INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

IT IS WIDELY APPRECIATED that the mitochondrial genome is prone to oxidant-mediated damage, being 10- to 100-fold more sensitive than nuclear DNA (36). Moreover, mutations and deletions in the mitochondrial genome in cells of the central nervous system and elsewhere have been linked to neurodegenerative disorders and other age-related diseases (11, 27, 30, 32, 33). Depletion of mitochondrial (mt) DNA with ethidium bromide is also known to suppress ATP synthesis and cause defects in cell function (8). These findings indicate that mtDNA damage could play a causal role in disorders linked to excessive generation of reactive oxygen species.

Pulmonary vascular endothelial cells (ECs) are among the most important targets of reactive oxygen species elaborated in acute lung injury (6, 7, 23). Emerging data suggest that there may be important differences between phenotypically distinct EC populations residing at different segments of the pulmonary circulation (1, 9, 24, 25). Although the effects of oxygen radicals on mtDNA in ECs from the lung have yet to be studied, a recent report by Ballinger and colleagues (3) demonstrated that the mitochondrial genome in human umbilical vein ECs was damaged by hydrogen peroxide and peroxynitrite. Accordingly, a proximate goal of the present study was to determine if mtDNA in pulmonary arterial (PA), microvascular (MV), and pulmonary venous (PV) ECs was sensitive to reactive species-induced damage. In addition, because cytotoxic reactive species can be generated from both exogenous and endogenous sources, mtDNA damage in lung ECs exposed to xanthine oxidase (XO), which produces reactive species directly, was compared with that induced by menadione, which promotes endogenous oxidant generation by redox cycling within the mitochondria. Finally, we used pharmacological tools to confirm that the mtDNA damage evoked by these stimuli in PAECs involved the iron-catalyzed production of hydroxyl radicals as it does in other systems (19, 22).

Accumulating data show that mitochondria are well endowed with enzymatic pathways that repair oxidative DNA damage (4, 14). Defective mtDNA repair has also been associated with certain human pathologies (13, 15). Interestingly, there may be cell type-specific differences in mtDNA repair capabilities as evidenced by recent findings (20) that primary cultures of selected glial cell populations exhibit disparities in the rates at which oxidative mtDNA damage is repaired. Even more provocative is the observation that sensitivity to oxidant-induced mtDNA damage and death in these central nervous system cell populations do not seem to be related to the mitochondrial antioxidant burden and, instead, are closely linked to the capacity for mtDNA repair (20). Glial cell types that exhibit avid mtDNA repair are less sensitive to oxidant-mediated cytotoxicity than are closely related cell populations that are relatively deficient in eliminating mtDNA lesions. To determine whether lung EC phenotypes are different in terms of their capacities for mtDNA repair, we examined the rates at which XO- and menadione-induced mtDNA lesions were removed from rat main PAECs, MVECs, and PVECs. In addition, to probe the association between mtDNA repair capacity and EC sensitivity to oxidant stress, we evaluated the relationship between repair kinetics and cell death in the three lung EC phenotypes.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Rat main PAEC, MVEC, and PVEC cultures. Main PAs and PVs were isolated from 250- to 300-g Sprague-Dawley rats killed with an overdose of Nembutal. Isolated arteries were opened, and the intimal lining was carefully scraped with a scalpel. Pulmonary MVECs were harvested from tryptic digests of peripheral lung tissue (24). The harvested cells were then placed into flasks (Corning, Corning, NY) containing Ham's F-12 nutrient mixture and DMEM mixture (1:1) supplemented with 10% fetal bovine serum, 100 U/ml of penicillin, and 0.1 mg/ml of streptomycin (GIBCO BRL, Grand Island, NY). The culture medium was changed once a week, and after reaching confluence, the cells were harvested with a 0.05% solution of trypsin (GIBCO BRL) and passsaged up to 15 times. The EC phenotype was confirmed by acetylated low-density lipoprotein uptake, factor VIII-related antigen immunostaining, and the lack of immunostaining with smooth muscle cell -actin antibodies (Sigma, St. Louis, MO).

Detection of mtDNA damage with quantitative Southern and ligation-mediated PCR analyses. PAECs cultured on 100-mm petri plates were challenged with ascending doses of either XO plus hypoxanthine (0.5 mM) or menadione. Quantitative Southern analysis was used to examine changes in the equilibrium density of the lesions in the entire mitochondrial genome (14). DNA isolated from the ECs was precipitated, resuspended in Tris-EDTA buffer (pH 8.0), and treated with DNase-free RNase. Purified DNA was then digested with BamHI, and complete digestion was verified on minigels. After restriction, the samples were resuspended in a small volume of Tris-EDTA buffer and precisely quantified with a Hoefer TKO minifluorometer and standards kit. Samples containing 5-10 µg of DNA were heated, cooled at room temperature, and then incubated for 15 min with sodium hydroxide (0.1 N) to cleave DNA at the sites of oxidative injury to the deoxyribose backbone. Samples were then combined with 5 µl of loading dye, loaded onto a 0.6% alkaline gel, and electrophoresed in an alkaline buffer. After the gel was washed, DNA was transferred by vacublotting (Millipore, Bedford, MA) onto a Zeta-Probe GT nylon membrane (Bio-Rad Laboratories, Hercules, CA) and cross-linked to the membrane with a GS Gene Linker (Bio-Rad). After a 10-min preincubation, the membrane was hybridized with a PCR-generated genomic DNA probe for the mitochondrial genome. The probe used to hybridize to mtDNA was generated via PCR from a mouse mtDNA sequence with the use of the following primers: 5'-GCAGGAACAGGATGAACAGTCT-3' from the sense strand and 5'-GTATCGTGAAGCACGATGTCAAGGGATGTAT-3' from the antisense strand. The 725-bp product recognized a 10.8-kb restriction fragment when hybridized to rat mtDNA digested with BamHI. Using a similar strategy, we also probed a 3.4-kb sequence in the 5' promoter of the nuclear vascular endothelial growth factor (VEGF) gene. The membranes were washed according to the manufacturer's instructions, and hybridization bands were detected with a Bio-Rad GS-250 molecular imager. Changes in the equilibrium "lesion" density were calculated as ln of the quotient of hybridization intensities in the treated and control bands (5).

Cytotoxicity assay. The cytotoxic effects of ascending doses of either XO or menadione were determined 24 h after removal of the oxidant generators with the commercially available MTT assay (Promega, Madison, WI).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Initial studies evaluated dose-response relationships for XO- and menadione-induced mtDNA damage in the three EC phenotypes. Cells were exposed to the free radical-generating systems for 1 h, a duration shown by preliminary experiments to be sufficient for mtDNA damage to plateau at any given XO concentration. Representative Southern blot analyses of the concentration-dependent effects of XO are shown in Fig. 1. It is evident that XO decreased hybridization intensity, which is indicative of mtDNA damage, in all three phenotypes, but PVECs tended to be the most sensitive compared with the other two EC populations. Changes in steady-state mtDNA lesion density were calculated as a function of the XO concentration and pooled for four experiments and are displayed for each EC phenotype in Fig. 1. The apparent rank order of sensitivity to XO-induced mtDNA damage was (from most to least) PVECs > PAECs > MVECs.

View larger version (38K): [in this window] [in a new window]   Fig. 1.   Rat pulmonary venous (PV), pulmonary arterial (PA), and pulmonary microvascular (MV) endothelial cells (ECs) were cultured under control (Con) conditions or challenged with indicated concentrations of xanthine oxidase (XO) plus 0.5 mM hypoxanthine for 1 h, after which the cells were lysed. High molecular weight DNA was isolated and digested to completion with BamHI. DNA samples were exposed to 1 N NaOH before Southern blot analysis and hybridization with a mitochondrial (mt) DNA-specific probe. Top: representative autoradiogram of quantitative Southern analysis of alkali-labile mtDNA damage. There are 2 lanes each for control and XO-treated samples. Note diminished hybridization intensity, indicative of mtDNA damage. Bottom: calculated increases in equilibrium lesion density normalized to 10 kb. Values are means ± SE for 4 experiments.

Representative Southern blot analyses depicting the effects of menadione are shown in Fig. 2. Like XO, the reactive species generated endogenously in response to menadione-damaged mtDNA, but the relative sensitivities of the EC phenotypes appeared to be different. Calculated increases in steady-state lesion density, also displayed in Fig. 2, indicate that MVECs and PVECs were roughly equal in terms of the ability of menadione to damage mtDNA, whereas PAECs were somewhat less sensitive, especially at the lowest menadione dose tested.

View larger version (35K): [in this window] [in a new window]   Fig. 2.   Rat MVECs, PVECs, and PAECs were cultured under control conditions or challenged with menadione for 1 h, after which the cells were lysed. High molecular weight DNA was isolated and digested to completion with BamHI. DNA samples were exposed to 1 N NaOH before Southern blot analysis and hybridization with a mtDNA-specific probe. Top: representative autoradiogram of quantitative Southern analysis of alkali-labile mtDNA damage. Note diminished hybridization intensity indicative of mtDNA damage. Bottom: calculated increases in equilibrium lesion density normalized to 10 kb. Values are means ± SE for 4 experiments.

In the nonvascular cell systems examined thus far, the iron-catalyzed production of hydroxyl radicals seems to play an important role in oxidant-mediated DNA damage (19, 22). To determine whether XO- and menadione-induced mtDNA damage in lung ECs involved a similar mechanism, PAECs were treated with XO (5 mU/ml) and menadione (100 µM) in the absence and presence of the putative hydroxyl radical scavenger dimethylthiourea (100 µM) (17, 34) or the iron chelator deferoxamine (150 µM). Cells were harvested 60 min after the simultaneous addition of the free radical generators and inhibitory agents, and quantitative Southern analysis was used to quantify changes in the steady-state density of mtDNA lesions. As shown in Fig. 3, both dimethylthiourea and deferoxamine suppressed XO- and menadione-induced mtDNA damage to an approximately equal extent, although deferoxamine tended to be less effective in attenuating XO-mediated damage than damage induced by menadione.

View larger version (44K): [in this window] [in a new window]   Fig. 3.   Rat PAECs were challenged with either XO (5 mU/ml) or menadione (MEN; 100 µM) in the presence and absence of 100 µM dimethylthiourea (DMTU) or 150 µM deferoxamine (DFX) for 60 min, after which cells were harvested and the extent of mtDNA damage was determined by quantitative Southern analysis. Representative autoradiogram (top) indicates that although neither DMTU nor DFX altered the baseline hybridization intensity, both agents suppressed the profound decrease in hybridization caused by either XO or menadione. Increases in the steady-state lesion density as determined in 4 separate experiments were pooled, and the effects of DMTU and DFX are expressed as percent inhibition of the responses to XO and menadione alone (bottom). All treatment groups were significantly depressed relative to control groups (P < 0.05) by 1-way ANOVA and Dunnett's test).

The deleterious effects of XO and menadione on the nuclear VEGF gene also were explored. No damage to the VEGF gene was detected at any XO or menadione dose examined (data not shown).

To examine mtDNA repair capacity, the three EC phenotypes were treated with 10 mU/ml of XO to produce an increase of ~1.5 strand breaks/10 kb in equilibrium lesion density. The cells were harvested immediately after a 1-h treatment with XO or were allowed to repair the damage for 4 or 24 h in XO-free medium. The representative Southern blot analyses shown in Fig. 4 indicate that PVECs exhibited a rather low capacity for mtDNA repair compared with the other two EC phenotypes. As also displayed in Fig. 4, when mtDNA integrity is expressed as the percentage repaired, i.e., as a percentage of the lesion burden present immediately after XO treatment, it was evident that PVECs were virtually incapable of repairing XO-induced mtDNA damage, at least over a 24-h period. Conversely, time-dependent repair in PAECs and MVECs resulted in almost complete repair of the lesions within 24 h. The outcome of a companion experiment that used 200 µM menadione to promote reactive species generation is shown in Fig. 5. In this instance, repair in PVECs and MVECs was almost complete within 4 h and approached 100% removal of the lesions. Repair in PAECs was somewhat slower but was still nearly complete by 24 h.

View larger version (32K): [in this window] [in a new window]   Fig. 4.   Rat PVECs, PAECs, and MVECs were cultured under control conditions or treated for 1 h with 10 mU/ml of XO plus 0.5 mM hypoxanthine. After XO treatment, cells were harvested immediately (0 h) or placed in XO-free medium and allowed to recover (repair) for 4 or 24 h. At the indicated times, high molecular weight DNA was isolated, digested, and subjected to quantitative Southern blot analysis. Top: representative autoradiogram of quantitative Southern blot analysis of alkali-labile mtDNA damage in control and XO-treated EC phenotypes as a function of time after the removal of XO. There are 2 lanes each for control and XO-treated samples. Note recovery of hybridization intensity at 4 and 24 h, indicative of repair of mtDNA damage evident immediately after XO treatment. Bottom: calculated repair of mtDNA lesions expressed as percent reversal of the initial lesion density. Values are means ± SE for 4 experiments.

View larger version (38K): [in this window] [in a new window]   Fig. 5.   Rat PVECs, PAECs, and MVECs were cultured under control conditions or treated for 1 h with 200 µM menadione. After menadione treatment, cells were harvested immediately (0 h) or placed in menadione-free medium and allowed to recover for 4 or 24 h. At the indicated times, high molecular weight DNA was isolated, digested, and subjected to quantitative Southern blot analysis. Top: representative autoradiogram of quantitative Southern blot analysis of alkali-labile mtDNA damage in control and menadione-treated EC phenotypes as a function of time after removal of menadione. Note recovery of hybridization intensity at 4 and 24 h, indicative of repair of mtDNA damage Bottom: calculated repair of mtDNA lesions expressed as percent reversal of the initial lesion density. Values are means ± SE for 4 experiments.

We next assessed whether the phenotypic differences in the XO- and menadione-induced mtDNA damage and repair described above were accompanied by differences in oxidant-mediated cytotoxicity determined 24 h after removal of the oxidant-generating stimuli. As shown in Fig. 6, PVECs were the most sensitive to XO-mediated cell death followed by PAECs and MVECs. In the case of menadione-induced cytotoxicity, MVECs and PAECs were of approximately equal sensitivity, followed by PVECs.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   Concentration-dependent cytotoxicity evoked by XO (top) and menadione (bottom) as assessed by MTT assay in PAECs, MVECs, and PVECs. Note different rank order of sensitivities of the different EC phenotypes to the free radical generators. Each point is the mean ± SE of at least 4 determinations.

Ligation-mediated PCR was used to determine whether the nucleotide specificity of XO-mediated mtDNA damage differed among the EC phenotypes or from damage induced by menadione. These analyses focused on a 200-bp sequence containing one of the break sites for the "common" 5-kb deletion known to occur in aging (11, 28). Representative autoradiograms depicting ligation-mediated PCR determinations of XO- and menadione-induced damage in the three EC phenotypes are shown in Figs. 7 and 8, respectively. A qualitative analysis of these data reveals two significant features. First, as would be predicted from the Southern blot analyses described previously, the increased hybridization intensities associated with increasing XO and menadione concentrations indicate that both agents caused dose-related damage to specific nucleotides within the 200-bp segment of the mitochondrial genome. Second, and more important, the pattern of damage for each agent did not differ markedly between the EC phenotypes.

View larger version (143K): [in this window] [in a new window]   Fig. 7.   Ligation-mediated PCR analysis of nucleotide-specific damage within the common deletion sequence of the mtDNA heavy strand in control PAECs, PVECs, and MVECs and in cells treated with either 10 or 5 mU/ml of XO plus 0.5 mM hypoxanthine (nos. at top). DNA samples were treated with (+) and without () 0.1 N NaOH to cleave to the deoxyribose backbone at sites of oxidative damage. A Maxim-Gilbert sequence ladder shows the nucleotide bases guanine (G), adenine (A), thymine (T), and cytosine (C). Letters and numbers at far right, base and position, respectively. Damage at a given nucleotide is indicated by an increase in hybridization intensity. Note 1) increasing hybridization intensities at specific nucleotides with increasing XO concentration and, most importantly, 2) the similar patterns of damage between the 3 EC phenotypes.

View larger version (160K): [in this window] [in a new window]   Fig. 8.   Ligation-mediated PCR analysis of nucleotide-specific damage within the common deletion sequence of the mtDNA heavy strand in control PAECs, PVECs, and MVECs and in cells treated with either 200 µM or 300 µM menadione (nos. at top). Letters and nos. at far right, base and position, respectively. Damage at a given nucleotide is indicated by an increase in hybridization intensity. Note 1) increasing hybridization intensities at specific nucleotides with increasing menadione concentration and, most importantly, 2) the similar patterns of damage between the 3 EC phenotypes.

On the other hand, the pattern of damage to specific nucleotides did differ between menadione and XO. Figure 9 shows a comparison of damage maps for XO- and menadione-induced mtDNA damage in PAECs. In this analysis, hybridization intensities for each nucleotide, as determined from four ligation-mediated PCR analyses, were scored from 1 to 6, with 1 reflecting the least prevalent and 6 indicative of the most prevalent lesions, and were plotted as a function of the specific nucleotide. The resulting maps revealed that there were a number of nucleotides frequently damaged by XO but not by menadione and vice versa. In further support of such differences in nucleotide specificity, the proportion of four nucleotides with severe damage, defined as a damage level  3 from the ligation-mediated PCR analyses, was determined and, as shown in Table 1, was guanine = adenine > thymine > cytosine in the case of XO-mediated damage and guanine > adenine > thymine > cytosine for damage evoked by menadione. The rate of damage to guanine in menadione-treated cells exceeded that in XO-treated ECs.

View larger version (12K): [in this window] [in a new window]   Fig. 9.   Ligation-mediated PCR-generated map of nucleotide-specific damage evoked by XO (A) and menadione (B) in cultured PAECs. The analysis focused on a 54-nucleotide span within the common deletion sequence of the mtDNA heavy chain. Position of bars indicates damage to a specific nucleotide; bar height indicates relative prevalence of damage as determined by hybridization intensities normalized to a scale of 1 (least lesions) to 6 (most prevalent lesions). Note differences in patterns of damage between XO and menadione.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Multiple studies (11, 27, 30, 32, 33) show that mutations and deletions in the mitochondrial genome contribute to pathogenesis in a spectrum of diseases. The first report examining the sensitivity of systemic vascular EC mtDNA to oxidant stress appeared recently (3), and it showed that mtDNA damage evoked by reactive oxygen and nitrogen species was accompanied by diminished ATP synthesis, impaired mitochondrial RNA expression, and altered mitochondrial membrane potential. It is unknown whether mtDNA in lung vascular ECs is sensitive to oxidant-mediated damage; resolution of this issue was the proximate goal of the present study. In a related context, in light of the increasing appreciation that there are segmental differences among EC phenotypes in the pulmonary circulation (1, 9, 24, 25), we determined if there were disparities between ECs harvested from PA, MV, and PV segments of the rat lung in terms of their sensitivities to oxidant stress. Finally, because the pulmonary vascular endothelium can be damaged by oxidants elaborated from both exogenous and endogenous sources, we compared the effects of XO and menadione, respectively, for their abilities to erode mtDNA integrity.

The present results show that XO- and menadione-generated oxidants cause prominent damage to the mitochondrial genome in three lung EC phenotypes, with no detectable effects on nuclear DNA. The relative sensitivity of mtDNA compared with that in the nucleus was also noted by Ballinger et al. (3) for systemic vascular smooth cells and ECs as it has been for other nonvascular cells (36). The reason that mtDNA is more sensitive than nuclear DNA has traditionally been ascribed to the lack of protective histone muscle proteins and the relative underdevelopment of DNA repair mechanisms, although more recent work shows that mitochondria do exhibit effective DNA repair for some types of damage (4).

We found that the EC phenotypes exhibited different sensitivities to mtDNA damage, with MVECs being less sensitive to XO than PAECs and PVECs. On the other hand, MVECs were most sensitive to menadione-induced mtDNA damage, with PAECs somewhat more resistant. The factors underlying these divergent sensitivities are unknown. Certainly, different abundances of cellular antioxidants and scavenging enzymes could play a role, although to what extent these pathways protect mtDNA from oxidant-mediated damage has not been systematically addressed. It is also possible that the EC phenotypes are intrinsically different in terms of the ability of XO and menadione to promote oxidant generation. For example, ECs are known to internalize XO by endocytosis, after which the free radicals generated intracellularly are capable of disrupting nitric oxide-dependent signaling (21). It is thus possible that the different sensitivities of mtDNA in the EC phenotypes studied in the present report are related to different capacities for XO binding and endocytosis. Counter to this idea, a recent preliminary finding from our laboratory that XO-induced mtDNA damage in cultured PAECs is unaffected by suppression of XO uptake argues that free radicals generated by extracellular XO are sufficient to erode mtDNA integrity (31). In the case of menadione, it should be considered that mitochondrial metabolism is a determinant of its ability to produce superoxide anion (16). As such, EC phenotype-dependent metabolic differences could account for their differing sensitivities to menadione-induced mtDNA damage.

It is widely appreciated that the iron-catalyzed formation of hydroxyl radicals plays an important role in oxidant-mediated DNA damage (19, 22). The oxidizing species are probably not diffusible hydroxyl radicals but rather hydroxyl radicals generated in the immediate vicinity of iron bound to the DNA target (18, 22). In this regard, it appears uncertain whether in the intact cell there is sufficient mtDNA-bound iron for the Fenton reaction to proceed. To address this issue, we examined the effects of the putative hydroxyl radical scavenger dimethylthiourea (17, 34) and found that it suppressed both XO- and menadione-induced mtDNA damage. More interestingly, deferoxamine, an iron chelator, also blocked mtDNA damage evoked by both stimuli in PAECs. These observations suggest that mtDNA, like nuclear DNA, appears to bind sufficient iron to support hydroxyl radical formation via the Fenton reaction and that the ultimate mtDNA-damaging reactive species is hydroxyl radical. It should be noted that these experiments were performed only in PAECs, but it seems likely that the involvement of iron and the hydroxyl radical can be extended to the other EC phenotypes insofar as the nucleotide pattern of damage in the mitochondrial genome failed to display any phenotype-dependent differences (see below).

Another mechanism suspected of governing the sensitivity of mtDNA to oxidant-mediated damage is the mtDNA repair pathways. In contrast to the traditional concept that mitochondria are deficient in DNA repair (10), experiments in nonvascular cells indicate that they are, at a minimum, equipped to perform base excision repair of the mitochondrial genome (4). Because mtDNA repair has not been documented in vascular cell types, the present study determined if lung ECs were able to remove oxidant lesions evoked by XO and menadione. As might be expected, pulmonary vascular ECs were generally able to repair lesions evoked by both stimuli, but there were rather prominent phenotype-dependent differences. XO-induced damage was repaired well by PAECs and MVECs but not in cells derived from PVs. Menadione-induced mtDNA lesions were repaired proficiently by all three phenotypes, but PAECs exhibited a somewhat slower rate than the other two. The concept that different phenotypes can exhibit divergent mtDNA repair capacity is not without precedent. In a study of closely related glial cell types, Hollensworth and coworkers (20) showed that astrocytes, which are highly resistant to oxidant-induced mtDNA damage, exhibit more efficient mtDNA repair than oligodendrocytes or microglia, which are relatively prone to oxidant-induced mtDNA damage. The abundance of various antioxidant pathways or scavengers did not correlate with mtDNA sensitivity to oxidants in these glial cell lines.

As expected on the basis of numerous previous reports (e.g., Refs. 6, 7), the oxidant stresses examined in the present study were lethal to lung ECs. A provocative pattern emerged when the sensitivity to XO-mediated cell death was compared with phenotype-dependent differences in mtDNA repair. The EC phenotype that repaired XO-induced mtDNA damage most avidly (MVECs) was resistant to its cytotoxic actions, whereas PVECs, which were deficient in repair of XO damage, were most sensitive to its cytotoxicity. In this regard, the aforementioned study (20) on glial cell phenotypes found that the propensity for menadione-induced apoptosis, which is associated with activation of caspase-9, was inversely related to the rate of repair of menadione-induced damage to mtDNA. Moreover, enhancement of mtDNA repair in either HeLa cells (12) or PAECs (unpublished observations) through stable or transient transfection, respectively, with the gene for the DNA repair enzyme, Ogg1, linked to a mitochondrial localization sequence confers protection against oxidant-mediated mtDNA damage and cytotoxicity. Although the available data are far from conclusive, it is tempting to speculate that the capacity to repair mtDNA damage or, stated differently, the persistence of mtDNA damage, could be a determinant of sensitivity to oxidant-mediated EC death.

Using ligation-mediated PCR, we found that the nucleotide specificity by which XO and menadione damaged an ~200-bp sequence located within the common deletion fragment of the mitochondrial heavy chain was internally consistent; with only minor exceptions, the same nucleotides were damaged regardless of the EC phenotype examined. On the other hand, there were differences between XO and menadione in terms of the nucleotide specificity of the damage, with the rate of damage to guanine and adenine similar for XO, whereas guanines seemed to be preferentially damaged by menadione. Neither the reasons for these differences nor their biological significance can be explained at present. Nevertheless, it is likely that although the proximate radical species generated by both XO and menadione is superoxide anion, the different anatomic locations of superoxide anion generation, extracellular or cytoplasmic for XO versus intramitochondrial for menadione, could afford different opportunities for interactions with other reactive species. As a consequence, the rates and/or nature of the radical species interacting with mtDNA could differ, thus accounting for the divergent nucleotide specificities. In a related context, an emerging concept relative to the biology of DNA damage is that certain nucleotides within a given sequence may be prone to damage or deficiently repaired and could thus play more important roles in determining cellular outcomes (29). Against this background, perhaps phenotype-dependent differences in the cytotoxic actions of XO and menadione are somehow linked to the different nucleotide specificities of the damage they induce. Clearly, much additional study will be required to address this possibility.

In summary, the present study shows that XO- and menadione-generated oxidants preferentially erode mtDNA integrity relative to nuclear DNA in lung EC phenotypes. There appear to be phenotype-dependent differences in the propensity for mtDNA damage, in the capacity for mtDNA repair, and in the cytotoxicity associated with these oxidant stresses that warrant additional investigation. An outstanding question prompted by our study as well as by the preceding report on oxidant mtDNA damage in systemic vascular ECs by Ballinger et al. (3) pertains to its biological significance. An editorial (35) accompanying the Ballinger paper appropriately emphasized the vagaries in causally linking EC mtDNA mutations and deletions to progressive vasculopathies like atherosclerosis because it is believed that 80-90% of the total mtDNA pool must be affected before disease progression takes place (26). On the other hand, our observation that mtDNA repair kinetics may be predictive of XO-induced lung EC death, coupled with the finding of Ballinger et al. (3) that mtDNA damage was associated with impaired ATP synthesis and reduced mitochondrial gene expression, raises the interesting possibility that acute damage to mtDNA could be a proximate cause of EC dysfunction and death. Thus even in the absence of mtDNA mutation or deletion, damage to mtDNA could be significant in the various forms of acute lung EC injury in which oxidants are suspected of playing a pathogenic role.