|Home | About | Journals | Submit | Contact Us | Français|
Fibroblasts from long-lived mutant mice are resistant to many forms of lethal injury as well as to the metabolic effects of rotenone and low-glucose media. Here we evaluated fibroblasts from young adult naked mole-rats (NMR; Heterocephalus glaber), a rodent species in which maximal longevity exceeds 28 years. Compared to mouse cells NMR cells were resistant to cadmium, MMS, paraquat, heat, and low glucose media, consistent with the idea that cellular resistance to stress may contribute to disease resistance and longevity. Surprisingly, NMR cells were more sensitive than mouse cells to H2O2, UV light, and rotenone. NMR cells, like cells from Snell dwarf mice, were more sensitive to tunicamycin and thapsigargin, which interfere with the function of the endoplasmic reticulum (ER stress). The sensitivity of both Snell dwarf and NMR cells to ER stress suggests that alterations in the unfolded protein response might modulate cell survival and aging rate.
The natural variation of lifespan among the species of mammals may prove to be a valuable resource for understanding the molecular mechanisms involved in regulating the aging process. Among mammals, maximum lifespan can differ by over 100 years, and even within a single order the longest-lived rodents can in some cases live more than 10-fold longer than the shortest-lived (1–3). These differences in mammalian longevity are much greater than those that can be created through single-gene mutations or dietary restriction (4). Species and mutant strains within species with exceptionally long lifespan tend to exhibit delayed rates of age-rated physical decline in multiple organ systems (2;5). Therefore, studies examining differences in the physiology and cellular biology of species that vary in their maximum lifespan may provide insights into the mechanisms that modulate the rate of age-related physiological decline.
It has long been proposed that the mechanisms that regulate the aging process may be similar to those that regulate resistance to stress (6;7). Further support for this idea has come from experimental evidence that single gene mutations that extend lifespan in the nematode C. elegans typically render these animals resistant to numerous forms of lethal cellular insults, including oxidative stress, UV light, heavy metals and heat (8–11). To a very limited extent, similar results have been seen in mammals; mutations in the IGF-I signaling pathway that extend mouse longevity also tend to enhance survival of mice injected with paraquat, an agent that induces oxidative stress (12–14).
In our own laboratory, we have found that fibroblasts derived from the long-lived Snell dwarf (Pit1dw/dw), Ames dwarf (Prop1dw/dw) and growth hormone receptor knockout (GHR-KO) mice are resistant to cell death induced by multiple agents including heat, external and internal oxidative stressors (specifically H2O2 and paraquat), heavy metal, UV light and the DNA alkylating agent methyl methanesulfonate (MMS) (15;16). All three of these mutant mouse lines share a phenotype of very low circulating levels of IGF-I as adults and a lifespan extension of 40–60% (17–19). Cells derived from newborn Snell dwarf mice show little or no difference in stress resistance in comparison to cells from newborn controls, suggesting that the level of fibroblast stress resistance is regulated by post-natal factors, presumably hormonal, in these endocrine-deficient mice (16). Snell dwarf-derived fibroblasts also exhibit metabolic abnormalities independent elicited by exposure to non-lethal forms of stress. For example, Snell dwarf-derived fibroblasts are relatively resistant to the metabolic effects of depleting glucose from the growth media and to incubation with the mitochondrial inhibitor rotenone (20). While changes in stress resistance of skin-derived fibroblasts are unlikely to play an important causal role in the control of mouse longevity, these in vitro data suggest a model in which abnormalities of the hormonal milieu during development may lead to augmented stress resistance in a variety of cell types from long-lived mutant mice and thus contribute to the unusual delayed pathology of multiple forms of late-life illnesses in these animals (18;21;22).
Tests of stress resistance of fibroblasts derived from the skin of adult donors have also shown that maximum species longevity is correlated with cellular resistance to some forms of stress and some forms of metabolic inhibition when evaluated in multiple long- and short-lived species of rodents (23). In that study, we found that fibroblasts from long-lived species tended to be more resistant to the lethal effects of cadmium, H2O2, heat and MMS and also more resistant to the metabolic effects of glucose withdrawal and treatment with rotenone. These data suggest that in vitro studies of cellular stress resistance might help to evaluate the hypothesis that shared mechanisms contribute to the regulation of longevity across the mammalian evolutionary tree.
Prompted by the results of the rodent comparison, we have now tested to see if cells from the longest-living rodents known, naked mole-rats (NMR; (24)), are resistant to various forms of in vitro stress. The NMR (Heterocephalus glaber) has been observed to live to ages of at least 28 years of age in captivity, about seven times longer than the maximum captive lifespan of laboratory mice (5;24). These animals are roughly the same size as mice, yet live about ten times longer than predicted by linear regression of lifespan against size for non-volant eutherians (i.e. those that neither fly nor glide) (25;26). NMR show little physiological or reproductive decline even when approaching 30 years of age, have never been observed to develop any spontaneous neoplasm, and do not show the typical age-associated acceleration in mortality risk that characterizes nearly every other species for which detailed survival data are available (5;27;28). Thus, physiological and biochemical mechanisms have evolved in this species that can dramatically extend lifespan compared to rodents of comparable size. As such, we thought it of interest to learn if cells from NMR, like those of long-lived mutant mice, are resistant to the effects of cellular stress. Fibroblast cell lines were derived from the skin of young adult NMR to test the hypothesis that the long lifespan of these animals is accompanied by cellular resistance to stress. Our results show that cells from these animals are, as predicted, resistant to some agents, but are, unexpectedly, quite sensitive to others.
NMR skin samples were obtained from a captive colony maintained in the Department of Biology at City College of New York. The husbandry conditions of these animals were as presented previously (5;28–32). Mouse skin samples were prepared from the genetically heterogeneous DC stock. DC mice are the intercrossed progeny of UM-HET3 mice, themselves generated by a cross between females of the (BALB/c × C57BL/6)F1 stock and males of the (DBA/2 × C3H/He)F1 stock. Biopsies for this study were taken from the G4 generation of DC mice. Skin samples were taken from male NMR of approximately 2 years of age and from male DC mice of 3–4 months of age. These ages were chosen as representatives of a physiologically equivalent age representing post-pubertal young adults of each species (33); 2 year old NMR have reached an age of about 7% of the species’ maximal lifespan and 3 month old mice have reached an age of about 6% of the species’ maximal lifespan.
Snell dwarf (Pitdw/dw) mice, and heterozygote (Pitdw/+) controls were bred as the progeny of (DW/J × C3H/HeJ)-dw/+ females and (DW/J × C3H/HeJ)F1-dw/dw males. Their sires had been previously treated with growth hormone and thyroxine to increase body size and fertility. Tail skin biopsies were taken from male mice at approximately 3 months of age.
Tail from NMRs were sterilized with 70% ethanol wipes and biopsies of 3–5 mm in length were obtained from the distal half of the tail and placed in complete medium (CM) made of Dulbecco’s modified Eagle medium (DMEM, high-glucose variant, Gibco-Invitrogen, Carlsbad, CA) supplemented with 10% heat-inactivated fetal bovine serum, antibiotics (100 U/ml penicillin and 100 µg/ml streptomycin; Sigma, St. Louis, MO) and 0.25 μg/ml of fungizone (Biowhittaker-Cambrex Life Sciences, Walkersville, MD) on ice and shipped overnight from New York to Michigan. For laboratory-raised DC mice and Snell dwarf and controls, tails were washed with 70% ethanol, and tail-skin biopsies of 3–5 mm in length obtained from the distal half of the tail were placed in CM prior to fibroblast isolation.
Fibroblast cultures from NMR and DC biopsies were developed using methods similar to those previously reported (15;34) except that the present study used 10% CO2 in air and an incubator temperature of 33° C. CM described above was used to subculture cells from both NMR and mice. We, like others (35), have found that fibroblasts from NMR grow poorly at the standard 37° C incubation temperature, perhaps because the animals have a body temperature several degrees below this in their natural environment (36). For this reason, both NMR and DC fibroblast cultures were incubated at 33° C. Initial cultures of cells derived from NMRs (“P0”) grew to confluence at a slower rate than those of DC-derived cells (not shown) so all cells were trypsinized for subculturing when NMR cells reached approximately 90% confluence. Cell cultures were fed at day 3 (replacing 2/3rds of the medium), again at day 6 and subcultured at Day 10 at a density of 7.5 × 105 in tissue culture flasks of 75 cm2 surface area to produce P1 cultures. This protocol was retained for all subsequent subculturing. Following an additional passage using the same protocol, P3 cells were harvested and cryopreserved as in previous studies (23). These cryopreserved cells were stored in liquid N2 for up to 24 months before assessment of stress resistance. Cell aliquots were thawed and grown two further passages as previously shown, with all incubations at 33° C and 10% CO2 in air (23).
NMR and DC cells lines were evaluated in two sets of experiments, each of which evaluated equal numbers of NMR and DC donors. P4 cultures were used to assess resistance to all agents. Tests for resistance to the lethal stresses cadmium, H2O2, paraquat, UV light, MMS and heat were performed as previously described (15;23;34;37). After treatment, cells were washed at 33° C with 1X PBS, and incubated at 33° C in serum-free DMEM supplemented with 2% BSA, antibiotics and fungizone; survival was measured 18 hours later using conversion of the extracellular tetrazolium dye WST-1 to its colored formazan product (15;23;34;37). All incubations (except for heat) were at 33° C in a humidified incubator with 10% CO2 in air.
For each set of glucose or rotenone assay experiments, representatives of P4 cell lines from both naked mole-rats and DC mice were assayed in parallel to minimize the effects of day to day variation. Tests for the effect of glucose withdrawal and rotenone were performed as previously shown (20;23). All incubations were at 33° C in a humidified incubator with 10% CO2 in air. It is important to note that although WST-1 is used in tests of both lethal stresses and in the evaluation of responses to low glucose, the decline in WST-1 reduction under low glucose conditions or at the rotenone concentrations employed is not accompanied by cell death (20).
For Snell dwarf and littermate control experiments, P3 cells were prepared in 96 well plates as previously described. These cells had not been previously cryopreserved and all assays with them were performed in a humidified incubator with 5% CO2 in air and at a temperature of 37° C. Thapsigargin and tunicamycin were suspended in a solution of DMSO (Sigma) and doses of these agents were diluted in DMEM with a final DMSO concentration <0.5%. Cells were treated for 24 hours after which they were washed with 1X PBS and incubated with serum-free DMEM supplemented with 2% BSA, antibiotics and fungizone; survival was measured 18 hours later using WST-1. DC and NMR assays for resistance to these agents were performed similarly, except all incubations and treatments were at 33° C in a humidified incubator with 10% CO2 in air.
For calculation of the resistance of each cell line to chemical stressors, at each dose of chemical stressor, mean survival was calculated for duplicate wells for each cell line. The LD50, i.e. dose of stress agent that led to survival of 50% of the cells, was then calculated using probit analysis as implemented in NCSS software (NCSS, Kaysville, UT). For this analysis, extremely low doses of stress agents that caused no cell death in fibroblasts, as measured by the WST1 assay, were censored from all data sets. ED50 values for glucose withdrawal and rotenone treatments were calculated in a similar manner to estimate the level of glucose or rotenone associated with a 50% reduction in cellular metabolic activity. Because some groups of data were not normally distributed, differences between groups in mean LD50 levels were evaluated by Mann-Whitney U non-parametric tests, with each day’s work containing equal numbers of cultures from each species or each genotype. For some agents, the day of assay seemed to affect the calculated LD50 or ED50, therefore data were also analyzed by two way ANOVA with species and day of assay as the factors.
To see if fibroblasts from NMR resembled cells from other long-lived rodent species or from long-lived mutant mice, we evaluated their resistance to a series of lethal stresses and non-lethal inhibitors of cellular metabolism. Skin fibroblasts were derived in each case from young adult animals, in which approximately 5% to 10% of each species’s maximal lifespan had been completed, and cell cultures for DC and NMR cultures were maintained at 33° C, a temperature at which NMR cells proliferate and survive.
The data in Figure 1 and Table 1 represent the results from cell lines tested for their resistance to the lethal agents cadmium, paraquat, heat, MMS, H2O2 and UV light. Our working hypothesis, based on the longer lifespan of NMR, was that fibroblasts from NMR would be relatively resistant to all these agents, similar to cells from long-lived Snell dwarf mice (15;34). Our data supported this notion for experiments assaying resistance to cadmium (Figure 1A), to paraquat (Figure 1B), to heat (Figure 1C) and to MMS (Figure 1D) in cell lines from 11 individual animals of each species. The difference in resistance to lethal stressors ranged from 269% for paraquat to 23% for MMS. These differences reached statistical significance by non-parametric Mann-Whitney test with a p < 0.04 in each case.
In contrast, however, we found that NMR fibroblasts were markedly more sensitive than DC mouse fibroblasts to H2O2 (approximately 4-fold more sensitive, Figure 1E) and to irradiation with UVC light (approximately 3-fold more sensitive, Figure 1F). In both cases, these differences reached significance by Mann-Whitney U test with p < 0.001. ANOVA calculations that adjust for day-to-day variations, as compiled in Table 1, confirmed the conclusion that NMR fibroblasts were more resistant to cadmium, paraquat, heat, and MMS, and more sensitive than mouse cells to H2O2 and UV light, with p < 0.01 in each case.
The data in Figure 2 (and also included in Table 1) represent the results from cell lines tested for their resistance to the metabolic inhibition caused by withdrawal of glucose from the growth medium or by the mitochondrial inhibitor rotenone. Fibroblasts from the NMR, like Snell dwarf cells and cells from the long-lived rodents previously tested, were relatively resistant to the effect of glucose withdrawal from the growth medium (Figure 2A, Mann-Whitney U test p = 0.01). Surprisingly, however, NMR-derived fibroblasts were approximately 2-fold more sensitive than DC fibroblasts to rotenone, at the non-lethal doses used for these tests (Figure 2B, Mann-Whitney U test p = 0.003). An ANOVA approach confirmed both inferences (Table 1).
The LD50 and ED50 values calculated for the resistance of mouse cells to most agents were similar to those noted in previous studies which had used cultures grown at 37° C, except for somewhat increased resistance to cadmium and rotenone (15;20;23;34;37). In a set of experiments not shown, cell lines from DC mice were assayed in parallel under growth conditions of 33° or 37° C, with 10% CO2 in air. The 4° temperature decrease alone resulted in a mean LD50 and ED50 of approximately 10-fold higher for cadmium and rotenone resistance, and 2-fold increase for resistance to peroxide; there was no difference in resistance to UV light.
It is speculated that some of the agents used in this and our other studies might partly act by inducing endoplasmic reticulum (ER) stress (38–40), a term coined to signify an imbalance between cellular demand for ER function and ER capacity (41). As a prelude to assessing NMR cells for resistance to ER stress agents, we compared the lethal effects of thapsigargin and tunicamycin on fibroblasts from Snell dwarf mice and their littermate controls. Thapsigargin disrupts ER Ca2+ homeostasis through irreversible inhibition of the endoplasmic reticulum Ca2+-ATPase (42), which in turn interferes with the function of Ca2+-sensitive chaperones such as calreticulin and ERp57, and hampers the post-translational processing, folding and export of proteins from the ER (43). Tunicamycin abolishes N-linked glycosylation of proteins in eukaryotic cells by blocking the function of N-acetylglucosamine-1-transferase, and thus interferes with the assembly and transport of glycoproteins from the ER to the Golgi complex (44;45). Figure 3 shows the results of six experiments involving cell lines derived from 8 or 9 pairs of Snell dwarf and control mice, showing that, in contrast to their relative resistance to most sources of lethal injury, Snell fibroblasts are significantly more sensitive than cells from littermate controls to cell death induced by thapsigargin (Figure 3A, Mann-Whitney U test p = 0.002) or tunicamycin (Figure 3B, Mann-Whitney U test p = 0.03). To see if NMR cells were, like Snell dwarf fibroblasts, unusually sensitive to agents that induced death via ER stress, we evaluated six pairs of NMR and mouse DC cultures. Figure 4 shows that NMR cells were indeed significantly more sensitive to both thapsigargin (Figure 4A, Mann-Whitney U test p = 0.01) and tunicamycin (Figure 4B, Mann-Whitney U test p = 0.004) compared to cells from DC mice tested in parallel. These data are also included in Table 1.
Table 2 presents a summary of these stress resistance data, together with published data, for comparison, on cells from long-lived mutant mice and from species of long-lived rodents.
In this study, we tested the hypothesis that fibroblasts derived from the skin of the longest-lived rodent, the NMR, exhibit enhanced resistance to lethal and non-lethal cellular stresses as has been reported for cell lines from long-lived mutant mice (15;20;34;37) and for cell lines from long-lived rodent species (23). We found that NMR-derived cells were indeed significantly more resistant than mouse cells to some forms of stress, i.e. cadmium, paraquat, heat, and MMS, and relatively resistant to the effects of low glucose concentrations. Surprisingly, however, we found that cells from NMRs were significantly more susceptible than mouse cells to the effects of H2O2, UV light and rotenone. This unexpected pattern of differential resistance raises a number of questions about the extent to which regulated responses to various forms of stress are coordinated, and provides some insights into the potential role of stress resistance in the evolution of NMR longevity.
Interspecies comparisons can sometimes be confounded by differences in body size and the plethora of parameters that scale with size, such as metabolic rate. Larger species of mammals tend to live longer, so that almost any trait that is linked to body size is likely to show a correlation with lifespan (26;46). Interpretation of NMR results, however, is not confounded in this manner, because adult NMR (~35 g) are similar to mice in size (~40 g for adult DC mice). Our design also focused on comparisons between cells taken from mice and NMRs that were at equivalent stages of development and at physiological equivalent ages, i.e. from young adult animals that had reached full maturity and had lived ~6% of their lifespan. At this physiologically age (3–4 months for mice and ~2 years for NMRs), neither mice nor NMR have yet begun to show any of the signs of aging, including increased mortality risk (5).
Our previous work has shown that fibroblast resistance to the lethal effects of cadmium, heat, and MMS toxicity and to the metabolic effects of glucose withdrawal from growth medium show significant or suggestive correlations with maximum longevity of the donor species (23). Fibroblasts from NMR are relatively resistant to each of these agents and are also resistant to paraquat. These data suggest that there may be pathways that regulate the resistance to these agents, and that the level of activity could vary with maximum lifespan among rodent species. These same hypothetical pathways may be those that produce resistance to these forms of stress in cells from Snell and Ames dwarf mice and from GHR-KO mice. Studies of cell lines from normal genetically heterogeneous mice have shown that those individual mice whose skin-derived fibroblasts are resistant to cadmium are also resistant to low glucose, again consistent with the idea that resistance to these two forms of stress is co-regulated, perhaps involving effects of polymorphic loci (20).
Many experts have proposed that cellular damage due to oxidative stress contributes to many of the diseases of aging, and that resistance to oxidative damage is a key element in extended longevity within or across species (47;48). Studies similar to those presented here have shown that primary fibroblasts from long-lived animals tend to be resistant to peroxide damage in comparisons of mammals across many taxonomic orders (49), within the rodents alone (23) and in long-lived mutant mouse strains (15;34). Some tissues from the NMR also seem to be resistant to oxidative injury, including arterial endothelial tissue and mitochondrial membranes from the liver and muscle (28;32). Our results showing that fibroblasts from NMR are resistant to cell death induced by paraquat are consistent with this model (Figure 1B). It is thought that paraquat causes acute cellular toxicity through the production of superoxide radicals in the mitochondria, which in turn result in disruption of NADPH-requiring biochemical processes (50;51). In contrast, however, NMR cells were dramatically more sensitive than mouse cells to the lethal effects of H2O2, and are indeed more sensitive to this agent than any other rodent we have previously studied (23). This result is provocative because these results do not support most models of oxidative stress and longevity and its basis may lie in the particulars of both species-specific oxidative damage and the conditions under which our assays are performed.
The toxicity of H2O2 differs from that of paraquat in that H2O2 causes oxidative stress through the production of hydroxyl radicals that damage the integrity of membranes, lipids and nucleic acids (52), conditions that may be exacerbated by the serum-free conditions of our assay (53;54). Recent studies have shown higher levels of oxidative damage to proteins and nucleic acids and particularly to lipids in tissues from NMR compared to mice of the same chronological or physiological age (29–31). If NMR-derived fibroblasts exhibit a similar phenotype, these cells may be sensitive to the effects of further oxidative damage, such as that caused by H2O2. In theory, this sensitivity could be distinct from the actions of paraquat and other oxidative stressors upon NADPH-dependent processes. These results may also represent our use of standard culture conditions of atmospheric oxygen levels, approximately 20% in air, which are higher than typical tissue oxygen levels (55) and could potentially induce either oxidative damage, or compensatory responses to damage, that would not have been seen in cells grown under lower oxygen tension (56;57). Alterations of the oxygen radicals produced by cellular metabolism due to cell culture at a lower temperature might affect the stress sensitivity of fibroblasts.
NMRs exhibit many characteristics that are contrary to those predicted by the oxidative stress theory of aging in addition the relatively high levels of oxidative molecular damage just mentioned. For example, mice and NMR show similar levels of reactive oxygen species production in heart mitochondria (A.J. Lambert, R. Buffenstein, unpublished) despite their disparities in maximum longevity. Similarly, NMRs have activity levels in the liver of the antioxidants superoxide dismutase, catalase, and glutathione that are similar to those seen in mice (29–31). In addition, glutathione peroxidase activity is much lower in NMR livers compared to that observed in livers from mice (29). Peroxide treatment in vitro induces lipid peroxidation and DNA oxidation that cause cell death (58), effects that can be modulated in part by the activity of glutathione peroxidase (59). It is not known if NMR fibroblasts have lower levels of superoxide dismuatase, catalase, and glutathione peroxidase than do mouse cells, but such differences might in principle contribute to the sensitivity of these cells to peroxide exposure in vitro. It seems clear from the NMR fibroblast data that resistance to H2O2 and resistance to paraquat are likely to be regulated by different defense mechanisms. Our data support the general conclusion that the exceptional lifespan of NMR is not likely to reflect merely a dramatic resistance to all forms of oxidative damage (29–31).
Loss of cellular resistance to H2O2 and UV light may reflect evolutionary adaptations to the subterranean niche of the NMR in its natural environment, where the gaseous atmosphere within the burrows is both hypoxic and hypercapnic (60). Fossil records reveal that Bathyergid ancestors have inhabited a subterranean milieu since the early Miocene (ca 24 million years ago (61). These animals have evolved a set of morphological and physiological phenotypes well suited to life underground, such as a markedly atrophied visual cortex and expanded somatosensory cortex, in keeping with sensory needs in the dark, and their skin is poorly pigmented (62). Physiological adaptation has included low rates of gas exchange and heat production leading to low body temperatures, as well as tolerance of hypoxia and hypercapnia (63;64). Adaptation to a hypoxic environment also may have driven the loss of cellular defenses to peroxide discussed above. Similarly, sensitivity to UV light may result from evolving in an underground environment in which sunlight-induced damage rarely occurs. The repair of UV genomic lesions is closely correlated with survival of cells following irradiation and occurs primarily through nucleotide excision repair, the rate of which is thought to be limited by the recognition of genomic lesions (65;66). Nucleotide excision repair (NER) also recognizes lesions caused by agents other than UV light, and it has been shown that cells that have lost the ability to repair UV-induced damage nevertheless can still repair these other lesions with varying levels of efficiency (67). It would be of interest to test if skin-derived NMR cells, which we show here are sensitive to UV light, are able to resist the toxic effects of other DNA lesions repaired by NER such as the interstrand cross-links induced by cisplatin. It would also be of interest to test if cells derived from NMR internal organs display sensitivity to UV light compared to those from other rodents. Alternatively, it might be possible that the UV sensitivity of NMR-derived cells could contribute to the remarkably low rate of spontaneous cancer in these animals (5); enhanced apoptosis associated with cells with damaged genomes might inhibit the development and progression of neoplasia (68).
NMR-specific cellular responses to some of these agents may result from altered cellular metabolism. Cells from poikilothermic reptiles tend to exhibit small inner mitochondrial surface area and greater membrane permeability of the inner mitochondria compared to mammals and both traits are associated with decreased mitochondrial activity (69;70). Because NMRs are thermally labile mammals outside the warm confines of their thermally buffered equatorial burrows, they may exhibit mitochondrial phenotypes similar to those seen in other poikilotherms in other classes of vertebrates, and this may result in decreased metabolic activity and metabolite levels. The non-lethal effects measured by the rotenone assay may be regulated by cues from the mitochondria, including the levels of metabolic products such as NADH and NADPH (20), and thus related to metabolic activity. Diminished activity of mitochondria may also contribute to resistance to the toxicity of paraquat; the mechanism of paraquat-mediated cell death is most often attributed to its induction of superoxide formation in the mitochondria (50). Investigation into the nature of mitochondria in these animals may prove useful in understanding the relationship between the resistance (or sensitivity) to these agents and longevity.
Phylogenetic history may also contribute to the unexpected sensitivity of NMR cells to H2O2 and UV light. NMRs are members of the rodent suborder Hystricognathi and are more closely related to other rock rats, cane rats, guinea pigs and porcupines than to mice, rats and other rodents in other suborders (5). Our previous comparison of cells from multiple rodent species included only a single species of hystricognath, the North American porcupine (23). Although the data set as a whole showed a correlation between peroxide resistance and longevity, the average resistance of cells from the long-lived porcupine (Mean ± SE, 41 µM ± 15) was not different from those from short-lived DC mice (47 µM ± 12, t-test p = 0.8) or from cells derived from wild-caught mice (55 µM ± 14, t-test p = 0.5). Further, cells from porcupines (47 J/m2 ± 34) tended to be sensitive to UV light compared to cells from DC mice (57 J/m2 ± 12, t-test p = 0.8) and were significantly more sensitive than cells from wild-caught mice (147 J/m2 ± 28, t-test p = 0.04). Glutathione peroxidase activity has been shown to be low in tissues from the hystricognath guinea pig and NMR compared to mice and rats (29;71–74) which might contribute to a relative sensitivity to peroxide among the members of the suborder. It is plausible that these two suborders may have diverged, early in their history, in the biochemical factors by which they respond to various forms of cell injury. It would thus be of interest to evaluate stress resistance of cells derived from hystricognaths whose lifespan was shorter than NMR and porcupines.
As part of this study, the compounds thapsigargin and tunicamycin were used to examine the effect of ER stress on fibroblasts from dwarf mice and NMR. Both of these agents act to hamper the transport of proteins from the ER, thapsigargin by disrupting ER Ca2+ homeostasis and tunicamycin through abolishing N-linked glycosylation of proteins (Rutkowski and Kaufman, 2004). To our surprise, fibroblasts from both dwarf mice and NMR proved to be more sensitive to the induction of ER stress than their respective controls. The accumulation of unfolded or misfolded proteins in the endoplasmic reticulum causes ER stress, and instigates a concerted adaptive program known as the unfolded protein response (UPR). The UPR alleviates ER stress by down-regulating the synthesis of secreted proteins, up-regulating ER chaperone and foldase expression levels, and activating ER-associated degradation (ERAD), thereby easing the burden on the stressed ER by reducing its protein load, increasing its folding capacity, and eliminating irreparably misfolded proteins (41;75;76). In higher eukaryotes, PERK, a double-stranded RNA-activated protein kinase-like ER kinase is the proximal transducer of protein synthesis inhibition, whereas ATF6, a basic leucine zipper transcription factor, and IRE1, a protein kinase/endoribonuclease orchestrate the up-regulation of chaperones and foldases, and activation of ERAD. If these pro-survival efforts are overwhelmed, ER stress-related apoptosis ensues.
Recent studies tend to implicate ER stress in the pathogenesis of many late-life illnesses (41;77). It has also been contended that the ER stress response might affect the aging process itself (78). So far, the C. elegans model provides the most compelling evidence in this respect (79). Treatment of C. elegans with the small molecule resveratrol extends life span in a daf-16-independent manner. Resveratrol induces the transcription of abu-11 (activated in blocked UPR), a member of a gene family with a role in the ER stress response. RNA interference data have shown that the resveratrol effect requires abu-11, and that worms overexpressing abu-11 displayed long lifespan. Furthermore, age has been shown to modulate certain effectors of the UPR. A decline with age in eIF2α phosphorylation together with an increase in GADD34 levels (80), as well as an increase in CHOP and phospho-JNK levels (80;81) in rat livers; and the carbonylation of BiP, protein disulfide isomerase (PDI), and calreticulin in livers of aged mice (82) are cases in point. Our data show that dwarf and NMR fibroblasts are sensitive to the lethal effects of ER stress. It is possible that this sensitivity, if it occurred in appropriate cells in vivo, might help to rid the organism of defective cells and thus postpone pathological outcomes. It is also possible that heightened sensitivity to ER stress might induce compensatory increases in pathways that protect against the lethal effects of some of the other agents tested in our in vitro analyses. More information on the molecular basis for sensitivity of dwarf and NMR cells to thapsigargin and tunicamycin, and on the relationship of the UPR to cellular responses to other forms of stress, is thus likely to be informative.
Understanding what mechanisms Nature uses to create long-lived species is a fundamental problem in biogerontology. Overall, the data presented here contribute to the growing evidence associating unusual longevity with resistance to cellular injury, both within and across species. Disparities among the models shown in Table 2, however, suggest that the specific pattern of cellular resistance may well vary from clade to clade. Developing methods to extend these analyses to other cell types, in vitro and in intact animals, is likely to provide additional insight into these relationships. Our data further illustrate the value of using non-traditional models, such as the NMR, for tests of specific hypotheses about the control of aging.
We are grateful to Maggie Lauderdale and Jessica Sewald for development of the DC mouse stock and the Snell dwarf colony, to James Harper for allowing us to use his unpublished data on Ames dwarf and GHR-KO cells in Table 2, and to Andrzej Bartke, Michal Masternak, and John Kopchick for providing Ames and GHR-KO skin samples. Yael Edrey, Mario Pinto, Roxanne Cheung, Ting Yang and the animal care staff at CCNY are sincerely thanked for their assistance with the NMR husbandry. This work was supported by NIA grants R01-AG022891, T32-AG000114, AG023122 and NIGMS S06-GM08168.