioned above, the exact nature of rotenone binding to complex I is not known. It is conceivable that the generation of H2O2 in the mouse kidney mitochondria is dependent on a complex interplay of redox status and membrane potential (24) of the mitochondria as well as the binding of rotenone. Nonetheless, unlike heart and brain mitochondria the main site of pyruvate/malate-supported H2O2 generation in the kidney mitochondria appears to be at complex III. Discrepant results on the effect of senescence on the rate of H2O2 generation have been reported (23, 24). Since the rate of H2O2 generation is dependent on the metabolic status of the mitochondria, different meth-
arca
126
KWONG AND SOHAL
ods for mitochondria isolation may have an effect on the rate of H2O2 generation. Mouse heart mitochondria isolated by the proteolytic enzymatic method (6) had a very low rate of H2O2 generation compared to those isolated by the method used in this study (data not shown). Using antimycin to maximally stimulate the rate of H2O2 generation, a small increase in the rate H2O2 generation was observed in the heart mitochondria from 23-month-old mice compared to those 4 months old. This agrees well with the previous finding of an age-dependent increase in the rates of mitochondrial oxidant generation (16). In conclusion, H2O2 generation in mitochondria isolated from different organs has defined substrate preferences as well as distinct inhibitor responses. The site(s) of generation is both substrate and organ specific. Thus, studies using a single condition to compare mitochondrial H2O2 generation in different organs may be fraught with potential flaws. REFERENCES 1. Turrens, J. F., and Boveris, A. (1980) Biochem. J. 191, 421–427. 2. McCord, J. M., and Fridovich, I. (1969) J. Biol. Chem. 244, 6049– 6055. 3. Turrens, J. F., Alexandre, A., and Lehninger, A. L. (1985) Arch. Biochem. Biophys. 237, 408–414. 4. Cadenas, E., Boveris, A., Ragan, C. E., and Stoppani, A. O. M. (1977) Arch. Biochem. Biophys. 180, 248–257. 5. Turrens, J. F. (1997) Biosci. Rep. 17, 3–8. 6. Hansford, R. G. (1978) Biochem. J. 170, 285–295. 7. Cino, M., and Del Maestro, R. F. (1989) Arch. Biochem. Biophys. 269, 623–638. 8. Sohal, R. S. (1993) Free Radical Biol. Med. 14, 583–588. 9. Zoccarato, F., Cavallini, L., Deana, R., and Alexandre, A. (1988) Biochem. Biophys. Res. Commun. 154, 727–734. 10. Forman, H. J., and Azzi, A. (1997) FASEB J. 11, 374–375. 11. Shigenaga, M. K., Hagen, T. M., and Ames, B. N. (1994) Proc. Natl. Acad. Sci. USA 91, 10771–10778.
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
12. Sohal, R. S., Ku, H. H., Agarwal, S., Forster, M. J., and Lal, H. (1994) Mech. Ageing Dev. 74, 121–133. 13. Sohal, R. S., and Sohal, B. H. (1991) Mech. Ageing Dev. 57, 187– 202. 14. Sims, N. R. (1993) in Methods in Toxicology: Mitochondrial Dysfunction (Lash, L. H., and Jones, D. P., Eds.), Vol. 2, Academic Press, San Diego. 15. Rouslin, W. (1993) in Methods in Toxicology: Mitochondrial Dysfunction (Lash, L. H., and Jones, D. P., Eds.), Vol. 2, pp. 313, Academic Press, San Diego. 16. Sohal, R. S., Agarwal, S., Candas, M., Forster, M. J., and Lal, H. (1994) Mech. Ageing Dev. 76, 215–224. 17. Lash, L. H., and Sall, J. M. (1993) in Methods in Toxicology: Mitochondrial Dysfunction (Lash, L. H., and Jones, D. P., Eds.), Vol. 2, pp. 8–12, Academic Press, San Diego. 18. Hyslop, P. A., and Sklar, A. (1984) Anal. Biochem. 141, 280– 286. 19. Harrington, P. A. (1990) Anal. Biochem. 186, 285–287. 20. Azzi, A., Montecucco, C., and Richter, C. (1975) Biochem. Biophys. Res. Commun. 65, 597–603. 21. Boveris, A., and Cadenas, E. (1980) in Superoxide Dismutase (Oberley, L. W., Ed.), Vol. 2, pp. 16–30, CRC Press, Boca Raton, FL. 22. Chance, B., Sies, H., and Boveris, A. (1979) Physiol. Rev. 59, 527–605. 23. Sohal, R. S., and Dubey, A. (1994) Free. Radical Biol. Med. 16, 621–626. 24. Hansford, R. G., Hogue, B. A., and Mildaziene, V. (1997) J. Bioenerg. Biomembr. 29, 89–95. 25. Muscari, C., Caldarera, C. M., and Guarnieri, C. (1990) Basic Res. Cardiol. 85, 172–178. 26. Gonzalez-Flecha, B., and Boveris, A. (1995) Biochim. Biophys. Acta 1243, 361–366. 27. Patole, M. S., Swaoop, A., and Ramasama, T. (1986) J. Neurochem. 47, 1–8. 28. Schatz, G., and Racker, E. (1966) J. Biol. Chem. 241, 1429– 1438. 29. Turrens, J. F., and McCord, J. M. (1990) in Free Radicals, Lipoproteins, and Membrane Lipids (Paulet, A. C., Douste-Blazy, L., and Paoletti, R., Eds.), pp. 203–212, Plenum Press, New York. 30. Singer, T. P., and Ramsay, R. R. (1994) Biochim. Biophys. Acta 1187, 198–202.
arca
Vol. 350, No. 1, February 1, pp. 118–126, 1998 Article No. BB970489
Substrate and Site Specificity of Hydrogen Peroxide Generation in Mouse Mitochondria1 Linda K. Kwong and Rajindar S. Sohal2 Department of Biological Sciences, Southern Methodist University, Dallas, Teaxs 75275
Received September 15, 1997, and in revised form October 20, 1997
The objective of this study was to elucidate the mechanisms of mitochondrial H2O2 generation in mouse organs by determining the nature of their differences in substrate utilization, inhibitor sensitivity, and the site specificity affecting H2O2 production. Mitochondria were isolated from heart, brain, and kidney and the rate of H2O2 generation was measured using the FADH-linked substrates succinate and a-glycerophosphate as well as the NADH-linked substrates pyruvate/malate, b-hydroxybutyrate, and glutamate. Respiratory inhibitors, antimycin and rotenone, were added singly and sequentially to each substrate-supported H2O2 generation reaction mixture to determine the mitochondrial site(s) of generation and the optimal condition(s) for maximal rates of generation. Succinate supported the highest rate of mitochondrial H2O2 generation. Moreover, it was the preferred substrate for the heart mitochondria. a-Glycerophosphate is a poor substrate for H2O2 generation in heart mitochondria. Inhibitor studies showed that heart mitochondria were the most sensitive and responsive to antimycin, while brain was the most sensitive to rotenone. A surprising finding was that NADH-linked substratesupported H2O2 generation in kidney mitochondria was not responsive to rotenone. The contribution from each of the three sites (ubiquinone, NADH dehydrogenase, and a-glycerophosphate dehydrogenase) of mitochondrial H2O2 generation to the total was both substrate and organ dependent. Results indicate that assay conditions must be considered before comparisons of sites and rates of mitochondrial H2O2 generation among different organs can be made. q 1998 Academic Press
1 This work was supported by Grants R01AG7657 and R01AG13563 from the National Institute of Health/National Institute on Aging. 2 To whom correspondence and reprint requests should be addressed at Department of Biological Sciences, Southern Methodist University, 220 Fondren Science Building, Dallas, TX 75275. Fax: 214/768-3955. E-mail: [email protected].
Key Words: mitochondria; reactive oxygen species; free radicals; H2O2 generation; antimycin; rotenone.
A widely recognized attribute of aerobic metabolism in eukaryotic cells is the production of reactive oxygen species (ROS3) by mitochondria. One-electron autooxidation of reduced mitochondrial respiratory components is believed to lead to production of superoxide anion radical (O2r0) (1) which is dismutated by superoxide dismutase to form hydrogen peroxide (H2O2), a progenitor of the highly reactive hydroxyl free radical, in a stoichiometric ratio of Ç2 (2). Two main sites of O2r0 and H2O2 generation in the inner mitochondrial membrane, namely ubiquinone at complex III (3) and NADH dehydrogenase at complex I (4), have been identified [see review by Turrens (5)]. However, there appears to be variation among different tissues and species in the sites of ROS generation. The major site of succinate-supported H2O2 generation in the bovine heart submitochondrial particles is at complex III, the ubiquinone site (3); however, in the rat heart (6) and brain (7) mitochondria, quantitatively more H2O2 is generated at complex I than at complex III via reverse electron flow. In the housefly mitochondria, NADH dehydrogenase is not a source of H2O2 (8). Furthermore, in guinea pig cerebral cortex mitochondria, a third site of H2O2 generation, autooxidation of a-glycerophosphate dehydrogenase has also been proposed (9). Thus, comparisons of the locations of ROS generation may be biased by experimental conditions such as substrates and inhibitors used as well as by the source of the samples. Recently, the validity of in vivo mitochondrial O2r0 3 Abbreviations used: PHPA, p-hydroxyphenylacetate; HRP, horseradish peroxidase; ROS, reactive oxygen species; Or0 2 , superoxide; AA, antimycin; rot, rotenone; mln, malonate; SMPs, submitochondrial particles; BSA, bovine serum albumin; SOD, superoxide dismutase; CoQ, coenzyme Q.
118
AID
0003-9861/98 $25.00 Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
arca
H2O2 GENERATION BY MITOCHONDRIA
and H2O2 production was questioned (10). The authors suggested that the production of O2r0 was thermodynamically unfavorable and that it is unreasonable to expect an increase in mitochondrial O2r0 and H2O2 production even under most pathological conditions. However, the current view of oxidative stress is that the mitochondria accumulated oxidative damages as the organism aged which resulted in mitochondrial dysfunction (11). An age-dependent increase in the production of mitochondrial O2r0 and H2O2 has been reported (8, 12). In addition, H2O2 generation was found to be increased in the in vitro oxidatively damaged mitochondria (13). Thus, additional information on the mechanism and the sites of H2O2 production may help to clarify this controversy. The objective of the present study is to validate the generation of H2O2 in mitochondria from mouse heart, brain, and kidney and to compare the relative amounts generated at the different mitochondrial sites using various NADH-linked and FADHlinked substrates in conjunction with the respiratory inhibitor rotenone and antimycin. MATERIALS AND METHODS Materials. All chemicals were of analytical grade and purchased from Sigma Chemical Co. (St. Louis, MO). Concentrated stock solutions of antimycin in ethanol and rotenone in chloroform were stored at 0207C until use. Diluted working solutions were made daily in ethanol for use in each experiment. Stock solution of horseradish peroxidase (EC 1.11.1.7, type VI) was made in H2O2 assay buffer at 100 U/ml. Aliquots were stored at 0207C until use. Animals. C57BL/6NNia male mice were obtained from the National Institute on Aging/National Center for Toxicological Research Project on Caloric Restriction Colonies (Jefferson, AR) and were housed individually at the University of North Texas Health Science Center at Fort Worth. The mice were fed ad libitum with NIH-31 diet (Purina Mills, Inc., Richmond, IN) and maintained on a 12-h light, 12-h dark cycle with the light cycle beginning at 06:00. Detailed description of animal husbandry is described elsewhere (12). For this study, 13- to 15-month-old mice were transported a short distance to the Southern Methodist University Vivarium. After an adaptation period of at least 1 week where the mice were maintained under similar conditions as described above, they were killed by carbon dioxide asphyxiation. Brain, heart, and kidney were rapidly removed, rinsed free of blood, and placed in ice-cold mitochondrial isolation buffer. Male Sprague–Dawley rats were bred at the Southern Methodist University Vivarium and maintained under the Federal Guidelines for Ethical Animal Experimentation. Rats were killed and their hearts removed as described above. Isolation of mitochondria and preparation of submitochondrial particles. Procedures for the isolation of mitochondria were conducted at 47C. Mitochondria were isolated from the brain by the method of Sims (14). The brain was finely minced and homogenized in 10 vol of isolation medium containing 320 mM sucrose, 1 mM K2EDTA, and 10 mM Tris–HCl, pH 7.1, using a ground-glass homogenizer. The homogenate was centrifuged at 1330g for 3 min. The pellet was rehomogenized in 5 vol of isolation medium and centrifuged as before. The combined supernatants were centrifuged at 21,200g for 10 min. The pellet was resuspended in isolation medium containing 15% Percoll (10 ml/g of starting material). The suspension was separated on a discontinuous gradient of Percoll: 3 ml 15%, 3.5
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
119
ml 23%, and 3.5 ml 40%. Separation was achieved by centrifugation at 30,700g for 5 min. Mitochondria banded at the interface of 23 and 40% Percoll (third band). The material was removed, diluted 1:4 with isolation medium, and centrifuged at 16,700g for 10 min. The supernatant was removed to within a few millimeters of the loose pellet. The pellet was gently resuspended in the remaining buffer and fatty-acid-free bovine serum albumin (BSA, 10 mg/ml; 0.5 ml/ brain) was added. The mixture was diluted with isolation medium (3 ml/brain) and centrifuged at 6900g for 10 min. The final pellet was rinsed and resuspended in isolation medium using a Teflon-toglass homogenizer. Mitochondria were isolated from the heart using a modified procedure of Rouslin (15). The ventricles were finely minced and homogenized in 10 vol of isolation medium containing 180 mM KCl, 10 mM EGTA–Tris base, pH 7.2, 0.5% fatty-acid-free BSA, and 10 mM Mops–KOH using a Polytron homogenizer (10 s, Brinkmann Instructions, Westbury, NY). The homogenate was centrifuged at 1000g for 10 min. The resulting supernatant was centrifuged at 17,500g for 10 min. The final pellet was rinsed free of any light or loosely packed damaged mitochondria and resuspended in buffer containing 250 mM sucrose, 1 mM EGTA, 10 mM Mops–KOH, pH 7.2. Submitochondrial particles (SMPs) were prepared as described previously (16). Kidney mitochondria were isolated according to the method of Lash and Sall (17). The cortices of the kidneys were minced and homogenized in 10 vol of isolation buffer containing 220 mM sucrose, 5 mM KH2PO4 , 5 mM MgCl2 , 20 mM KCl, 20 mM triethanolamine–HCl, 0.1 mM phenylmethylsulfonyl fluoride, and 2 mM EGTA, pH 7.4, using a Polytron homogenizer. The homogenate was centrifuged at 600g for 10 min. The pellet was resuspended in the original volume of buffer using a ground-glass homogenizer and centrifuged at 600g for 10 min. The combined supernatants were recentrifuged at 15,000g for 5 min. The pellet was carefully rinsed and resuspended in the isolation buffer without EGTA. Measurement of the rates of H2O2 generation and state 4 respiration. The rate of H2O2 released from the mitochondria was measured fluorometrically by the method of Hyslop and Sklar (18). The reaction mixture (3 ml) contained 75–150 mg of mitochondrial protein, 500 mg p-hydroxyphenylacetate (PHPA), 4 U HRP, and 7 mM substrate in 154 mM KCl, 10 mM KH2PO4 , 3 mM MgCl2 , and 0.1 mM EGTA (pH 7.4). The rate of H2O2 released was determined by following the increase in fluorescence at 377C using a Perkin–Elmer LS-5 spectrofluorometer equipped with a thermal-controlled and magnetic stirring sample compartment (Ex: 320 nm; Em: 400 nm). The specific standard curve for each substrate and inhibitor was determined by measuring the fluorescence of known amounts of H2O2 in assay containing the substrate or inhibitor with no mitochondrial protein. The rate of mitochondrial state 4 (substrate in excess but no exogenous ADP) respiration was measured polarographically using a Clark-type electrode and a YSI Model 5300 biological oxygen monitor (YSI Inc., Yellow Springs, OH) at 307C. The incubation mixture consisted of buffer (same as that used for H2O2 generation) and 80–400 mg of mitochondrial protein. The rate of oxygen consumption was measured after the addition of substrate and after the addition of inhibitor(s). Mitochondrial protein concentration was estimated by the method of Harrington (19) using bovine serum albumin as standard. Measurement of Or0 The rate of Or0 2 generation. 2 generation was measured in SMPs as the SOD-inhibitable reduction of acetylated ferricytochrome c (20). The SOD-inhibitable reaction mixture (1 ml) contained 15–30 mg submitochondrial protein, 7.8 mM acetylated ferricytochrome c, 1.2 mM antimycin, 5 mM rotenone, 7 mM succinate, and 100 U SOD in 100 mM potassium phosphate, pH 7.4. The sample mixture was similar to the SOD-inhibitable except no SOD was added and the inhibitors were added as indicated in the text. The reduction of acetylated ferricytochrome c was monitored kinetically at 550 nm and 377C.
arca
120
KWONG AND SOHAL
Effect of respiratory substrates and inhibitors on the quantitation of H2O2 . The milieu of dye-coupled fluorescent reactions may have an effect on the quantitation of the fluorescence. Thus, the effect of different mitochondrial respiratory substrates and inhibitors on PHPA–HRP fluorescent detection of H2O2 was assessed. As shown in Fig. 2, antimycin and rotenone had minimal effect on the fluorescent response of the H2O2 standards. However, succinate slightly lowered the fluorescent response, and pyruvate/malate appeared to significantly quench the fluorescence. Therefore, different H2O2 standard curves were established for each substrate used in this study.
FIG. 1. Catalase inhibition of PHPA–HRP quantitation of hydrogen peroxide generated in the mouse heart mitochondria. H2O2 generation in the mouse heart mitochondria (200 mg) was monitored as described under Materials and Methods. Succinate (7 mM) was added to initiate the reaction. Catalase (200 U) was added as indicated.
Respiratory substrate selectivity of H2O2 generation by mitochondria from different organs. Mitochondria metabolize a variety of respiratory substrates to generate NADH and FADH2 . It is conceivable that in vivo the mitochondria from different organs have selective substrate preference. Therefore, the abilities of various NADH and FADH-linked substrates to support in vitro H2O2 generation in mitochondria isolated from heart, brain, and kidney were compared (Table I). In the mouse heart mitochondria, succinate supported the highest rate of H2O2 generation with a-glycerophosphate supporting the lowest, Ç5% of the rate measured with succinate. The pyruvate/malate-, glutamate-, and b-hydroxybutyrate-supported rates were Ç15–30% that with succinate. In the brain and kidney mitochondria, succinate also supported the highest rate. However, in the presence of the other four substrates brain
RESULTS
Specificity of the dye-coupled fluorescent assay for H2O2 . The rates of extramitochondrial release of H2O2 , i.e., H2O2 generation, are often measured by assays coupled to horseradish peroxidase formation of compound II, which reacts with an electron donor that either gains or loses fluorescent capacity (21). In this study, the rate of H2O2 generation is monitored by the increase in fluorescence due to the accumulation of the oxidized form of PHPA (18). However, other reaction(s) in the mitochondrial sample may also oxidize PHPA and contribute to the measured fluorescence. Therefore, catalase, which catalyzes the reduction of H2O2 without the coupled oxidation of PHPA, is added to the reaction to determine if the increase in fluorescence is H2O2 dependent. As seen in Fig. 1, the rate of succinatesupported increase in fluorescence was lessened with the addition of catalase. Further addition of catalase continued to diminish the rate. Thus, the measured fluorescent response is dependent on the availability of H2O2 .
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
FIG. 2. Effects of respiratory substrates and inhibitors on the PHPA–HRP quantitation of hydrogen peroxide. Reaction condition was as described under Materials and Methods. (j) No addition; (n) 0.5 mM antimycin; (l) 1 mM rotenone; (h) 7 mM succinate; and (.) 5 mM/1 mM pyruvate/malate.
arca
H2O2 GENERATION BY MITOCHONDRIA TABLE I
Substrate Specificity of Hydrogen Release from Mouse Heart, Brain, and Kidney Mitochondria Extramitochondrial H2O2 releasea Substrate Succinate (7 mM) Pyruvate/malate (5 mM/1 mM) a-Glycerophosphate (7 mM) b-Hydroxybutyrate (7 mM) Glutamate (7 mM)
Heart
Brain
Kidney
317 { 43 (n Å 7) 51 { 18 (n Å 6) 18 { 15 (n Å 2) 71 { 34 (n Å 5) 98 { 16 (n Å 4)
387 { 37 (n Å 7) 149 { 22 (n Å 9) 148 { 29 (n Å 5) 118 { 39 (n Å 5) 157 { 42 (n Å 4)
222 { 26 (n Å 6) 171 { 37 (n Å 3) 151 { 20 (n Å 6) 145 { 16 (n Å 2) 180 { 11 (n Å 4)
Note. Extramitochondrial H2O2 release was measured as described under Materials and Methods. The reaction was initiated by the addition of specific substrate. Values represent means { SE. a Rate is expressed as pmol H2O2/mg protein/min.
and kidney mitochondria generated H2O2 at a rate near 50% of that with succinate. Effect of respiratory inhibitors on H2O2 generation in mitochondria from different organs. The rate of H2O2 generation by isolated mitochondria depends on their metabolic state (22). The highest rate was observed when the components of the electron transport in the mitochondria were highly reduced (state 4). Thus, respiratory inhibitors which block electron flow through the mitochondria and produce a full reduction of all components located on the substrate side of the inhibition site would increase the rate H2O2 generation at a site(s) between the substrate and inhibitor. The effects of rotenone, an inhibitor at complex I, and antimycin, an inhibitor at complex III, on the rate of H2O2 generation in mitochondria from heart, brain, and kidney were compared. In the succinate-supported H2O2 generation, heart mitochondria were the most responsive to antimycin (Fig. 3A) and were also the most sensitive. At 0.25 mM antimycin, the rate of H2O2 generation in the heart mitochondria was maximally stimulated, greater than 4-fold that of control. However, at the same concentration of antimycin, the rates in the brain and kidney mitochondria increased approximately 2and 1.4-fold that of control, respectively. The rate in the brain and kidney mitochondria reached maximal at 0.75 mM, approximately 4- and 2-fold that of control, respectively. When rotenone was added to the reaction mixture containing pyruvate/malate, a NADH-linked substrate, the rate of H2O2 generation in brain mitochondria was the most sensitive and responsive (Fig. 3B). At 0.25 mM rotenone, it was maximally stimulated to approximately sevenfold that of control. Although the pyr-
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
121
uvate/malate mixture was a marginal substrate for H2O2 generation in heart mitochondria (see Table I), the addition of rotenone (0.5 mM) maximally stimulated the rate to approximately sixfold that of control. However, in the kidney mitochondria, the rate of H2O2 generation was not stimulated by the addition of rotenone. Furthermore, concentrations above 1 mM rotenone decreased the rate of H2O2 generation by Ç50%. Similar concentration effects of rotenone were observed when another NADH-linked substrate, b-hydroxybutyrate, was used (data not shown). Effects of sequential addition of inhibitors on H2O2 generation. To better understand the characteristics of the two main sites of H2O2 generation in the mammalian mitochondria, rotenone and antimycin were added sequentially to FADH-linked and NADH-linked substrate-supported reactions. FADH-linked substrates succinate and a-glycerophosphate were used to measure H2O2 generation at the ubiquinone site. Antimycin was added to block electron flow between cytochrome b and ubiquinone, while rotenone was added to prevent backflow of electrons to the NADH dehydrogenase site. The addition of antimycin differentially increased the rates of H2O2 generation to 4.0-fold (heart), 3.2-fold (brain), and 1.8-fold (kidney) that of the controls (Table II). Further addition of rotenone decreased the rates, but the levels were still higher than those measured with succinate alone. When rotenone was added before antimycin, the rates decreased to approximately 50% of succinate controls. Subsequent addition of antimycin increased the rates to levels similar to those measured with reversed addition of the inhibitors. These results suggest that with succinate as substrate H2O2 generation occurred at both the ubiquinone site and the NADH dehydrogenase site in the heart, brain, and kidney mitochondria. In a-glycerophosphate-supported H2O2 generation, the addition of antimycin greatly increased the rate of generation in mitochondria from all three organs (Table II). Sequential addition of rotenone to the antimycin-supplemented reaction further increased the rates slightly. On the other hand, incubation with rotenone only slightly stimulated the rates above those of substrate control, but further addition of antimycin greatly increased the rates. This suggests that for this substrate the majority of H2O2 was produced at the ubiquinone site. Thus, the main site of a-glycerophosphate-supported H2O2 generation in the mouse heart, brain, and kidney mitochondria was different from that of succinate. NADH-linked substrates pyruvate/malate, b-hydroxybutyrate, and glutamate were used to measure H2O2 generation at both the NADH dehydrogenase site and the ubiquinone site. Again, rotenone and antimycin were added singly and sequentially to estimate the pro-
arca
122
KWONG AND SOHAL
FIG. 3. Effect of respiratory inhibitors on hydrogen peroxide generation. Reaction condition was as described under Materials and Methods. (A) Effect of antimycin on succinate (7 mM)-supported H2O2 generation in the heart (j), brain (h), and (.) kidney mitochondria. (B) Effect of rotenone on pyruvate/malate (5 mM/1 mM)-supported H2O2 generation in the heart (j), brain (h), and (.) kidney mitochondria.
portion of H2O2 generated at each of the sites. As shown in Table III, rotenone when added to pyruvate/malatesupported H2O2 generation increased the rate in the heart and brain mitochondria approximately six- and eightfold that of controls, respectively. Further addition of antimycin did not change the rate substantially. Addition of antimycin before rotenone also greatly increased the rate. Subsequent addition of rotenone had a slight stimulatory effect on the rate in the brain mitochondria. In contrast, it had a slight inhibitory effect on the rate in the heart mitochondria. Surprisingly, in the kidney mitochondria rotenone had little effect on H2O2 generation, while antimycin had a stimulatory effect. However, the order of addition appeared to have an influence on H2O2 generation. Rotenone obliterated the stimulatory effect of antimycin, while antimycin increased the rate of rotenone-supplemented H2O2 generation. In addition, the observed antimycin-stimulated increases in the mitochondria of all organs were not sensitive to malonate, a potent competitive inhibitor of succinate dehydrogenase. This suggests that succinate dehydrogenase was not involved in pyruvate/ malate-supported H2O2 generation. In the b-hydroxybutyrate-supported H2O2 generation, rotenone did not appear to have an effect on heart mitochondria, but had a small stimulatory effect on brain and kidney mitochondria. Furthermore, subsequent addition of antimycin decreased the rate in heart and brain mitochondria, but not in kidney. However, addition of antimycin before rotenone greatly increased the rates in the mitochondria of all three organs and these increases were not sensitive to malonate, but were to rotenone. Similar results were obtained from
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
the glutamate-supported system with the inhibitors except that the antimycin-stimulated increases were sensitive to malonate as well as to rotenone. Maximum potential rate of H2O2 generation in heart, brain, and kidney mitochondria. The generation of
TABLE II
Effects of Rotenone and Antimycin on FADH-Linked Substrate-Supported Hydrogen Peroxide Release from Heart, Brain, and Kidney Mitochondria Relative extramitochondrial H2O2 release Substrates Succinate Succinate / AA Succinate / AA / rot Succinate / rot Succinate / rot / AA a-Glycerophosphate a-Glycerophosphate / a-Glycerophosphate / a-Glycerophosphate / a-Glycerophosphate /
AA AA / rot rot rot / AA
Heart
Brain
Kidney
1 4.0 1.9 0.8 2.0 1 4.1 5.1 2.8 5.5
1 3.2 1.4 0.5 1.9 1 7.2 7.6 1.2 8.8
1 1.8 1.3 0.4 1.2 1 5.9 6.7 1.2 5.2
Note. Relative extramitochondrial H2O2 release is expressed as a ratio of the rate measured with inhibitor(s) to that of substrate alone. The initial rate of H2O2 generation in the mitochondrial reaction mixture was measured in the presence of substrate, and then the inhibitor(s) was added in the order indicated and the resulting rate was measured. Concentrations of substrates and inhibitors were as followed: 7 mM succinate, 7 mM a-glycerophosphate, 0.25 mM (heart) and 0.75 mM (brain and kidney) antimycin (AA), and 0.5 mM (heart and kidney) and 1 mM (brain) rotenone (rot).
arca
H2O2 GENERATION BY MITOCHONDRIA TABLE III
Effects of Rotenone and Antimycin on NADH-Linked Substrate-Supported Hydrogen Peroxide Release from Heart, Brain, and Kidney Mitochondria Relative extramitochondrial H2O2 release Substrates
Heart
Brain
Kidney
Pyruvate/malate Pyruvate/malate / rot Pyruvate/malate / rot / AA Pyruvate/malate / AA Pyruvate/malate / AA / rot Pyruvate/malate / AA / mln b-Hydroxybutyrate b-Hydroxybutyrate / rot b-Hydroxybutyrate / rot / AA b-Hydroxybutyrate / AA b-Hydroxybutyrate / AA / rot b-Hydroxybutyrate / AA / mln Glutamate Glutamate / rot Glutamate / rot / AA Glutamate / AA Glutamate / AA / rot Glutamate / AA / mln
1 6.2 6.6 11.0 8.5 11.0 1 1.0 0.3 6.6 1 5.4 1 1.2 1.7 4.8 1.6 1.4
1 8.0 7.5 5.4 6.7 5.4 1 2.2 1.1 5.6 2.2 5.1 1 1.0 0.6 6.7 1.0 3.1
1 0.9 3.3 2.5 .8 2.5 1 1.4 2.4 2.8 1.3 2.8 1 0.7 0.4 3.8 0.8 2.2
Note. Relative extramitochondrial H2O2 release was measured as described in the note to Table 2. Concentrations of substrates and inhibitors were as followed: 5 mM/1 mM pyruvate/malate, 7 mM bhydroxybutyrate, 7 mM glutamate, 0.5 mM (brain and kidney) and 0.75 mM (heart) antimycin (AA), 1 mM rotenone (rot), and 5 mM malonate (mln).
H2O2 by mitochondria is considered to have much physiological significance. It is thought to be associated with oxidative damage in the aging process as well as pathogenesis of many diseases. Although no direct comparison of in vitro and in vivo rates can be made, the in vitro rate of H2O2 production in the mitochondria may lend insights to the in vivo potential of oxidative stress (damage). Thus, the maximal rates of H2O2 production under optimal conditions in heart, brain, and kidney mitochondria were compared. Heart and brain mitochondria had the highest rate of H2O2 generation with the supplement of succinate and antimycin, 1161 { 119 and 1382 { 218 (mean { SD) pmol/min/mg protein, respectively. Brain mitochondria also had a similar high rate of generation with the addition of a-glycerophosphate, antimycin, and rotenone (Table IV). Kidney mitochondria had the lowest maximal rate, Ç50% that of heart and brain. The rate of O atom consumption was also measured in mitochondria from these organs under similar assay conditions (Table IV). The rates did not correlate directly with the rate of H2O2 release, implying that varying degrees of free radical leakage occurred in the mitochondria during electron flow. Thus, in vivo the heart and brain mitochondria may
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
123
produce more ROS than the kidney and therefore may be at higher risk of oxidative damage. However, the rate of production may not be related to the rate of oxygen consumption. Furthermore, the maximal rate of H2O2 generation in mouse heart mitochondria isolated from 23-monthold animals was Ç14% higher than that from 4-monthold animals (1393 { 122 pmol/min/mg protein versus 1220 { 111 pmol/min/mg protein; mean { SD; P õ 0.05). Similarly, in a previous study an increase of Ç50% in the rate of succinate-supported heart mitochondrial H2O2 generation was observed in 23-monthold mice when compared to 9-month-old animals (16). This suggests that heart mitochondria from older animals may be at higher risk of oxidative stress than those from younger animals. Mitochondrial state 4 respiration in heart, brain, and kidney. It is well established that mitochondrial O2r0 and H2O2 generation occurs by the autooxidation of certain components of the electron transport chain under condition of state 4 respiration (22). To determine whether a relationship existed between oxygen consumption and H2O2 generation, the rates of state 4 oxygen consumption in heart, brain, and kidney mitochondria were measured with selected substrates. As seen in Table V, succinate-supplemented heart and brain mitochondria consumed oxygen most efficiently. Moreover, varying rates of oxygen consumption were observed in the different substrate and mitochondria combinations. However, when compared to the rates of H2O2 generation (Table I), no clear relationship between state 4 oxygen consumption and H2O2 generation was evident. Superoxide and hydrogen peroxide generation in mouse and rat heart SMPs. The inhibitory effect of rotenone on the mitochondrial rate of succinate-supported H2O2 generation suggests that O2r0 was generated at complex I as well as at complex III. Thus, the rate of succinate-supported generation of O2r0 and H2O2 was studied in the mouse and rat heart SMPs using the inhibitors antimycin and rotenone (Table VI). The ratio of O2r0 to H2O2 was approximately 2 which is in good agreement with the previously reported values (21). Contrary to the results observed with mitochondria, rotenone did not appear to suppress the antimycin-stimulated rate of O2r0 and H2O2 generation in mouse SMPs. A similar finding was obtained with rat SMPs. Thus, the results indicate that the major site of succinate-supported O2r0 and H2O2 generation in mouse and rat heart SMPs is located in the ubiquinone site, consistent with previous findings (3). DISCUSSION
The present study demonstrates that mitochondria isolated from mouse heart, brain, and kidney have se-
arca
124
KWONG AND SOHAL TABLE IV
Rates of Hydrogen Peroxide Release and Oxygen Consumption in Heart, Brain, and Kidney Mitochondria in the Presence of Respiratory Inhibitors
Organ
Substrate (mM)/inhibitor (mM)
Heart Kidney Brain Brain
Succinate (7)/antimycin (0.5) a-Glycerophosphate (7)/antimycin (0.75) Succinate (7)/antimycin (0.75) a-Glycerophosphate (7)/antimycin (0.5)/rotenone (1)
Rate of H2O2 release (pmol/min/mg protein) 1161 618 1382 1470
{ { { {
119 52 218 132
Rate of O consumption (nmol/min/mg protein) 11.4 11.9 11.7 16.9
{ { { {
1.9 3.0 6.6 3.6
Note. Rates of H2O2 release and O atom consumption were measured as described under Materials and Methods. The reaction was initiated with the addition of the specific substrate which was followed by the addition of the inhibitor. Values represent means { SD.
lective substrate and inhibitor preferences for in vitro H2O2 generation and that the apparent sites of H2O2 generation are both substrate and organ specific. Of those used in this study, succinate was the most effective substrate for mitochondrial H2O2 generation in all three organs. Brain and kidney mitochondria also utilized the other substrates, but the rates were much lower than those supported by succinate. However, heart mitochondria appeared to have a high preference for succinate as the substrate. Mitochondria from the different organs also varied in their sensitivity and responsiveness to the respiratory inhibitors rotenone and antimycin. Antimycin had the greatest effect on heart mitochondria, while rotenone had the greatest effect on brain. In addition, our results also suggest that the sites of mitochondrial H2O2 generation are dissimilar in the heart, brain, and kidney; that the relative amounts of H2O2 generated at the different sites are dependent on the substrate used; and that there are varying amounts of free radical leakage during electron flow. The higher rates of H2O2 generation measured in the heart and brain mitochondria when compared to kidney in this study may be due to the interplay of several factors. Superoxide is produced from the autooxidation of respiratory components during electron transport. It is believed that O2r0 is formed on the inner
TABLE V
State 4 Respiration in Mouse Heart, Brain, and Kidney Mitochondria
Substrates
Organ
Succinate
Heart Brain Brain Kidney Kidney Heart
a-Glycerophosphate Pyruvate/malate b-Hydroxybutyrate
AID
ABB 0489
/
6b49$$$281
Rate of O consumption (nmol/min/mg protein) 118 115 27.8 17.1 25.7 60.4
{ { { { { {
6.4 16 8.5 4.4 3.4 8.1
01-02-98 19:25:43
side of the inner mitochondrial membrane because if it was formed on the outer side, it would be eliminated by cytochrome c which is located in the intermembrane space (5). However, in the rat heart mitochondria, the presence of cytochrome c was required for the succinate-supported generation of H2O2 (3). Superoxide is released to the mitochondrial matrix where it is dismutated via superoxide dismutase to form H2O2 . H2O2 is either reduced to water by mitochondrial glutathione peroxidase or diffused out of the mitochondria. Thus, the measured extramitochondrial release of H2O2 is the net balance of these complex reactions as well as the rate of O2r0 generation. Furthermore, the in vitro measured rate of H2O2 generation may not be directly comparable to the in vivo rate. Nonetheless, the present results indicate that the heart and brain may potentially be more at risk for oxidative stress than the kidney.
TABLE VI
Superoxide and Hydrogen Peroxide Generation in Mouse and Rat Heart Submitochondrial Particles Superoxide generation (nmol/min/mg protein) Mouse AA AA / rot Rot Rot / AA Rat AA AA / rot Rot Rot / AA
H2O2 generation (nmol/min/mg protein)
2.96 3.46 0.97 3.43
{ { { {
0.42 0.53 0.13 0.42
1.79 { 0.36 2.13 { 0.13 Not detectable 1.55 { 0.44
3.87 4.28 1.41 4.34
{ { { {
0.27 0.56 0.12 0.56
2.64 { 0.14 2.36 { 0.18 Not detectable 2.68 { 0.24
Note. Values are means { SD. The reactions were essentially as described under Materials and Methods. Succinate was the substrate in both reactions. SOD (300 U/reaction) was included in the H2O2 generation reaction to catalyze the formation of H2O2 from Or0 2 . The inhibitors were added in sequence separated by a 5-min incubation interval.
arca
H2O2 GENERATION BY MITOCHONDRIA
Succinate has been used extensively in studies of H2O2 generation in mammalian mitochondria (9, 23– 26). It is used generally in combination with antimycin (24, 26, 27) and occasionally with rotenone (24) to determine the site(s) of H2O2 generation. It is commonly accepted that antimycin has a stimulatory effect on the rate of generation. The present study of mouse heart, brain, and kidney mitochondria confirms that antimycin does potentiate the rate of succinate-supported H2O2 generation. Furthermore, the extent of potentiation varies in the different organs. Since antimycin is purported to bind to cytochrome b, the magnitude of its potentiation may be related to the cytochrome b content of the mitochondria. An important finding of this study is that rotenone inhibited the rate of succinate-supported H2O2 generation in mouse heart, brain, and kidney mitochondria with or without the prior addition of antimycin. This inhibitory effect of rotenone on succinate-supported H2O2 generation has been reported in rat heart (24) and brain (7) mitochondria. Thus, the present study supports the findings that there may be a backflow of electrons from coenzyme Q to complex I and that succinate-supported mitochondrial H2O2 generation occurs not only at complex III, but also at a site upstream from the rotenone block, complex I. However, contrary to the earlier findings that more H2O2 is generated at complex I than at complex III, the present results show that a quantitatively similar rate of H2O2 generation occurs at both sites. Furthermore, the results from the study on SMPs (Table VI) clearly indicate that much of the O2r0 and H2O2 is formed at complex III, consistent with the earlier finding that the flavin of NADH dehydrogenase in SMPs does not undergo reduction in the presence of succinate (28). Rotenone is believed to bind at the Fe–S center of NADH dehydrogenase (29), but appears to have a dual binding site (30) whose subunit location still has not been established. Its interaction with mitochondria may be substantially different from that with SMPs due to the presence of mitochondrial components absent in SMPs. Thus, it can be concluded that succinate-supported H2O2 generation in the mitochondria occurs at both complex I and complex III. The in vivo significance of this reverse flow of electrons is not known. On the other hand, a-glycerophosphate-supported H2O2 generation occurs mainly at the ubiquinone site. Contrary to other studies (7, 9), the rate of H2O2 generation was measurable in brain and kidney mitochondria in the absence of inhibitor; however, it was extremely low in heart mitochondria. Addition of antimycin greatly stimulated the rates. As seen previously, rotenone also stimulated the a-glycerophosphate-supported rate of H2O2 generation (13). It was postulated that autooxidation of a-glycerophosphate dehydrogenase contributed to H2O2 generation in guinea pig cere-
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
125
bral cortex mitochondria (9). Hence, the observed increase in this study may be due to a rotenone-dependent autooxidation of the flavoprotein dehydrogenase in the a-glycerophosphate shuttle. The results from the sequential addition of antimycin and rotenone study confirm that complex III is the major site of a-glycerophosphate-supported H2O2 generation with minor contribution from the autooxidation of a-glycerophosphate flavoprotein dehydrogenase. No reverse flow of electrons to complex I was evident. Thus, it appears that although the electrons from a-glycerophosphate enter the respiratory chain at the level of CoQ similarly to succinate, the paths must be different such that reverse flow of electrons is unlikely. Another interesting, but puzzling finding of this study was that whereas complex I is the main site for pyruvate/malate-supported H2O2 generation in the mouse heart and brain mitochondria, this is not so for b-hydroxybutyrate, another NADH-linked substrate. Its main site appears to be at complex III, and not at complex I (Table III), since rotenone had a minimal effect on its supported rate of H2O2 generation, while antimycin greatly potentiated it. It seems unlikely that the interaction of rotenone and NADH dehydrogenase with NADH is dependent on the initial substrate. It is possible that a metabolite(s) of b-hydroxybutyrate may enter the electron transport chain via the CoQ site. However, our data do not support this possibility. Thus, further studies are needed to understand this phenomenon. Another NADH substrate, glutamate, proved to be a poor substrate for the study of H2O2 generation at complex I because its metabolite(s) also enters the respiratory chain at the CoQ site via succinate dehydrogenase as evident by malonate inhibition of antimycin-stimulated H2O2 generation. An unanticipated finding was the lack of effect of rotenone on NADH-linked substrate-supported H2O2 generation in the kidney mitochondria. Although rotenone inhibited the pyruvate/malate-supported state 4 oxygen consumption by more than 80% in the kidney mitochondria, a measurable amount of oxygen was still consumed (data not shown). However, no increase in H2O2 generation was evident. As mentioned above, the exact nature of rotenone binding to complex I is not known. It is conceivable that the generation of H2O2 in the mouse kidney mitochondria is dependent on a complex interplay of redox status and membrane potential (24) of the mitochondria as well as the binding of rotenone. Nonetheless, unlike heart and brain mitochondria the main site of pyruvate/malate-supported H2O2 generation in the kidney mitochondria appears to be at complex III. Discrepant results on the effect of senescence on the rate of H2O2 generation have been reported (23, 24). Since the rate of H2O2 generation is dependent on the metabolic status of the mitochondria, different meth-
arca
126
KWONG AND SOHAL
ods for mitochondria isolation may have an effect on the rate of H2O2 generation. Mouse heart mitochondria isolated by the proteolytic enzymatic method (6) had a very low rate of H2O2 generation compared to those isolated by the method used in this study (data not shown). Using antimycin to maximally stimulate the rate of H2O2 generation, a small increase in the rate H2O2 generation was observed in the heart mitochondria from 23-month-old mice compared to those 4 months old. This agrees well with the previous finding of an age-dependent increase in the rates of mitochondrial oxidant generation (16). In conclusion, H2O2 generation in mitochondria isolated from different organs has defined substrate preferences as well as distinct inhibitor responses. The site(s) of generation is both substrate and organ specific. Thus, studies using a single condition to compare mitochondrial H2O2 generation in different organs may be fraught with potential flaws. REFERENCES 1. Turrens, J. F., and Boveris, A. (1980) Biochem. J. 191, 421–427. 2. McCord, J. M., and Fridovich, I. (1969) J. Biol. Chem. 244, 6049– 6055. 3. Turrens, J. F., Alexandre, A., and Lehninger, A. L. (1985) Arch. Biochem. Biophys. 237, 408–414. 4. Cadenas, E., Boveris, A., Ragan, C. E., and Stoppani, A. O. M. (1977) Arch. Biochem. Biophys. 180, 248–257. 5. Turrens, J. F. (1997) Biosci. Rep. 17, 3–8. 6. Hansford, R. G. (1978) Biochem. J. 170, 285–295. 7. Cino, M., and Del Maestro, R. F. (1989) Arch. Biochem. Biophys. 269, 623–638. 8. Sohal, R. S. (1993) Free Radical Biol. Med. 14, 583–588. 9. Zoccarato, F., Cavallini, L., Deana, R., and Alexandre, A. (1988) Biochem. Biophys. Res. Commun. 154, 727–734. 10. Forman, H. J., and Azzi, A. (1997) FASEB J. 11, 374–375. 11. Shigenaga, M. K., Hagen, T. M., and Ames, B. N. (1994) Proc. Natl. Acad. Sci. USA 91, 10771–10778.
AID
ABB 0489
/
6b49$$$281
01-02-98 19:25:43
12. Sohal, R. S., Ku, H. H., Agarwal, S., Forster, M. J., and Lal, H. (1994) Mech. Ageing Dev. 74, 121–133. 13. Sohal, R. S., and Sohal, B. H. (1991) Mech. Ageing Dev. 57, 187– 202. 14. Sims, N. R. (1993) in Methods in Toxicology: Mitochondrial Dysfunction (Lash, L. H., and Jones, D. P., Eds.), Vol. 2, Academic Press, San Diego. 15. Rouslin, W. (1993) in Methods in Toxicology: Mitochondrial Dysfunction (Lash, L. H., and Jones, D. P., Eds.), Vol. 2, pp. 313, Academic Press, San Diego. 16. Sohal, R. S., Agarwal, S., Candas, M., Forster, M. J., and Lal, H. (1994) Mech. Ageing Dev. 76, 215–224. 17. Lash, L. H., and Sall, J. M. (1993) in Methods in Toxicology: Mitochondrial Dysfunction (Lash, L. H., and Jones, D. P., Eds.), Vol. 2, pp. 8–12, Academic Press, San Diego. 18. Hyslop, P. A., and Sklar, A. (1984) Anal. Biochem. 141, 280– 286. 19. Harrington, P. A. (1990) Anal. Biochem. 186, 285–287. 20. Azzi, A., Montecucco, C., and Richter, C. (1975) Biochem. Biophys. Res. Commun. 65, 597–603. 21. Boveris, A., and Cadenas, E. (1980) in Superoxide Dismutase (Oberley, L. W., Ed.), Vol. 2, pp. 16–30, CRC Press, Boca Raton, FL. 22. Chance, B., Sies, H., and Boveris, A. (1979) Physiol. Rev. 59, 527–605. 23. Sohal, R. S., and Dubey, A. (1994) Free. Radical Biol. Med. 16, 621–626. 24. Hansford, R. G., Hogue, B. A., and Mildaziene, V. (1997) J. Bioenerg. Biomembr. 29, 89–95. 25. Muscari, C., Caldarera, C. M., and Guarnieri, C. (1990) Basic Res. Cardiol. 85, 172–178. 26. Gonzalez-Flecha, B., and Boveris, A. (1995) Biochim. Biophys. Acta 1243, 361–366. 27. Patole, M. S., Swaoop, A., and Ramasama, T. (1986) J. Neurochem. 47, 1–8. 28. Schatz, G., and Racker, E. (1966) J. Biol. Chem. 241, 1429– 1438. 29. Turrens, J. F., and McCord, J. M. (1990) in Free Radicals, Lipoproteins, and Membrane Lipids (Paulet, A. C., Douste-Blazy, L., and Paoletti, R., Eds.), pp. 203–212, Plenum Press, New York. 30. Singer, T. P., and Ramsay, R. R. (1994) Biochim. Biophys. Acta 1187, 198–202.
arca