Testing hypotheses of aging in long-lived mice of the genus Peromyscus: association between longevity and mitochondrial stress resistance, ROS detoxification pathways, and DNA repair efficiency

Abstract

In the present review we discuss the potential use of two long-lived mice of the genus Peromyscus—the white-footed mouse (P. leucopus) and the deer mouse (P. maniculatus) maximum lifespan potential ∼8 years for both—to test predictions of theories about aging from the oxidative stress theory, mitochondrial theory and inflammatory theory. Previous studies have shown that P. leucopus cells exhibit superior antioxidant defense mechanisms and lower cellular production of reactive oxygen species (ROS) than do cells of the house mouse, Mus musculus (maximum lifespan ∼3.5 years). We present new data showing that mitochondria in P. leucopus cells produce substantially less ROS than mitochondria in M. musculus cells, and that P. leucopus mitochondria exhibit superior stress resistance to those of M. musculus. We also provide evidence that components of the DNA repair system (e.g., pathways involved in repair of DNA damage induced by γ-irradiation) are likely to be more efficient in P. leucopus than in M. musculus. We propose that mitochondrial stress resistance, ROS detoxification pathways and more efficient DNA repair contribute to the previously documented resistance of P. leucopus cells toward oxidative stress-induced apoptosis. The link between these three pathways and species longevity is discussed.

Long-lived mice of the genus Peromyscus: a useful model for successful aging in small, mouse-like rodents

Comparative studies of species with unusually long lifespans may provide useful insights into the mechanisms determining successful aging, and identify genetically encoded mechanisms responsible for differences in mammalian longevity. The practical problem is to identify species pairs that are reasonably closely related taxonomically but that differ significantly in longevity, have genomic information available, are relatively easy to obtain, and can be bred and maintained in the laboratory at low cost. In this review we focus on the potential use of the long-lived mice of the genus Peromyscus [including the white-footed mouse (P. leucopus) and the deer mouse (P. maniculatus)] to test predictions of the oxidative stress theory, mitochondrial theory and inflammatory theory of aging.

Sacher and Hart (1978) originally proposed P. leucopus and Mus musculus (house mouse) as a longevity contrast pair. Both species belong to the superfamily of mouse-like rodents (Muroidea) and share a common ancestor 20–25 million years ago (Steppan et al. 2004). However, mice of the genus Peromyscus, despite close physical resemblance to the house mouse, have unusually long lifespans for their size: the maximum recorded longevities for P. leucopus and P. maniculatus in captivity are 8.2 years (Sacher and Hart 1978), and 8.33 years (Nowak and Paradiso 1983), respectively, approximately twice as long as the maximum lifespan potential of M. musculus. Previous data showed that, in aged P. leucopus, the hypothalamic-pituitary-ovarian axis remains intact and fertility is maintained at least until 5.5 years (Steger et al. 1980; Burger and Gochfeld 1992), the rate of accumulation of DNA damage (in liver and kidney cells) is delayed (Su et al. 1984), and cells and tissues exhibit a number of anti-aging adaptations to combat oxidative damage (Csiszar et al. 2007c). Given these considerations, P. leucopus is a useful model of longevity in small mouse-like rodents. There are also practical advantages to using rodents of the Peromyscus genus in aging studies (Burger and Gochfeld 1992; Guo et al. 1993), including the availability of information on husbandry, toxicology, and pathology, and the fact that the genome of at least four Peromyscus species will be sequenced in the foreseeable future (P. leucopus and P. maniculatus are scheduled for 2× and 7× coverage, respectively; for current status see http://www.genome.gov/10002154). Since the selection of animal models in biogerontology is often driven by feasibility considerations, it is important to note that mice of the genus Peromyscus are readily available from the Peromyscus Genetic Stock Center at the University of South Carolina (http://stkctr.biol.sc.edu/). These mice are the descendants of 38 ancestors captured between 1982 and 1985 in North Carolina.

Cellular oxidative stress resistance

Harman originally proposed the free radical theory of aging half a century ago (Harman 1956), yet the relationship between oxidative stress and aging is still much debated (Sohal et al. 1990, 1993; Sohal and Brunk 1992; Sohal and Orr 1992). There is strong evidence that aging in mammals is associated with an increased production of reactive oxygen species (ROS), and oxidative macromolecular damage accrues with age in virtually every tissue studied (Van Remmen and Richardson 2001; Van Remmen et al. 2003a; Csiszar et al. 2005). However, genetic knockout mice for major cellular antioxidant enzymes often fail to exhibit significant changes in longevity (Van Remmen et al. 2003b; Mansouri et al. 2006; Sentman et al. 2006). In contrast, the general concept that oxidative stress is involved in many age-related diseases, including atherosclerosis, hypertension, diabetic vasculopathy and Alzheimer’s disease, appears robust.

We are currently testing predictions of the oxidative stress hypothesis of aging in long-lived mice of the genus Peromyscus. Because oxidative stress clearly plays a central role in cardiovascular aging (Csiszar et al. 2002, 2005; Ungvari et al. 2004; Labinskyy et al. 2006b), we have compared vascular ROS homeostasis and oxidative stress resistance in P. leucopus and M. musculus. Our recent studies have demonstrated that in vascular tissues of P. leucopus there is attenuated production of ROS (O₂⁻; H₂O₂) from NAD(P)H oxidase (Csiszar et al. 2007c). The differences in NAD(P)H oxidase activity in short- and long-lived species is significant, because upregulation of NAD(P)H oxidase has been shown to contribute to age-related oxidative stress in the cardiovascular system of rats (Hamilton et al. 2001; Csiszar et al. 2002; Adler et al. 2003). NAD(P)H oxidase also plays a major role in vascular pathophysiological alterations in metabolic diseases (e.g., diabetes), which many investigators consider “accelerated vascular aging” (based on similarities of the gene expression profile in senescent vessels to those observed in metabolic diseases). It is unknown how NAD(P)H oxidase activity differs among other (e.g., brain) tissues of P. leucopus and M. musculus. Future studies should also elucidate whether P. leucopus is protected from age-related increases in NAD(P)H oxidase activity.

We have recently reported a higher glutathione peroxidase and catalase content in P. leucopus, and a more abundant expression of eNOS (endothelial nitric oxide synthase) associated with increased endothelial NO production in large arteries (Csiszar et al. 2007c). Previous studies have also shown that brain and heart of P. leucopus have higher activities of catalase and glutathione peroxidase (Sohal et al. 1993) and lower levels of protein oxidative damage, as well as lower susceptibility to oxidative damage in response to experimental oxidative stress (Sohal et al. 1993) than those of M. musculus. We also showed that endothelial cells of P. leucopus are substantially more tolerant of oxidative stress (induced by oxidized low-density lipoprotein or H₂O₂ treatment) than those of M. musculus (Csiszar et al. 2007c). Previous studies demonstrated that inhibition of glutathione peroxidase in various cell types enhances oxidative stress-induced apoptosis (Ran et al. 2004; Van Remmen et al. 2004; Ungvari et al. 2007a). Also, there is a positive correlation between glutathione peroxidase activity and oxidative stress resistance in human cell lines (Marklund et al. 1984). These findings suggest that higher cellular glutathione peroxidase content may contribute to the superior oxidative stress resistance in P. leucopus. Inhibition of hemeoxygenase-1 (HO-1) also can enhance oxidative stress-induced apoptosis (Abraham and Kappas 2005; Kruger et al. 2006; Ungvari et al. 2007a). Because expression of HO-1 is also greater in the vascular tissue of P. leucopus than in M. musculus vessels (Csiszar et al. 2007c), it is possible that HO-1 contributes to cellular resistance to oxidative damage in P. leucopus cells.

There is an emerging view that ROS, in addition to inactivating NO (which acts as a cellular survival factor) and causing oxidative damage, play important signaling roles in the cell. Oxidative/nitrosative stress and consequent activation of numerous downstream effector pathways are thought to be implicated in the inflammatory process in most organs (Pacher et al. 2007). Many age-related degenerative diseases, including atherosclerosis (Ross 1993), are, in part, inflammatory diseases, and recent studies have shown that even in “healthy aging” there is a pro-inflammatory shift in skeletal muscle (Pedersen et al. 2003; Roubenoff et al. 2003; Phillips and Leeuwenburgh 2005), vascular (Csiszar et al. 2002, 2003, 2004) and cardiac (Lee et al. 2002) gene expression profiles (including upregulation of TNFα, IL-6 and iNOS). There is growing evidence that high levels of inflammatory cytokines contribute to a pro-inflammatory microenvironment that facilitates the development of vascular disease and sarcopenia in aging. In particular, it has become established that both vascular and skeletal muscle aging are associated with dysregulation of TNFα expression (Csiszar et al. 2002, 2003, 2004, 2007b; Pedersen et al. 2003; Roubenoff et al. 2003; Bátkai et al. 2007). TNFα is a master regulator of NAD(P)H oxidase activity, leukocyte infiltration of tissues, and apoptosis. Despite the increasing evidence for a role of inflammation in aging, there are no current studies comparing inflammatory processes in short- and long-lived species. We propose that the M. musculus–Peromyscus longevity contrast pair can be exploited to test for species differences in the link between age-related inflammation, oxidative stress and longevity.

Recent studies have uncovered important cross-talk between inflammatory cytokines, ROS and pro-inflammatory genes in the pathogenesis of age-related diseases. Of particular note are studies demonstrating that aging-induced oxidative stress activates NF-κB in the vascular endothelium (Helenius et al. 1996; Donato et al. 2007; Ungvari et al. 2007b), which is likely a major factor contributing to increased expression of adhesion molecules and iNOS (Cernadas et al. 1998; Csiszar et al. 2002; Ungvari et al. 2007b). NF-κB activation and chronic inflammation seem to be a generalized phenomenon in aging, because increases in NF-κB activity have been observed in aged rat skeletal muscle, liver, brain and also cardiac muscle (Helenius et al. 1996; Korhonen et al. 1997; Radak et al. 2004; Zhang et al. 2004). The finding that scavenging of H₂O₂ attenuates NF-κB activation in aged vessels (Ungvari et al. 2007b) suggests a role for mitochondria-derived H₂O₂ in regulation of endothelial NF-κB activity in aging. Local leukocyte recruitment is an early step in atherogenesis and many other degenerative diseases, and is controlled by the expression of cell adhesion molecules. It is significant that inhibition of mitochondrial ROS production has been shown in vitro to decrease expression of the adhesion molecule ICAM-1 and attenuate monocyte adhesiveness to the endothelium in aged rat arteries (Ungvari et al. 2007b). In aging mice that overexpress human catalase in mitochondria, cardiac pathology has been shown to be delayed, oxidative damage reduced, H₂O₂ production and H₂O₂-induced aconitase inactivation attenuated, and mitochondrial deletions reduced (Schriner et al. 2005). On the basis of these findings, future studies should test the hypotheses that (1) cells of long-lived mice of the Peromyscus genus are protected against oxidative stress induced by inflammatory stimuli; and (2) oxidative stress elicits a blunted inflammatory response in Peromyscus as compared to the shorter-lived M. musculus.

Mitochondrial oxidative stress resistance

The mitochondrial theory of aging, put forward by Harman (Harman 1972), postulates that mitochondria are the main source of ROS in aged cells. According to this theory, mitochondria-derived H₂O₂ diffuses readily through cell membranes and contributes to a variety of macromolecular oxidative modifications. The original working version of this theory invoked accumulation of oxidative damage to proteins, lipids and DNA as the primary causal factor in the aging process (Harman 1972). It was proposed that, over time, oxidative macromolecular damage accumulates because the anti-oxidant and repair systems that prevent or remove the damage are imperfect. The long-term effect of this accumulation of macromolecular damage is the decline in cell function characteristic of aging (Sohal and Brunk 1992; Sohal and Orr 1992, 1998; Ku and Sohal 1993; Sohal et al. 1993; Sohal and Weindruch 1996; Lass et al. 1998; Bayne et al. 2005). As discussed above, mitochondria-derived ROS (mtROS; most importantly, H₂O₂) also act as pro-inflammatory signaling molecules, which have important implications for the aging process. In addition to being a source of ROS, mitochondria also determine cellular energetics, regulate NAD(P)/NAD(P)H levels— and consequently, the activity of several enzymes—and contribute to the induction of apoptosis. Comparative studies addressing mtROS homeostasis, mitochondrial function and mitochondrial stress resistance are especially useful to address predictions of the mitochondrial theory of aging. In this context, previous studies by us and others have revealed that heart, liver, smooth muscle and fibroblast mitochondria of P. leucopus produce less O₂⁻ and H₂O₂ than shorter-living species (Sohal et al. 1993; Brunet-Rossinni 2004; Csiszar et al. 2007c) (Fig. 1). These findings agree with the predictions of the oxidative stress and mitochondrial hypotheses of aging. Our recent studies using the mitochondrial O₂⁻ specific dye MitoSox have demonstrated that mitochondria in P. leucopus fibroblasts produce low amounts of ROS even when stimulated with high glucose (Z. Ungvari and P. Pacher, unpublished observation). The underlying mechanisms for differences in mtROS generation and mitochondrial oxidative stress resistance are likely multifaceted. mtROS generation is affected by mitochondrial number per se, as well as the respiratory substrate, presence or absence of inhibitors of electron transport, mitochondrial membrane potential (ΔΨ) and ΔpH. In addition, differential expression of antioxidant enzymes, which neutralize ROS (e.g., glutathione peroxidase, Mn-SOD, catalase, peroxyredoxin), is likely to contribute to the lower mtROS generation and/or oxidative stress resistance of P. leucopus mitochondria.

Fig. 1.

a Representative histograms of flow cytometry experiments demonstrating that the mean fluorescent intensity of oxidized MitoSox is greater in cultured Mus musculus fibroblasts than in Peromyscus leucopus cells. MitoSox staining was used to measure mitochondrial O₂⁻ generation (for methods, see Mukhopadhyay et al. 2007a, b). Cell debris (low forward and side scatter), dead (Sytox Green and annexin V-positive) and apoptotic cells (annexin V-positive) were excluded from the analysis (Mukhopadhyay et al. 2007a, b). b Summary data for MitoSox fluorescence intensities in cultured human, P. leucopus and M. musculus fibroblasts. Cells were grown in 96-well plates and buildup of MitoSox fluorescence was assessed using a Tecan Infinite M200 fluorescent plate reader as described in Csiszar et al. (2007c). Hoechst 33258 fluorescence representing DNA content/cell mass was used for normalization. Data are means ± SEM (n = 8 in each group), * P < 0.05. vs human, # P < 0.05 vs P. leucopus

The critical role played by mitochondria in programmed cell death is attributed to the opening of the mitochondrial permeability transition (MPT) pore (Zorov et al. 1997; Pacher et al. 2001; Pacher and Hajnoczky 2001; Zorov et al. 2007). There is growing evidence that cytoplasmic Ca²⁺ signaling and mitochondrial Ca²⁺ homeostasis are coupled through highly regulated Ca²⁺ uptake and release processes driven by ΔΨ (Akerman 1978; Pacher et al. 2001, 2002; Pacher and Hajnoczky 2001; Zorov et al. 2007). Cross-talk between cytoplasmic Ca²⁺ signals and mitochondrial Ca²⁺ waves (resulting in MPT) is believed to be important in cellular commitment to apoptosis (Pacher et al. 2001; Pacher and Hajnoczky 2001). It is thought that the capability of mitochondria to withstand stimuli that elicit MPT and thus to maintain ΔΨ and to retain calcium is very important for cellular survival. Recent studies by us and by others have demonstrated that cells from longer-living, relatively small mammalian species (including naked mole rats, bats and P. leucopus; Labinskyy et al. 2006a; Csiszar et al. 2007a, 2007c; Harper et al. 2007; Ungvari et al. 2008) are resistant to the pro-apoptotic effects of multiple stressors. Thus far, however, the link between mitochondrial stress resistance and longevity has not been investigated.

We are currently testing the hypothesis that long-lived animals not only have lower cellular mtROS production, but also exhibit superior mitochondrial stress resistance. Here we report the first results of recent studies comparing stress resistance of mitochondria isolated from the livers of M. musculus and P. leucopus. The comparison was based on physiological responses of mitochondria of these species to sequential loads of Ca²⁺ (in vitro model of Ca²⁺ oscillations). A standard experimental model of supplementing isolated mouse liver mitochondria with optimal (5 mM) amounts of succinate (the respiratory substrate for complex II) in the presence of rotenone (respiratory complex I inhibitor) was used (Krasnikov et al. 1997, 2005; Kuzminova et al. 1998). Our results showed no difference in basal oxygen uptake, Ca²⁺ retention, ΔΨ maintenance or volume change between P. leucopus and M. musculus liver mitochondria (Fig. 2). After mitochondria were added to the buffer and endogenous respiratory substrates were consumed (Fig. 2a), mitochondria were energized with succinate plus rotenone. Under these conditions complete mitochondrial energization occurs almost immediately (Fig. 2c). Oxygen uptake was linear and the three other measured parameters were stable during the experiment (∼40 min). Ca²⁺ retention capacity was tested for both mitochondrial samples using an in vitro model of Ca²⁺ oscillations (sequential additions of Ca²⁺ in the presence of succinate and rotenone). Under these conditions complete mitochondrial energization took ∼3 min (Fig. 3c).

Fig. 2a–d.

Simultaneous measurement of four parameters in isolated mouse liver mitochondria (MLM). Black tracesP. leucopus, red tracesM. musculus (strain C57BL/6). The incubation medium contained 300 mM sucrose, 5 mM HEPES, 1 mM KH₂PO₄, pH 7.4. MLM (1 mg/ml) were energized by addition of 5 mM succinate plus 1 μM rotenone (final concentrations) to the medium according to published methods (Krasnikov et al. 1997, 2005; Kuzminova et al. 1998). a–d O₂ uptake, Ca²⁺ flux, ΔΨ, and swelling (as reflected by changes in optical density), respectively. Decrease in O₂ concentration reflects O₂ uptake by mitochondria. Note stable baseline levels of the measured parameters after addition of succinate in b– d

Fig. 3a–d.

Effects of sequential additions of Ca²⁺ on four parameters measured simultaneously in isolated MLM. Black tracesP. leucopus, red tracesM. musculus (strain C57BL/6). The incubation medium contained 300 mM sucrose, 5 mM Hepes, 1 mM KH₂PO₄, and 5 mM succinate, pH 7.4 according to previously described methods (Krasnikov et al. 1997, 2005; Kuzminova et al. 1998). Addition of MLM (1 mg/ml) was followed by addition of rotenone (1 μM). The initial rate of oxygen uptake (a) was measured in a closed cuvette. To prevent anoxic conditions during the experiment, the cuvette was periodically opened to allow equilibration of the incubation medium with ambient oxygen. At the end of the experiment the cuvette was sealed and respiration rate for uncoupled mitochondrial state recorded. Additions of Ca²⁺ were 20 μM each (final concentration). a–d O₂ uptake, Ca²⁺ flux, ΔΨ, and swelling, respectively. The decrease in O₂ concentration reflects O₂ uptake by mitochondria (a); increase in Ca²⁺ concentration is due to Ca²⁺ efflux from mitochondria (b); increase in TPP⁺ concentration reflects ΔΨ dissipation (c); decrease in absorbance reflects mitochondrial swelling (d)

After mitochondria were maximally energized, sequential additions of Ca²⁺ were introduced to mitochondria every 2 min. There was no difference in the rate of oxygen uptake between P. leucopus and M. musculus mitochondria throughout the experiments (Fig. 3a). However, during pulses of loading with Ca²⁺, substantial differences in Ca²⁺ retention capacity (Fig. 3b), ΔΨ maintenance (Fig. 3c), and mitochondrial volume (Fig. 3d) were observed. As predicted, mitochondria isolated from longer-living P. leucopus exhibited greater resistance to Ca²⁺ stress compared to M. musculus. We also compared ΔΨ changes induced by Ca²⁺ loading for both P. leucopus and M. musculus. Data were collected for each time Ca²⁺ was introduced into the experimental buffer. Because Ca²⁺ uptake into mitochondria is ΔΨ-dependent, Ca²⁺ uptake is accompanied by an initial partial, transient depolarization followed by ΔΨ restoration (Fig. 3c). In our system, ΔΨ is measured by means of a TPP⁺-selective electrode. Redistribution of TPP⁺ across the mitochondrial membrane is directly proportional to ΔΨ. Thus, the initial rate of TPP⁺ uptake by mitochondria reflects ΔΨ restoration. Values of ΔTPP⁺ in μM s⁻¹ plotted against representative amounts of Ca²⁺ loaded into mitochondria are shown in Fig. 4.

Fig. 4.

Dependence of ΔΨ restoration on [Ca²⁺] loading in P. leucopus and M. musculus liver mitochondria. For both species the dependence of ΔΨ restoration on [Ca²⁺] loaded is sigmoidal. Both data sets fit the equation . Coefficients a, b and x₀ for P. leucopus and M. musculus are a = 0.29 and 0.15,  b= −18.0 and −19.7, x₀ = 154 and 83.5, respectively

The dependence of ΔΨ restoration during increased Ca²⁺ loading for both species can be described by similar equations containing three independent variables (a, b and x₀; Fig. 4). The proposed model predicts that maintenance of ΔΨ by the respiratory chain (Y) depends on at least three parameters affecting Ca²⁺ transition through mitochondrial transport systems (coefficients a, b and x₀). An implication of this model is that the effects on Ca²⁺ transition caused by differences in the intensity of H⁺ leaks should be considered. Previous studies have suggested that, during the initial stages of MPT induction (when high-amplitude swelling of mitochondria is not yet detectable), the pore can exist in the H⁺-selective state (Novgorodov and Gudz 1996). This could increase H⁺ leaks, which in turn could lead to reversal of activity of the ΔΨ-dependent Ca²⁺ uniporter and, finally, to collapse of ΔΨ, concomitant high-amplitude swelling, and subsequent mitochondrial failure. In our model, coefficient ‘a’ reflects the rate of ΔΨ restoration after the first addition of Ca²⁺. Coefficient ‘b’ corresponds to the median rate of ΔΨ restoration (below this rate dissipation of ΔΨ begins). Coefficient ‘x₀’ is the median value of the Ca²⁺ corresponding to the maximal amount of Ca²⁺ that mitochondria are capable of retaining before dissipation of ΔΨ starts.

Taken together, our findings suggest that P. leucopus mitochondria exhibit superior resistance to Ca²⁺ loading compared to M. musculus mitochondria. We propose that mitochondria that possess greater Ca²⁺ retention capacity could significantly delay induction of apoptosis under conditions of cellular stress. We propose two working hypotheses to explain our results: (1) P. leucopus mitochondria have more effective Ca²⁺ buffering systems in the matrix; or (2) the MPT of P. leucopus exhibits different functional/structural characteristics from those of M. musculus mitochondria [e.g., relative deficiency of cyclophilin D (CypD)]. In support of the second hypothesis, neurons from CypD-knockout mice are indeed resistant to ROS. Moreover, brain mitochondria, which lack CypD, are capable of retaining considerably more Ca²⁺ compared to mitochondria obtained from wild-type mice (Forte et al. 2007). However, it is also possible that the respiratory chains for both M. musculus and P. leucopus mitochondria possess similar efficiencies and levels of H⁺ leaks, and that the earlier onset of the MPT in M. musculus is due to inferior efficiency of ROS detoxification systems.

In conclusion, our studies indicate that mitochondria from young P. leucopus produce substantially less ROS than young M. musculus mitochondria and exhibit superior resistance to stress. We hypothesize that mitochondrial stress resistance in P. leucopus contributes to the remarkable resistance of P. leucopus cells to oxidative stress-induced apoptosis (Csiszar et al. 2007c). Further studies should address this hypothesis in detail. There is little data available regarding age-related alterations in mitochondrial function and phenotype in P. leucopus. We are aware of only one electron microscopic study, which revealed no age-related structural alterations in P. leucopus mitochondria (King et al. 1982). Thus, further studies are definitely needed to elucidate whether cells of long-lived mice of the genus Peromyscus (especially postmitotic cells) are protected against age-related increases in mtROS generation, impairment of mitochondrial biogenesis, or altered mitochondrial gene expression and mitochondrial functional decline. Also, it should be determined whether mitochondrial stress resistance or cellular resistance to apoptotic stimuli in long-lived mice of the genus Peromyscus is altered in aging.

Repair of oxidative DNA damage

It has been estimated that eukaryotic cells must repair thousands of DNA lesions per day to counteract endogenous sources of DNA damage, such as DNA damage generated by cellular production of ROS/RNS (Lindahl 1993). It is assumed that failure to repair DNA lesions can lead to deleterious mutations, genomic instability, or cell death, thereby accelerating the aging process or causing life-threatening diseases such as cancer. Investigators in earlier studies have proposed a strong correlation between DNA repair efficiency and maximum species lifespan (Hart et al. 1979a). However, most of these earlier studies used relatively simple experimental approaches, and the role of DNA damage accumulation in the aging process and its contribution to species longevity are still a matter of debate. The mechanisms by which DNA is repaired vary greatly depending on the type of DNA lesion, and include repair of double-strand breaks, nucleotide excision repair, and base excision repair (Peterson and Cote 2004). DNA double-strand breaks arise in DNA due to ionizing radiation, chemical exposure or stalled or collapsed DNA replication forks.

Two highly conserved major pathways exist for repairing double-strand breaks: nonhomologous end joining and homologous recombination. The nucleotide excision repair pathway is used for the removal of a variety of DNA-distorting lesions, including UV-induced pyrimidine dimers (Peterson and Cote 2004). The base-excision repair pathway is responsible for repair of oxidized and alkylated DNA bases, as well as abasic sites generated by spontaneous depurination (Peterson and Cote 2004). The pathophysiological consequences of each type of damage are different (e.g., DNA lesions that are substrates for base excision repair are highly mutagenic). Therefore, interspecies differences in efficiencies of repair systems for individual damage components likely contribute to different aging and pathophysiological phenotypes (e.g., cancer, apoptotic cell loss, cellular dysfunction). In-depth comparative studies could elucidate interspecies differences in the various steps of repair pathways and the enzymatic machinery that facilitates access to chromatin. Below, we summarize the available information on DNA repair mechanisms in P. leucopus and provide a perspective on this field.

Previous data from the Hart laboratory suggested that the rate of accumulation of age-related DNA damage (single-strand breaks) in liver and kidney of P. leucopus is slower than in house mice (Su et al. 1984). Hart and coworkers selectively assessed the efficiency of repair for individual DNA damage components. In one set of experiments fibroblasts from various short- and long-lived species (from mice to Homo sapiens) were exposed to UV irradiation. The ability of fibroblasts to perform unscheduled DNA synthesis after UV irradiation (a measure of global nucleotide excision-repair) was measured autoradiographically (Hart and Setlow 1974). Using this method, a positive correlation between life-span and efficiency of repair of UV-induced DNA damage was proposed. Subsequent studies using the same methods to test this hypothesis using the M. musculus–P. leucopus longevity contrast pair yielded results in full agreement with this hypothesis (Hart et al. 1979b).

We are currently comparing relative rates of DNA repair in fibroblast cell lines from various long- and short-lived mammalian species after UV treatment, γ-irradiation or oxidant challenge using the single-cell gel electrophoresis ("comet") assay, a new, simple and sensitive method of evaluating DNA damage and repair in individual cells. We recently revisited the question of nucleotide excision-repair in P. leucopus using this technique. Interestingly, P. leucopus fibroblasts exhibited a greater extent of DNA damage immediately post-UV_{254 nm} treatment, and nucleotide excision-repair seemed to be less effective (Fig. 5b) than in M. musculus cells. In contrast, our results so far indicate that cells of P. leucopus exhibit less H₂O₂-induced DNA damage than cells of shorter-lived species, including M. musculus (Csiszar et al. 2007c). A substantial portion of H₂O₂-induced DNA damage is thought to be due to oxidants generated from iron-mediated Fenton reactions, and there appear to be at least two distinguishable classes of iron-mediated Fenton oxidants of DNA. However, some current studies argue against the importance of the Fenton reaction in this context (Ischiropoulos and Beckman 2003). Differences in the ability of H₂O₂ to cause DNA damage in M. musculus and P. leucopus therefore may reflect differences in free-radical detoxification systems or chromatin structure differences rendering it less accessible to H₂O₂. At present it is not clear how DNA repair efficiency compares in M. musculus and P. leucopus cells after H₂O₂ treatment. It is also unknown how repair of peroxynitrite (ONOO⁻)-induced DNA damage differs between the two species.

Fig. 5a,b.

DNA repair in P. leucopus. a Animals were γ-irradiated (6 Gy) and DNA damage assessed in peripheral blood mononuclear cells (PBMC) by single-cell electrophoresis (“comet assay”) (Ungvari et al. 2007a) at 30 min, 1 and 2 h post-irradiation. Damaged DNA migrates during electrophoresis from the nucleus towards the anode, forming the shape of a “comet” with a head (cell nucleus with intact DNA) and a tail (relaxed and broken DNA). The frequency distribution of tail DNA content was obtained (median values shown; P < 0.05). b Primary mouse and P. leucopus fibroblasts were treated with UV_{254 nm} (10 mJ/m²) and the extent of DNA damage assessed by comet assay at 0.5, 1, 2, and 3 h post-irradiation (Csiszar et al. 2007c). Median values of tail DNA content at each time point are shown

It is thought that the spectrum of damages due to H₂O₂ is similar to (but not congruent with) that caused by ionizing radiation—that is, mainly base-excision repair. With this in mind, the second objective of our ongoing studies is to investigate the effect of in vivo γ-irradiation on DNA damage in M. musculus and P. leucopus peripheral blood leucocytes. P. leucopus cells exhibited less DNA damage and greater efficiency of repair in response to the same dose of γ irradiation (6 Gy; Fig. 5). Most of the DNA-damaging effects of ionizing radiation are induced by •OH radicals (indirect effects) and by one-electron oxidation (direct effects). It is unknown which of these mechanisms is more important for the observed interspecific differences in cellular sensitivity to γ radiation. We suggest that the differences in the extent of DNA damage immediately post-irradiation likely reflect differences in the nuclear DNA organization or antioxidant defenses, since irradiation was performed at 0°C, which effectively slows enzymatic repair processes.

In conclusion, DNA repair is an important mechanism by which cells maintain genomic integrity, and the efficiency of DNA repair pathways may contribute to interspecies differences in both the aging process and longevity. Because ROS have been implicated in cancer and age-related degenerative diseases, the link between decline in DNA repair capacity and/or defects in repair factors, late-life diseases and accelerated aging in mammals is an active area of interest. The M. musculus–Peromyscus longevity contrast pair is an exceptionally good model for studying association between longevity and DNA repair efficiency, mitochondrial stress resistance and ROS-detoxification pathways. We hope that this review will stimulate interest among biologists in long-lived mice of the genus Peromyscus as study organisms and in a comparative approach to aging research.

Abbreviations

CR

Caloric restriction

AGE

Advanced glycosylation end-product

TNFα

Tumor necrosis factor-alpha

IL-6

Interleukin-6

iNOS

Inducible nitric oxide synthase

ROS

Reactive oxygen species

mtROS

Mitochondrial reactive oxygen species

RNS

Reactive nitrogen species

NF-κB

Nuclear factor-kappa B

Mn-SOD

Manganese superoxide dismutase

TPP

Tetraphenyl phosphonium

CypD

Cyclophilin D

ONOO-

Peroxynitrite

•OH

Hydroxyl radical

SEM

Standard error of the mean

MLM

Mouse liver mitochondria

MPT

Mitochondrial permeability transition