Mitochondrial Ubiquinone Homologues, Superoxide Radical Generation, and Longevity in Different Mammalian Species*

Rates of mitochondrial superoxide anion radical (O·̄2) generation are known to be inversely correlated with the maximum life span potential of different mammalian species. The objective of this study was to understand the possible mechanism(s) underlying such variations in the rate of O·̄2 generation. The hypothesis that the relative amounts of the ubiquinones or coenzyme Q (CoQ) homologues, CoQ9 and CoQ10, are related with the rate of O·̄2 generation was tested. A comparison of nine different mammalian species, namely mouse, rat, guinea pig, rabbit, pig, goat, sheep, cow, and horse, which vary from 3.5 to 46 years in their maximum longevity, indicated that the rate of O·̄2 generation in cardiac submitochondrial particles (SMPs) was directly related to the relative amount of CoQ9 and inversely related to the amount of CoQ10, extractable from their cardiac mitochondria. To directly test the relationship between CoQ homologues and the rate of O·̄2 generation, rat heart SMPs, naturally containing mainly CoQ9 and cow heart SMPs, with high natural CoQ10 content, were chosen for depletion/reconstitution experiments. Repeated extractions of rat heart SMPs with pentane exponentially depleted both CoQ homologues while the corresponding rates of O·̄2 generation and oxygen consumption were lowered linearly. Reconstitution of both rat and cow heart SMPs with different amounts of CoQ9 or CoQ10 caused an initial increase in the rates of O·̄2 generation, followed by a plateau at high concentrations. Within the physiological range of CoQ concentrations, there were no differences in the rates of O·̄2generation between SMPs reconstituted with CoQ9 or CoQ10. Only at concentrations that were considerably higher than the physiological level, the SMPs reconstituted with CoQ9 exhibited higher rates of O·̄2 generation than those obtained with CoQ10. These in vitrofindings do not support the hypothesis that differences in the distribution of CoQ homologues are responsible for the variations in the rates of mitochondrial O·̄2 generation in different mammalian species.

with CoQ 10 . These in vitro findings do not support the hypothesis that differences in the distribution of CoQ homologues are responsible for the variations in the rates of mitochondrial O 2 . generation in different mammalian species.
A current hypothesis of aging postulates that oxidative stress/damage is a major causal factor in the attrition of functional capacity occurring during the aging process (1)(2)(3)(4)(5)(6). The basic tenet of this hypothesis is that there is an intrinsic imbalance between the reactive oxygen species (ROS), 1 that are incessantly generated in the aerobic cells and the antioxidative defense against them, thereby resulting in the accrual of steady-state levels of oxidative molecular damage. The direct evidence in support of this hypothesis is that the augmentation of antioxidative defenses by simultaneous overexpression of Cu/Zn superoxide dismutase, which converts superoxide anion radicals (O 2 . ) into H 2 O 2 , and catalase, which removes H 2 O 2 , retards the age-associated increase in the levels of molecular oxidative damage and extends the life span of Drosophila melanogaster by one-third (7,8).
Although there are several intracellular loci for the generation of O 2 . (the first molecule in the ROS series), it is widely accepted that the mitochondrial electron transport chain is the main source of O 2 . (9,10). Previous studies in this laboratory have indicated that the rate of mitochondrial O 2 . generation varies greatly, even in the same type of tissue, among different mammalian species and is inversely related to the maximum life span potential (MLSP) of the species (11,12). The inverse relationship between the rate of O 2 . generation and MLSP was found to hold in a sample of mammalian species as well as a group of dipteran insect species (11)(12)(13).
The question that arose out of these studies and that is also the subject of this investigation is what is the underlying mechanism for the variations in the rates of mitochondrial O 2 . generation in different species? Although opinions vary (14), a number of experimental studies in the literature suggest that ubiquinones modulate the rate of mitochondrial O 2 . /H 2 O 2 generation (10,(15)(16)(17)(18). Ubiquinones (2,3-dimethoxy-5-methyl-6multiprenyl-1,4-benzoquinone), or coenzyme Q (CoQ), is a quinone derivative with a chain of 1-12 isoprene units in the different homologue forms (CoQ n ) occurring in nature. Relatively short-lived mammalian species such as the mouse and the rat primarily contain CoQ 9 , whereas the larger long-lived mammals such as man predominantly exhibit CoQ 10 (19). The present study tests the hypothesis that variations in the rate of O 2 . by cardiac submitochondrial particles (SMPs) in different mammalian species are related to the relative CoQ 9 and/or CoQ 10 content. The hypothesis was prompted by the fact that longevity of non-primate mammalian species tends to be inversely correlated with the rate of mitochondrial O 2 . generation and directly correlated with the body mass.
Animals-Hearts were obtained from mouse (Swiss), rat (Harlan Sprague Dawley), guinea pig (Hartley Albino), rabbit (New Zealand White), pig (Yorkshire), goat (Angora), sheep (Rambouillet), cow (Holstein), and horse (mixed) which range from 3.5 to 46 years in MLSP (21,29). All the animals were young, healthy, sexually mature adult males. The approximate ages of the animals were: mouse, rat, and guinea pig, 4 months; rabbit, 7 months; pig, 6 -7 months; goat and sheep, 1 year; cow and horse, 3 years. In smaller animals the entire heart was processed; however, in the pig, cow, and horse the hearts were cut into smaller pieces, and representative samples were selected. The values for the rates of O 2 . generation in different species are partially based on the result of previous studies in this laboratory (12,22). MLSP values for different species, obtained from the literature (20,21), in years, are: mouse, 3.5; rat, 4.5; guinea pig, 7.5; rabbit 13; goat, 18; sheep, 20; pig, 27; cow, 30; horse, 46. Isolation of Mitochondria and Preparation of SMPs-Mitochondria were isolated by differential centrifugation as described by Arcos et al. (23). Briefly, pieces of the heart were homogenized in 10 volumes (w/v) of isolation buffer containing 180 mM KCl, 0.5% bovine serum albumin, 10 mM MOPS, 10 mM EGTA Tris base, pH 7.2, at 4°C. The homogenate was centrifuged at 1,000 ϫ g for 10 min, and the supernatant was recentrifuged at 17,500 ϫ g. The resulting mitochondrial pellet was washed and resuspended in 0.25 M sucrose, 1 mM EGTA, 10 mM MOPS, pH 7.2. To prepare SMPs, the mitochondrial pellet was resuspended in 30 mM potassium phosphate buffer, pH 7.0, and sonicated three times, each consisting of a 30-s pulse burst, at 1-min intervals at 4°C. The sonicated mitochondria were centrifuged at 8,250 ϫ g for 10 min to remove the unbroken organelles; the supernatant was recentrifuged at 80,000 ϫ g for 45 min, and the resulting pellet was washed and resuspended in 0.1 M phosphate buffer, pH 7.4, as described previously (12).
Extraction and Quantitation of Coenzyme Q-CoQ was extracted from mitochondria using a hexane:ethanol mixture as described by Takada et al. (24). Briefly, 50 l of mitochondrial suspension, containing ϳ100 g of protein and 50 l of double-distilled H 2 O were mixed with 750 l of hexane:ethanol (5:2) for 1 min using a vortex mixture. The mixture was centrifuged for 3 min at 1,200 ϫ g, and 450 l of the hexane layer was collected, dried under helium, and dissolved in 100 l of ethanol. Quantitation of ubiquinones was performed by HPLC by the method of Katayama et al. (25). The ethanol extract (10 -20 l) was chromatographed on a reverse phase C 18 HPLC column (25.0 ϫ 0.46 cm, 5 m, Supelco), using a mobile phase consisting of 0.7% NaClO 4 in ethanol:methanol:70% HClO 4 (900:100:1) at a flow rate of 1.2 ml. The electrochemical and UV detectors consisted of an ESA Coulochem II and a Waters Associates Model 440 absorbance detector at a wavelength of 280 nm. The setting of the electrochemical detector was as follows: guard cell (upstream of the injector) at ϩ200 mV, conditioning cell at Ϫ550 mV (downstream of the column), followed by the analytical cell at ϩ150 mV. The concentrations of ubiquinones were estimated by comparison of the peak area with those of standard solutions of known concentration.
Coenzyme Q Depletion and Repletion-Submitochondrial particles were depleted of native CoQ by pentane extraction, as described by Maguire et al. (26) and selectively repleted with exogenous CoQ 9 or CoQ 10 . Aliquots of SMPs (100 l containing ϳ250 g of protein) were freeze-dried and extracted 6 times, each for 45 min at 4°C, with 1 ml of pentane containing 15 M ␣-tocopherol (for the last extraction pure pentane was used). The pentane layer was removed by centrifugation and discarded. In some cases the pentane layer was collected and brought to dryness under a stream of helium and resuspended in 100 l of ethanol, and the CoQ content was measured by HPLC. Various amounts of CoQ 9 or CoQ 10 were added to pentane-extracted and/or freeze-dried SMPs, dried under helium, resuspended in 100 mM potassium phosphate buffer (pH 7.4), and sonicated for up to 3 s in a Branson 2200 sonicator. CoQ depletion and incorporation of exogenous CoQ 9 or CoQ 10 into SMPs membranes were confirmed by hexane:ethanol extraction and HPLC analysis, as described above.

Measurement of Superoxide Anion Radical Generation
. generation by SMPs was measured as superoxide dismutase-inhibitable reduction of acetylated ferricytochrome c (27), as described previously (12). The reaction mixture contained 10 M acetylated ferricytochrome c, 6 M rotenone, 1.2 M antimycin A, 100 units of superoxide dismutase/ml (in the reference cuvette), and 10 -100 g of SMP protein in 100 mM potassium phosphate buffer, pH 7.4. The reaction was started by adding 7.5 mM succinate, and the reduction of acetylated ferricytochrome c was followed at 550 nm.

Measurement of Oxygen Consumption-
The rate of respiration of submitochondrial particles was measured polarographically using a Clark-type electrode at 37°C. The incubation mixture, to measure state 4 respiration, consisted of buffer (154 mM KCl, 3 mM MgCl 2 , 10 mM KPO 4 , 0.1 mM EGTA, pH 7.4) and 30 -100 g of SMP protein; 7 mM succinate and/or 7 mM NADH were used as substrates, and 1.2 M antimycin A, 6 M rotenone, or 0.5 mM TTFA were employed as specific respiratory inhibitors.

Variations in the Distribution of CoQ Homologues in Mitochondria of Different
Species-Comparisons of the concentrations of CoQ 9 and CoQ 10 extracted from the heart mitochondria were made in nine different mammalian species, namely mouse, rat, guinea pig, rabbit, pig, goat, sheep, cow, and horse. The data, presented in Table I and Fig. 1, indicate that both the total as well as the relative concentrations of CoQ 9 and CoQ 10 in heart mitochondria vary greatly in different species. The total concentration of mitochondrial CoQ, i.e. CoQ 9 ϩ CoQ 10 , varied about 2-fold in different species with the rank order: Although all nine species examined in this study contained both CoQ 9 and CoQ 10 the ratio of CoQ 10 /CoQ 9 varied Ͼ 600-fold. In species such as the mouse and the rat almost 90% of mitochondrial CoQ occurred as CoQ 9 while in the guinea pig CoQ 9 and CoQ 10 were present in roughly equal amounts. In mitochondria from rabbit, pig, goat, sheep, cow, and horse, CoQ 10 was the predominant form, with CoQ 9 constituting ϳ1.3 to 4.0% of the total CoQ content.
Correlation between CoQ and Superoxide Anion Radical Generation-To determine the relationship between mitochondrial CoQ content and the rate of O 2 . generation in different species, the amounts of CoQ 9 and of CoQ 10 were plotted against the average rates of O 2 . generation by SMPs, partially determined in the context of previous studies (12). As shown in Fig. 1A, the amount of CoQ 9 was directly correlated and that of CoQ 10 was inversely correlated (Fig. 1B)  . Generation in Rat Heart SMPs-Repeated extractions with pentane were found to exponentially deplete the amount of native CoQ 9 from the rat heart SMPs (Fig. 2, inset); the amount remaining after six serial extractions was about 4.5% of the total amount extractable by hexane. In contrast, apparently due to the much lower natural content of CoQ 10 , only three extractions with pentane were sufficient to deplete SMPs of CoQ 10 to a level below the detection threshold of 0.2 M (i.e. 0.015 nmol/mg of SMP protein).
To determine the effect of pentane extractions on the functional state of the SMPs, rates of oxygen consumption and O 2 .
generation were determined after each extraction procedure. The rate of succinate-supplemented oxygen consumption was highest in the unextracted SMPs, decreasing linearly following each extraction procedure, reaching 25% of the initial value after seven successive extraction procedures (Fig. 2). Addition of antimycin A and TTFA greatly reduced (to Ͻ2%) the rate of oxygen consumption by the depleted SMPs, whereas rotenone had no effect, indicating that O 2 consumption observed was specifically due to succinate oxidase activity. NADH did not, in most instances, stimulate the rate of oxygen consumption by the depleted SMPs. A similar study was conducted on the effect of various pentane extractions on the rate of O 2 . generation by the SMPs.
Again, the rate of O 2 . generation was highest in the unextracted SMPs and progressively declined with each sequential pentane extraction, reaching 45% of the control value after six extraction procedures, where less than 5% of the original CoQ was present (Fig. 3).
Overall the results of the depletion experiments indicated that even after six or seven serial extractions with pentane the SMPs exhibited succinate oxidase activity and were able to generate O 2 . albeit at rates lower than the unprocessed SMPs.

Effects of Reconstitution of Rat Heart SMPs with CoQ
Homologues-Rat heart SMPs that had been extracted with pentane six times, as described above, were reconstituted with different amounts of CoQ 9 and CoQ 10 . Reconstitution with increasing amounts of CoQ 9 or CoQ 10 caused an initial steep increase in the succinate-supplemented rate of oxygen consumption, which was followed by a plateau (Fig. 4B, inset). No significant differences in the rates of oxygen consumption were observed between the SMPs reconstituted with equal amounts of CoQ 9 or CoQ 10 .
Reconstitution of the depleted rat heart SMPs with increasing amounts of CoQ 9 resulted in an initial sharp rise in the rate of O 2 . generation, followed by a more gradual increase (Fig. 4A,   inset). At the highest concentration of repleted CoQ 9 used in FIG. 2. Rates of oxygen consumption in rat heart SMPs after partial depletion of CoQ. Freeze-dried SMPs were depleted of their native CoQ homologues by repeated pentane extractions. After each extraction procedure, the rate of oxygen consumption was measured polarographically with a Clark-type electrode using 7 mM succinate as a substrate. The inset shows the SMP CoQ content remaining after each extraction procedure with pentane. The (remaining) CoQ content was determined in hexane:ethanol extracts by HPLC. Data are mean Ϯ S.E. of three independent experiments. S.E. is not shown in the inset.

FIG. 3. Rates of O 2
. generation in rat heart SMPs following CoQ depletion. Freeze-dried SMPs were depleted of their native CoQ homologues by repeated pentane extractions. After each extraction procedure, the rate of O 2 . generation was measured as superoxide dismutaseinhibitable reduction of acetylated ferricytochrome c as described in the legend of Fig. 1. CoQ content of SMPs, after successive extractions, is shown in Fig. 2 (Fig. 4A, inset). The differences between CoQ 9 -and CoQ 10 -reconstituted SMPs emerged only at concentrations considerably greater than the in vivo level (Table I).
Effects of Reconstitution with CoQ 9 and CoQ 10 on Bovine Heart SMPs-In contrast to the rat, bovine cardiac SMPs contain a relatively high amount of CoQ 10 and a small amount of CoQ 9 (see Table I). Depletion of bovine SMPs by six serial extractions with pentane achieved a 96% extraction of CoQ 10 and virtually the entire amounts of the detectable CoQ 9 . Reconstitution of these depleted SMPs with varying concentrations of CoQ 9 or CoQ 10 indicated different patterns for the two homologues for the rate of oxygen consumption and of O 2 . generation. Augmentation of SMPs with relatively low amounts of CoQ 9 or CoQ 10 caused a sharp increase in both the rate of oxygen consumption (Fig. 5B, inset) and O 2 . generation (Fig. 5A, inset), but at higher concentrations these rates leveled off. Within the physiological range of CoQ content (Table I), there were no differences in the rates of O 2 . generation between SMPs reconstituted with CoQ 9 or CoQ 10 (Fig. 5A, inset). However, the maximal rates of O 2 . generation were greater for CoQ 9 than for CoQ 10 . For example, in the bovine SMPs, reconstituted with 50 nmol of CoQ 9 , the rate of O 2 . generation was 35% greater than in the SMPs reconstituted with an equal amount of CoQ 10 (Fig.  5A).
To further determine whether CoQ 9 and CoQ 10 content above the in vivo level had a different effect on the rate of O 2 .
generation, freeze-dried unextracted bovine SMPs were augmented with CoQ 9 or CoQ 10 . As shown in Fig. 6, the rates of O 2 . generation were stimulated to a greater extent by the addition of CoQ 9 than CoQ 10 . The differences in the rate of O 2 .

FIG. 4. Rates of oxygen consumption and O 2 . generation in
CoQ-depleted/reconstituted rat heart SMPs. Freeze-dried SMPs were depleted of native CoQ homologues by six repeated pentane extractions and reconstituted with specific amounts of CoQ 9 or CoQ 10 in pentane. The reconstituted SMPs were dried and suspended in phosphate buffer, and rates of O 2 . generation, shown in A, were measured as superoxide dismutase-inhibitable reduction of acetylated ferricytochrome c. Rates of oxygen consumption, shown in B, were determined polarographically with a Clark-type electrode using 7 mM succinate as a substrate. The insets depict the relationship between rates of O 2 .
generation and oxygen consumption by SMPs and CoQ concentrations, within the physiological range. Data are mean Ϯ S.E. of three independent experiments.

FIG. 5. Rates of oxygen consumption and O 2 . generation in
CoQ-depleted/reconstituted bovine heart SMPs. Freeze-dried SMPs were depleted of native CoQ by six repeated extractions with pentane and reconstituted with specific amounts of CoQ homologues as described in Fig. 4 Results of this study should be interpreted in light of the fact that the preparatory procedures involving freeze-drying and pentane extractions, although widely employed (15-17, 26, 27), have an irreversible effect on the functional state of SMPs. For example, freeze-drying of SMPs followed by depletion and reconstitution with the original (natural) amount of CoQ  (28) are relatively minor. Nevertheless, the relative length of the polyisoprenoid chain and the resultant effects on the hydrophobicity of the molecule have been shown to have an effect on the location of the molecule within the phospholipid bilayer of the cell membrane. Although the relative position of CoQ 9 and CoQ 10 in the phospholipid bilayer has not been precisely determined, the CoQ homologues with relatively short polyisoprenoid chains are believed to lie closer to the surface of the bilayer, whereas the long chained ones are thought to be nearer to the center of the bilayer (29 -32). For example, studies by Kagan et al. (33) have shown that short chain ubiquinols are relatively more efficient in inhibiting Fe 2ϩ -ascorbate-induced lipid peroxidation, suggesting that the polyisoprenoid chain length has an effect on the interaction between the quinols and the ROS present in the aqueous phase. Studies by Matsura et al. (34) on rat and guinea pig hepatocytes also implicate major differences in antioxidant efficiency between the reduced CoQ 9 and the reduced CoQ 10 . In response to the hydrophilic radical initiator, 2-2-azobis-(2-amidinopropane) dihydrochloride, CoQ 9 was found to be preferentially oxidized as compared with CoQ 10 homologue and thus may be more accessible to ROS present in the surrounding aqueous phase. This mechanism, albeit hypothetical, may underlie the relatively higher rates of O 2 . generation in highly CoQ 9 -rich SMPs observed in this study. Functional differences between CoQ 9 and CoQ 10 have also been reported by Edlund et al. (35), who found that treatment of mumps virus-infected cultured neurons with CoQ 10 protected the cells from degeneration, whereas no effects were observed in response to CoQ 9 treatment. Results of the present in vitro studies demonstrate that CoQ 9 and CoQ 10  cannot be explained on the basis of relative CoQ 9 or CoQ 10 content.