Chemically induced carcinogenesis in rodent models of aging: assessing organismal resilience to genotoxic stressors in geroscience research

Abstract

There is significant overlap between the cellular and molecular mechanisms of aging and pathways contributing to carcinogenesis, including the role of genome maintenance pathways. In the field of geroscience analysis of novel genetic mouse models with either a shortened, or an extended, lifespan provides a unique opportunity to evaluate the synergistic roles of longevity assurance pathways in cancer resistance and regulation of lifespan and to develop novel targets for interventions that both delay aging and prevent carcinogenesis. There is a growing need for robust assays to assess the susceptibility of cancer in these models. The present review focuses on a well-characterized method frequently used in cancer research, which can be adapted to study resilience to genotoxic stress and susceptibility to genotoxic stress-induced carcinogenesis in geroscience research namely, chemical carcinogenesis induced by treatment with 7,12-dimethylbenz(a)anthracene (DMBA). Recent progress in understanding how longer-living mice may achieve resistance to chemical carcinogenesis and how these pathways are modulated by anti-aging interventions is reviewed. Strain-specific differences in sensitivity to DMBA-induced carcinogenesis are also explored and contrasted with mouse lifespan. The clinical relevance of inhibition of DMBA-induced carcinogenesis for the pathogenesis of mammary adenocarcinomas in older human subjects is discussed. Finally, the potential role of insulin-like growth factor-1 (IGF-1) in the regulation of pathways responsible for cellular resilience to DMBA-induced mutagenesis is discussed.

Susceptibility to carcinogenesis critically determines mouse and rat lifespan

Cancer is a disease of aging. More than half of cancers are diagnosed in patients older than 70 and cancer is a leading cause of late-life mortality in the Western world (Siegel et al. 2018). Neoplasms are also a major cause of late-life mortality in laboratory mice and rats used in lifespan studies to understand fundamental longevity assurance mechanisms and to develop interventions that delay aging.

Progress in geroscience research in the past two decades has led to the identification of fundamental cellular and molecular mechanisms that contribute to aging processes (Kennedy et al. 2014). Such mechanisms include roles for genomic instability, altered telomere biology, increased oxidative stress, epigenetic changes, altered proteostasis, decreased nutrient sensing and metabolism, changes in senescence pathways, and stem cell function. It has been increasingly realized that there is a significant overlap between the cellular mechanisms of aging and pathways contributing to carcinogenesis (Hinkal and Donehower 2008). In particular, the ability to mount an effective, homeostasis-restoring defense response to environmental and cellular stressors also likely plays important roles in regulating the aging process and in determining the onset and progression of late-life diseases, including cancer. Among the mechanisms involved in cellular stress resilience, DNA damage response and DNA repair pathways are of special importance (Maynard et al. 2015). They are critical pillars of cellular homeostasis and organismal defenses, minimizing somatic mutations in both post-mitotic and mitotic cells in response to both intrinsic and extrinsic genotoxic stressors. These genomic maintenance systems prevent induction of cancers and maintain youthful tissue function and organismal longevity.

Long-lived animals (e.g., the longest living rodent, the naked model rat (Buffenstein 2005)) have evolved a robust set of cellular defense mechanisms, which maintain genomic integrity and prevent cancer over their entire lifespan (Gorbunova et al. 2014; MacRae et al. 2015; Seluanov et al. 2018). Genetic mouse models of successful aging that exhibit significant lifespan extension (e.g., growth hormone/insulin-like growth factor-1 [GH/IGF-1]-deficient Ames dwarf and Snell dwarf mice) also exhibit cellular resistance to genotoxic stresses (Maynard and Miller 2006; Murakami et al. 2003; Page et al. 2009; Podlutsky et al. 2017; Salmon et al. 2008; Salmon et al. 2005) and significantly decreased cellular mutation frequency and cancer incidence (Garcia et al. 2008). These findings accord with epidemiological findings of a positive correlation between tumor incidence and the rate of aging calculated on the basis of age-specific mortality in a given population (Anisimov 2001). Most pharmacological or dietary anti-aging interventions that extend lifespan usually also achieve this objective by inhibiting/delaying tumorigenesis in rodent models (Sharp and Richardson 2011). For example, caloric restriction was shown to significantly inhibit tumorigenesis and/or upregulate DNA repair pathways in responsive mouse and rat strains, which contributes to its lifespan-extending effects (Cheney et al. 1980; Cohen et al. 2004; Minor et al. 2010; Pearson et al. 2008b; Yamaza et al. 2010). Treatment with the mTOR inhibitor, rapamycin, which is known to extend mouse lifespan (Harrison et al. 2009), also delayed tumor onset and progression in cancer-prone and carcinogen-treated rodents (Granville et al. 2007; Mabuchi et al. 2007; Mosley et al. 2007). Further, rapamycin also prolonged the lifespan of cancer-prone mice (Anisimov et al. 2010). Of note, rapamycin failed to increase lifespan when given to mice with already established tumors, suggesting that rapamycin can increase lifespan primarily by decreasing tumorigenesis rather than by decelerating tumor growth (Anisimov et al. 2010). In contrast to longer living mice (Garcia et al. 2008; Gesing 2016; Smith et al. 1973), murine strains with the shortest lifespans frequently possess impaired genomic maintenance systems and are usually extremely susceptible to specific types of neoplasms. Genetically modified mice that exhibit defects in DNA repair and/or DNA damage response genes exhibit significantly increased cancer incidence, which often associates with systemic premature aging phenotypes (Hinkal and Donehower 2008). Virtually, all other genetic or dietary mouse models of accelerated aging (e.g., high-fat diet-fed mice (Sundaram and Yan 2016; Zhao et al. 2013; Zhu et al. 2017)) also demonstrate increased incidence of spontaneous tumors.

In the field of geroscience analysis of novel genetic mouse models with either a shortened, or an extended, lifespan provides a unique opportunity to evaluate the synergistic roles of longevity assurance pathways in cancer resistance and regulation of lifespan and to develop novel targets for interventions that both delay aging and prevent carcinogenesis (Anisimov 2001). There is a growing need for robust assays to assess the susceptibility of cancer in these models.

The present review focuses on a well-characterized method frequently used in cancer research, which can be adapted to study resilience to genotoxic stress and susceptibility to genotoxic stress-induced carcinogenesis in geroscience research namely, chemical carcinogenesis induced by treatment with 7,12-dimethylbenz(a)anthracene (DMBA) Figure 1. We review recent progress in understanding how longer-living mice may achieve resistance to chemical carcinogenesis and how these pathways are modulated by anti-aging interventions. Strain-specific differences in sensitivity to DMBA-induced carcinogenesis are also explored and contrasted with mouse lifespan. The clinical relevance of inhibition of DMBA-induced carcinogenesis for the pathogenesis of mammary adenocarcinomas in older human subjects is discussed. Finally, the potential role of IGF-1 in the regulation of pathways responsible for cellular resilience to DMBA-induced mutagenesis is discussed.

Fig. 1.

Concept figure illustrating the overlap between mechanisms of aging and mechanisms involved in carcinogenesis. Accordingly, impairment of DNA repair pathways and consequential DNA damage contribute both to impaired organismal stress resilience, accelerating the process of aging and increase susceptibility to carcinogenesis. Investigating DMBA-induced carcinogenesis in rodent models of longevity is a useful approach to understand the underlying mechanisms of both aging and age-related cancer development

DMBA-induced carcinogenesis

Several polycyclic aromatic hydrocarbons (PAHs) were identified as potent carcinogens in preclinical studies. Among these, DMBA has been extensively used to study the pathogenesis of mammary carcinoma in rodent models (Huggins et al. 1961). For example, it was demonstrated that even a single intra-gastric dose of DMBA was sufficient to induce mammary adenocarcinoma in female Sprague-Dawley rats (Huggins et al. 1961). DMBA-induced carcinogenesis is also well characterized in mouse models (Ethier and Ullrich 1982). The utility of DMBA-induced mouse mammary carcinogenesis models as valid equivalents for the study of human breast cancer has been established (Abba et al. 2016). The lipid-rich breast tissue is particularly susceptible to carcinogenesis induced by lipophilic PAHs, including DMBA. Like many other carcinogens, DMBA must be metabolized in order to induce DNA damage (Cavalieri and Rogan 1995). Xenobiotic-metabolizing enzymes have been shown to be responsible for the metabolism of DMBA through a two-step oxidation process to reactive intermediates and for the elimination of carcinogens from the body (Gonzalez 2001). Mice lacking the P450 enzyme, CYP1B1, or microsomal epoxide hydrolase (mEH) are less responsive to DMBA-induced tumorigenesis (Gonzalez 2001). Both natural and synthetic inhibitors of the P450 enzymes have been shown to inhibit DMBA-induced tumorigenesis (Cai et al. 1997). The ability of some antioxidants to inhibit carcinogenesis has been shown to be related to their abilities to inhibit the metabolism of DMBA (Slaga 1995). Initiation of the transformation of cells by carcinogens in vitro is dependent upon the cells being in the proliferative state so that the cells can metabolize the carcinogen to its reactive intermediate and replicate carcinogen-damaged DNA (Calaf and Russo 1993). In vitro studies demonstrate that, in mammary epithelial cells, PAHs are metabolically activated to cause significant DNA damage, mutations, and malignant transformation of the cells. Early studies in rats revealed that ovariectomy suppressed the susceptibility of these animals to DMBA-induced mammary carcinogenesis, confirming an important role for ovarian hormones in the pathogenesis of breast cancer (Lee and Oyasu 1974; Stevens 1966). It is generally accepted that cells need to be initiated with a carcinogenic insult before hyperstimulation with estrogen will result in carcinogenic progression (DiSaia and Creasman 1997). An initiated cell is one that has acquired a genetic mutation in an oncogene or a tumor suppressor gene (Kistner 1959). A promoter, such as estrogen, accelerates tumor formation without further DNA damage by stimulating the clonal expansion of initiated cells into a growing tumor (Butterworth 1990). Tamoxifen and some environmental estrogens have been shown to induce DNA mutations and therefore can contribute to tumor promotion and progression (Phillips 2001; Seidman and Kurman 2000). Progression of the tumor to a metastatic cancer is associated with further genetic changes that increase the invasive properties of the tumor cells (Kistner 1959).

Cellular resilience to genotoxic stressors and susceptibility to chemically induced carcinogenesis in longer-living versus shorter-living mice

Several lines of evidence support the view that genetics plays an important role in determining both mouse lifespan and susceptibility to cancer. The outstanding genetic resources for M. musculus include hundreds of inbred strains and also genetically modified mice that exhibit a significant variation of lifespan (Yuan et al. 2009). Accordingly, median lifespan ranges from less ~ 1.3 to ~ 3.1 years in various strains of mice used in aging research (Liao et al. 2010; Yuan et al. 2009). Neoplasms are a major cause of late-life mortality in laboratory mice, and mouse strains with the shortest lifespans are especially susceptible to carcinogenesis. In geroscience research, comparison of shorter-living and longer-living mouse strains is a promising approach to understand the roles of fundamental longevity assurance mechanisms in lifespan and cancer, and to develop interventions that delay aging and prevent carcinogenesis. Here, some important strain-specific differences in susceptibility to cancer are highlighted, which are relevant for geroscience studies.

C57BL/6 mice, which are used most frequently in aging studies, and BALB/c mice have a low incidence of spontaneously occurring mammary tumors. In contrast, other strains, including the C3H/Sm strain of mice, develop spontaneous mammary adenocarcinomas. Administration of DMBA to C57BL/6 mice and BALB/c mice (Ethier and Ullrich 1982) results in a moderate frequency of mammary tumors within 40 weeks after treatment. In BALB/c mice, DMBA-induced tumor incidence was reported to be 29% (Dusing-Swartz et al. 1979) to 68% (Medina 1974). In C57BL/6 mice, DMBA-induced mammary tumor incidences were reported to be 20% (Lydon et al. 1999) to 32% (Medina 1974), which increases to 60% in the presence of a pituitary isograft (Lydon et al. 1999). In contrast, in FVB/N mice treated with DMBA, mammary tumor incidences were reported to be 75% at 29 weeks after initiating DMBA treatment (Currier 2005). Hudson and coworkers provided a detailed analysis of mammary tumor development and survival rates in FVB and C57Bl/6 mice treated with DMBA (0.1 ml of 10 mg/ml DMBA dissolved in sesame oil by gavage once a week for 6 weeks) (Hudson et al. 2012). In these studies, the median time to death was 132 days in FVB mice and 180 days in C57Bl/6 mice (Hudson et al. 2012). Median time to mammary tumor onset was 166 days in FVB mice whereas 273 days in C57Bl/6 mice (Hudson et al. 2012). As expected, when C57BL/6 mice were crossed with the shorter-living and more cancer-prone DBA/2 strain of mice (which is the oldest inbred strain with a median lifespan of ~ 22.6 months), the resulting hybrids rapidly developed mammary cancers in response to DMBA treatment (Medina et al. 1980). Female hybrid C57BL/6 × DBA/2f F1 mice (derived from C57BL/6 females mated to DBA/2f males) treated with DMBA (1.0 mg dissolved in 0.2 ml cottonseed oil, given, i.e., once a week, for 6 weeks) were reported to exhibit a high incidence of mammary tumors (69 to 81%) (Medina et al. 1980). C3H/Sm mice are also sensitive to DMBA-induced mammary carcinogenesis (Drohan et al. 1982), with a reported incidence of DMBA-induced mammary tumors of ~ 57% (Medina and Smith 1999). Note that there appears to be an inverse correlation between susceptibility to DMBA-induced mammary carcinogenesis and mean lifespan of the FVB/N, C3H/Sm, BALB/c, and C57BL/6 mouse strains (~ 20, ~ 22, ~ 23.5, and ~ 30 months respectively). For a detailed analysis of the relationship between exposure of mice to DMBA and mammary tumor frequency over a wide range of doses as well as the relative effectiveness of DMBA given as single or multiple exposures, please consult the reference (Ethier and Ullrich 1982).

Topical application of DMBA induces skin cancer, which can also be exploited in geroscience studies. In animal models, numerous studies of organ sites, such as skin, utilize treatment with the tumor promoter, (TPA) after treatment with DMBA in a two-stage model of carcinogenesis, while animal studies in other organ sites, such as ovary, have shown that single or multiple treatments with DMBA are sufficient to induce carcinogenesis. The mechanism of TPA promotion is thought to involve the formation of hydrogen peroxide (H₂O₂). Chemoprevention agents that inhibit TPA promotion have been shown to inhibit H₂O₂ formation both in vivo and in vitro (Lim et al. 1992; Wei and Frenkel 1993). H₂O₂ is capable of promoting initiated cells both in vivo and in vitro (reviewed in (Huang et al. 1999)). In rat liver epithelial cells, H₂O₂ promotion is accompanied by induction of immediate early genes and disruption of gap junction potential, which can be blocked by antioxidants (Huang et al. 1999). Of note, there are also important strain differences in susceptibility to DMBA-induced skin tumorigenesis. Previous studies demonstrated that skin tumor induction by a 20-week treatment with DMBA produced papillomas, out of which 50.0% progressed to carcinomas in FVB/N mice, compared with 23.1% in BALB/c and 15.0% in C57BL/6 mice (Hennings et al. 1993). Note that there appears to be also an inverse correlation between the susceptibility to DMBA-induced skin carcinogenesis and mean lifespan of the FVB/N, BALB/c, and C57BL/6 mouse strains (~ 20, ~ 23.5, and ~ 30 months respectively). Susceptibility to DMBA-induced carcinogenesis may be organ-dependent. There are also striking strain differences in uterine carcinogenesis induced by a single administration of DMBA. The BALB/c, C3H, and C57BL/6 strains (mean lifespan ~ 23.5, ~ 23.6, and ~ 30 months) were reported to exhibit a high incidence of uterine tumors (87%, 90%, and 93%, respectively), while the SWR/J strain (mean lifespan ~ 22 months) showed a low incidence (17%) (Tsubura et al. 1993). There is also evidence that aging increases the susceptibility of the mouse skin to DMBA-induced carcinogenesis (Ebbesen 1974). Specific molecular events in DMBA carcinogenesis include G → T or A → T transversions in codons 12 and 61 of H-ras1, K-ras, and N-ras were detected in DMBA-induced skin, leukemia, and mammary tumors (Qing et al. 1997).

The molecular mechanisms responsible for strain-dependent differences in susceptibility to DMBA-induced carcinogenesis are likely multifaceted and are subject to intense research. Likely, mechanisms include differences in DNA repair pathways, chromatin accessibility, neuroendocrine factors, and differentially expressed cancer susceptibility genes (Chappell et al. 2017).

Important for the present overview, the apparent association of longevity and cancer resistance phenotypes suggests a key role of pathways involved in both carcinogenesis and longevity assurance (e.g., DNA repair pathways (Vermeij et al. 2016), IGF-1 signaling (Yuan et al. 2009)). Known factors that modulate both susceptibility of mice to DMBA-induced mammary carcinogenesis and regulation of lifespan include estrogen status (Fishman et al. 1995; Mehta et al. 2014), activity, and expression of Nrf2-driven antioxidant response pathways (Becks et al. 2010) and DNA repair mechanisms (de Vries et al. 1995; Hollander et al. 2001; Wijnhoven et al. 2001).

Both in humans and experimental animals, the efficiency of cellular DNA repair determines longevity and risk of cancer. Inherited mutations of DNA repair genes are known to lead to syndromes characterized by increased risk of cancer and/or a shortened lifespan in humans and are currently used to predict cancer risk (Corso et al. n.d.; Price et al. n.d.). Mouse models with genetic defects in DNA repair pathways are also characterized by increased susceptibility to mutagenesis and tumorigenesis induced by DMBA, as well as other genotoxic stressors (de Vries et al. 1995; Nakane et al. 1995; van der Horst et al. 1997; Wijnhoven et al. 2001; Wijnhoven et al. 2000). Lower DNA repair capacity also correlates with increased risk for breast cancer in humans (Kennedy et al. 2005; Machella et al. 2008; Santella et al. 2005; Shen et al. 2006). In addition, there is evidence that diet-induced obesity, which promotes an accelerated aging phenotype in mice (Minor et al. 2011; Pearson et al. 2008a; Ungvari et al. 2010a, 2011a; Zhang et al. 2015), also exacerbates DMBA-induced mammary carcinogenesis in this species (Zhao et al. 2013). Feeding a high-fat diet and the resulting obesity in mice is associated with complex neuroendocrine changes, including increases in leptin levels and dysregulation of insulin and insulin-like growth factor-1 (IGF-1) signaling. Further studies are warranted to elucidate the exact role of leptin-, insulin-, and/or IGF-1-dependent mechanisms in exacerbation of DMBA-induced carcinogenesis in obese mice.

From the aforementioned neuroendocrine factors, IGF-1 has special relevance for geroscience research (Yuan et al. 2009). Progress in biogerontology in the past two decades demonstrated that IGF-1-dependent pathways are involved in the regulation of aging processes (Ashpole et al. 2015, 2017; Bailey-Downs et al. 2012a, 2012b; Mitschelen et al. 2011; Podlutsky et al. 2017; Sonntag et al. 2013; Tarantini et al. 2016a, 2016b, 2017b; Toth et al. 2015; Toth et al. 2014; Yuan et al. 2009) and determine lifespan (Ashpole et al. 2017) in evolutionarily distant organisms (for a review, refer to the reference (Sonntag et al. 2012)). There are also several lines of evidence supporting the concept that IGF-1 signaling critically regulates carcinogenesis. First, patients with developmental IGF-1 deficiency due to GH insensitivity (Laron syndrome) do not appear to develop cancer during aging (Guevara-Aguirre et al. 2011). Mice with developmental GH/IGF-1 deficiency, including Ames dwarf mice, Snell dwarf mice, and growth hormone receptor gene-disrupted mice (GHRKO) also exhibit significantly decreased cancer incidence at old age and a marked extension of lifespan (Flurkey et al. 2001; Holzenberger et al. 2003; Liang et al. 2003; Sun et al. 2017). Cells derived from Ames dwarf, Snell dwarf, and GHRKO mice also exhibit cellular resistance to genotoxic stressors apparently due to their improved DNA repair capacity (Dominick et al. 2016; Leiser et al. 2006; Murakami et al. 2003; Page et al. 2009; Podlutsky et al. 2017; Salmon et al. 2005). Although detailed analysis of resistance to chemical carcinogenesis is lacking, studies on GH/IGF-1 deficient rat models (see below) demonstrate that early life disruption of GH/IGF-1 signaling results in marked resistance to DMBA-induced carcinogenesis (Ramsey et al. 2002). Early life restoration of IGF-1 levels in GH/IGF-1 deficient mice and rats with GH treatment reverses the cancer resistance and longevity phenotypes and also impairs cellular DNA repair capacity (Panici et al. 2010; Sun et al. 2017). These findings led to the hypothesis that GH/IGF-1 levels during a critical period early in life cause persisting changes in cellular DNA repair capacity, presumably by epigenetic regulation of genes involved in DNA repair (Podlutsky et al. 2017). The Aging Phenome Project (Bogue et al. 2016; Leduc et al. 2010; Yuan et al. 2015; Yuan et al. 2009), which aimed to characterize aging-related phenotypes of multiple inbred mouse strains, demonstrated that there is a significant variation of IGF-1 levels among these inbred strains. Importantly, early life IGF-1 levels in these strains are negatively correlated with median lifespan (Bogue et al. 2016; Leduc et al. 2010; Yuan et al. 2015; Yuan et al. 2009), providing indirect additional evidence that the early life hormonal milieu is an important determinant of the rate of aging and/or susceptibility to carcinogenesis in various mouse strains.

Susceptibility to chemically induced carcinogenesis in shorter-living and longer-living rats

Sprague-Dawley rats (median lifespan of 27 to 30 months (Masoro 1980)) and Lewis rats (median lifespan ~ 30.5 months (Sonntag et al. 2005a)) are highly sensitive to DMBA-induced mammary carcinogenesis (Wood et al. 2002). Wistar and Fisher F344 rats (median lifespan ~ 27 months and ~ 29 months, respectively (Masoro 1980)), Long-Evans, August, and ACI rats exhibit intermediate susceptibility to DMBA-induced mammary carcinogenesis (Wood et al. 2002). Other rat strains, including the Lewis dwarf (Ramsey et al. 2002), Copenhagen, SHR, Wistar-Kyoto, and DRH rats are highly resistant (Wood et al. 2002). It is of note that resistance to tumorigenesis is often organ-specific in these strains (Wood et al. 2002). For example, F344 rats exhibit the highest susceptibility for hepatocarcinogenesis, whereas Sprague-Dawley and Wistar rats have intermediate susceptibility (Wood et al. 2002).

Since the mid-60s, it has been established that the spontaneous development of mammary carcinomas during aging in the rat is modulated by neuroendocrine factors (Durbin et al. 1966). Strain-dependent differences in organismal sensitivity to DMBA-induced tumorigenesis have also been attributed previously, at least in part, to strain-specific differences in neuroendocrine regulation, including the function of the hypothalamo-pituitary-adrenal axis (De Jonage-Canonico et al. 2003; Jakubowski et al. 2002; Kerdelhue et al. 2016; Yon de Jonage-Canonico et al. 2005). Studies on cancer-resistant strains suggest that strain-specific resistance to DMBA-induced mammary carcinogenesis does not depend on strain-dependent differences in carcinogen metabolism, DMBA activation, and DNA adduct formation (Wood et al. 2002). The role of strain-specific differences in DNA repair pathways in differential sensitivity to both spontaneous and DMBA-induced carcinogenesis is not completely understood. The available data suggest that DMBA-induced preneoplastic lesions also do not progress well in cancer-resistant rat strains. Interestingly, susceptibility to DMBA-induced carcinogenesis also depends on the age of the rats (Sinha and Dao 1980). For a comprehensive review of potential genes involved in cancer resistance, consult the reference (Wood et al. 2002).

Lewis dwarf rats are homozygous for the spontaneous autosomal recessive dw-4 mutation, which causes a decrease in GH secretion from the pituitary gland beginning around post-natal day 26 (Carter et al. 2002a; Carter et al. 2002b; Charlton et al. 1988). Lewis dwarf rats have chronically low levels of GH and IGF-1 and make an excellent animal model of isolated GH/IGF-1 deficiency (Bailey-Downs et al. 2012b; Charlton et al. 1988; Ungvari et al. 2010b, 2011b; Yan et al. 2014). Importantly, Lewis dwarf rats were shown to exhibit marked resistance to DMBA-induced carcinogenesis (Ramsey et al. 2002) and cellular resistance to genotoxic stressors (Podlutsky et al. 2017). The GH-deficient spontaneous Dwarf rat is also resistant to both DMBA- and N-methyl-N-nitrosourea- induced mammary carcinogenesis (Swanson and Unterman 2002).

Susceptibility to chemically induced carcinogenesis in long-living rodent species

Differences in DNA repair capacity and cellular resilience to genotoxic stressors have been proposed to significantly contribute to the substantial variation of lifespan and differential susceptibility to carcinogenesis in mammalian species (Evdokimov et al. 2018; Ma et al. 2016; MacRae et al. 2015; Seluanov et al. 2018). The available evidence supports the view that longevity associates with resistance to DMBA-induced carcinogenesis in longer-living rodents. For example, the subterranean blind mole rats (Spalax) are long-lived (maximum lifespan potential > 20 years) rodents, which are resistant to both spontaneous and DMBA-induced carcinogenesis (Manov et al. 2013). Naked mole rats (maximum lifespan potential > 30 years) also exhibit marked cancer resistance (Buffenstein 2005). There is increasing evidence that there is a link between the ability of rodents to enter hibernation and longevity (Wu and Storey 2016). Interesting in this regard is the observation that hibernating and non-hibernating ground squirrels exhibit differential sensitivity to topically administered DMBA (Ruben 1982).

Clinical relevance: PAH-induced carcinogenesis in humans

Studies investigating genome maintenance pathways and cancer resistance in laboratory mouse and rat models of aging using DMBA-induced mammary carcinogenesis have important clinical relevance. Only ~ 5% of human breast cancer cases can be attributed to the inheritance of breast cancer susceptibility genes, whereas the remaining cases are considered to be sporadic in origin. It is highly likely that environmental influences, including carcinogen exposure, play a critical role in the pathogenesis of breast cancer.

PAHs exist and persist in the environment as a result of incomplete combustion of fossil fuels (coal and gas), hydrocarbons, and biomass for energy production and are common contaminants of terrestrial and aquatic ecosystems (Korsh et al. 2015). Many PAHs are persistent (remain in the environment for a long time without decomposition) and are bioaccumulative. There are over 100 PAHs with similar chemical properties, typically occurring as mixtures, which include benz(a)pyrene, dibenz(ah)anthracene, 1-nitropyrene, and DMBA, all of which are established experimental breast carcinogens. Exposure to PAHs occurs via inhalation or oral uptake, and via consumption of smoked and grilled meat and fish (e.g., barbecues). PAHs are found in tobacco smoke and automobile exhaust smoke. Additionally, dermal exposure may result from the use of PAH-containing consumer products made of rubber or soft PVC. PAHs are hydrophobic compounds and can be readily absorbed through the dermis. A wide range of consumer products designed for skin contact (e.g., sport equipment and household utensils, toys, tools, clothes, or wristbands) contain PAHs due to the frequent application of mineral oil products as plasticizers or lubricants during the manufacturing process or the addition of carbon black for coloring and abrasion resistance purposes (Bartsch et al. 2017). Athletically active individuals may be exposed to PAHs via skin contact with abrasions from synthetic turf fields. PAHs are also present in paraffins, widely used for medicinal and cosmetic purposes. Epidemiological evidence clearly demonstrates the carcinogenic effects of PAHs in occupationally exposed subjects (Mastrangelo et al. 1996). Exposure to PAHs and PAH-induced DNA damage has been linked to the pathogenesis of breast cancer (Korsh et al. 2015; Large and Wei 2017; Lee et al. 2019; Li et al. 1996; Maltoni et al. 1989; Niehoff et al. 2017a; Rundle et al. 2000; Rundle et al. 2002; Shen et al. 2017; White et al. 2016) as well as lung cancer, bladder cancer, and skin cancer in humans. Genetic and environmental factors (e.g., dietary factors, exogenous hormone exposure, including hormone replacement therapy) likely interact with each other to influence individual susceptibility to carcinogenesis.

Developmental reprogramming of cancer susceptibility and longevity

Over the past decades, experimental, clinical, and epidemiological studies provided strong support to the concept that early life events shape developmental trajectories and fundamentally impact health span and lifespan. The “developmental origins of health and disease” (DOHaD) concept posits that several types of cancer, including breast cancer, may have early life origins (Beinder et al. 2014; Biro and Deardorff 2013; Denholm et al. 2016; Ekbom et al. 1992; Michels et al. 1996). The DOHaD concept predicts that exposure to dietary factors or endocrine disruptors or alterations in the endocrine milieu during critical periods of development reprograms the epigenome of normal cells, altering cellular resilience to genotoxic stressors and predisposing the tissues to a higher risk of carcinogenesis, including chemically induced carcinogenesis, later in life (Chiam et al. 2009; Doherty et al. 2010). It is well established that windows of susceptibility relevant for the pathogenesis of breast cancer include the prenatal period (e.g., due to suboptimal maternal nutrition (Ekbom et al. 1992; Walker and Ho 2012)). More recent experimental and epidemiological studies confirm that the peripubertal period, which is a time of rapid breast development (Haslam and Schwartz 2011; Olson et al. 2010; Zhao et al. 2013), is also an important window of susceptibility in which the breast cells might be especially vulnerable to potential carcinogens, endocrine disruptors, and endocrine changes, which affect the risk of breast cancer in later life (Biro and Deardorff 2013). For example, the NIEHS (National Institute of Environmental and Health Sciences) Sister Study, which examines environmental and familial risk factors for breast cancer in sisters of women who have had breast cancer, demonstrated that lifestyle factors (e.g., higher level of physical activity) around puberty influences breast cancer risk later in life (Niehoff et al. 2017b). Animal studies support the concept that in addition to fetal exposure, the peripubertal period is also a susceptible window for epigenetic reprogramming, modulating risk for carcinogenesis.

The endocrine changes that occur during puberty are highly conserved across mammalian species and include dramatic increases in the anabolic hormone insulin-like growth factor-1 (IGF-1) (Carter et al. 2002b; D'Costa et al. 1993; Deak and Sonntag 2012; Sonntag et al. 2012; Sonntag et al. 2000; Sonntag et al. 1999; Sonntag et al. 2005b) in addition to changes in sex hormone levels. During puberty, increased levels of estrogen and IGF-I synergize for ductal morphogenesis (Kleinberg and Barcellos-Hoff 2011). These endocrine changes are important for normal breast development and their dysregulation is hypothesized to contribute significantly to alterations in cellular resilience to genotoxic stressors and susceptibility to tumorigenesis. Importantly, in humans, the peripubertal IGF-1 surge is highly variable. There is increasing epidemiological and experimental evidence to support the concept that timing and magnitude of the peripubertal IGF-1 surge in this sensitive developmental window modulate susceptibility to mammary carcinogenesis later in life.

Human patients with Laron dwarfism, which is caused by a mutation in the GH receptor resulting in a lack of the peripubertal IGF-1 surge (Guevara-Aguirre et al. 2011), do not develop breast cancer later in life. Lewis dwarf rats, which also do not exhibit a peripubertal IGF-1 surge, were also shown to exhibit marked resistance to DMBA-induced mammary carcinogenesis (Ramsey et al. 2002). Importantly, early life treatment of GH/IGF-1 deficient Lewis dwarf rats with GH, which temporarily restores IGF-1 levels, was shown to reverse the cancer resistance phenotype (Ramsey et al. 2002). Selective developmental GH/IGF-1 deficiency in this model also resulted in persisting increases in cellular DNA repair capacity and upregulation of several DNA repair-related genes, suggesting the cellular basis for the persisting influence of peripubertal GH/IGF-1 status on cancer resistance and cellular resilience to genotoxic stressors (Podlutsky et al. 2017). Studies in mouse models of GH and IGF-1 deficiency also confirm that modulation of the early life IGF-1 surge has significant late-life consequences, altering susceptibility to cancer and abolishing the longevity phenotype (Panici et al. 2010). Additionally, liver-specific IGF-I gene–deleted (LID) mice, in which circulating IGF-I levels during development are 25% of that in control mice were shown to exhibit marked resistance to DMBA-induced mammary carcinogenesis (Wu et al. 2003). Similarly, transgenic mice that constitutively express a GH antagonist exhibit a significant decline in early life IGF-1 levels, which associate with marked resistance to DMBA-induced mammary carcinogenesis later in life (Pollak et al. 2001). Interestingly, in inducible liver IGF-1-deficient (iLID) mice, a 75% decline in circulating IGF-1 induced at 1 year of age resulted in an increased tumor incidence, demonstrating the importance of the early lifetime window for cancer prevention (Gong et al. 2014).

Epidemiological studies provide also indirect evidence to support a link between IGF-1-dependent pubertal growth spurt and breast cancer later in life. A recent study reported that risk for breast cancer increases 11% for every 5-cm increase in adult height (an indirect indicator of the peripubertal IGF-1-dependent growth spurt) (Bouhours-Nouet et al. 2007). Similar conclusions were reached in the Million Women Study, demonstrating a 1.17-increased risk for every 10-cm adult height (Green et al. 2011). Similarly, if a woman reaches her maximum height at or before the age of 12 years (suggesting an earlier/larger peripubertal IGF-1 surge), the risk of developing breast cancer later in life is significantly increased (by 1.4) (Li et al. 2007). Another indirect indicator of the peripubertal IGF-1 surge is bone mineral density, as the majority of bone mineral content is deposited during the teenage years (Baxter-Jones et al. 2011) and circulating levels of IGF-1 directly regulate bone growth and density (Mohan et al. 2003; Yakar et al. 2002). There are several studies demonstrating the relationship between greater bone mineral density and later development of breast cancer (Cauley et al. 1996; Cauley et al. 2007; Zhang et al. 1997).

Nutritional factors are critical regulators of IGF-1 levels during development. Nutritional status during childhood has a significant effect on pubertal development and contributes significantly to the variation in the timing of puberty and the timing and magnitude of the peripubertal IGF-1 surge (Thissen et al. 1994). Variation in the hormonal milieu due to nutritional status likely plays an important role in the reprogramming of cancer susceptibility and longevity. Accordingly, young female patients with anorexia nervosa develop multiple endocrine abnormalities, including reduced IGF-1 levels during a sensitive developmental time window (Gianotti et al. 2002). Data from the Sister Study (which is a large, prospective, observational cohort study that enrolled women aged 35–74 who had a sister with breast cancer but had never had it themselves) demonstrate that there is a significant inverse association between having a history of an eating disorder and invasive breast cancer (O'Brien et al. 2017). These results accord with the conclusions of other population-based registry studies, showing that women diagnosed with anorexia at younger ages have a substantially reduced risk for developing breast cancer later in life (Karamanis et al. 2014; Mellemkjaer et al. 2001; Mellemkjaer et al. 2015; Michels and Ekbom 2004). The results strikingly show that early onset anorexia in pre-adolescent and adolescent girls result in a significant reduction in breast cancer risk (standardized incidence ratio/the ratio of observed-to-expected number of cases 0.4), whereas adult onset anorexia has no significant effect (Papadopoulos et al. 2009). Overconsumption of calories and childhood obesity also affect the hormonal milieu during development, including a significant rise in circulating IGF-1 (Silfen et al. 2002; Thissen et al. 1994), which contribute to developmental reprogramming of susceptibility to mammary carcinogenesis (Gunther et al. 2015; Jung et al. 2016). Indeed, recent data have shown a strong association between higher body mass index during adolescence and increased risk for several malignancies, including colorectal cancer and breast cancer and in adulthood (Weihrauch-Bluher et al. 2018). Interestingly, higher adolescent body mass index also doubles the risk for male breast cancer later in life (Keinan-Boker et al. 2018). Preclinical studies in rodent models also support the concept that interaction of nutritional status, obesity, and GH/IGF-1 signaling regulates susceptibility to DMBA-induced mammary carcinogenesis (Aupperlee et al. 2015; Gahete et al. 2014; Hakkak et al. 2007; Zhu et al. 2017).

Epigenetic regulation of organismal susceptibility to chemically induced carcinogenesis

Accumulating evidence demonstrates that epigenetic alterations regulate aging and determine susceptibility to the development of age-related diseases, including cancer (Gentilini et al. 2013; Jones et al. 2015; Jung et al. 2015; Jung and Pfeifer 2015; Sen et al. 2016; Thompson et al. 2010; Zampieri et al. 2015). Diet, lifestyle, and other environmental factors were demonstrated to induce epigenetic alterations resulting in persisting changes in the cellular transcriptome, which modulate aging processes and contribute to the altered risk of various age-related diseases including cancer. The most widely recognized epigenetic regulatory mechanisms include DNA methylation, histone modifications, and noncoding RNA expression. Epigenetic changes can lead to altered gene expression, which can impact cellular pathways involved in both aging processes and carcinogenesis, including molecular mechanisms underlying cellular resilience to genotoxic stressors. DNA methylation, catalyzed by a group of enzymes called DNA methyltransferases (DNMTs), is a key physiologic mechanism of epigenetic silencing of transcription. There is growing evidence suggesting that inactivation of genomic maintenance/tumor suppressor genes by DNA methylation is mechanistically linked to carcinogenesis. Epigenetic silencing of DNA repair genes via promoter methylation during carcinogenesis has been reported for several DNA repair pathways including base excision repair, nucleotide excision repair, DNA mismatch repair, and several other DNA damage processing mechanisms (Dobrovic and Simpfendorfer 1997; Esteller et al. 2001; Guan et al. 2008; Howard et al. 2009; Kane et al. 1997; Lahtz and Pfeifer 2011; Lawes et al. 2005; Lee et al. 2007; Narayan et al. 2004; Peng et al. 2005; Peng et al. 2006; Vo et al. 2004). Further, several risk factors that promote breast cancer and/or shorten lifespan, including obesity, were shown to alter DNA methylation patterns (Melnyk et al. 2017). The current view is that promoter methylation and silencing of DNA repair genes are early events in tumorigenesis that contribute to both cancer initiation and progression by increasing genomic instability. Genomic instability, in turn, promotes genetic aberrations at other important gene loci (Lahtz and Pfeifer 2011), leading to changes in the expression of genes critical to transformation pathways.

Regulation of DNA repair genes and other genes involved in resistance to genotoxic stressors via epigenetic modifications is a plausible explanation for the effects of early life influences (e.g., dysregulation of peripubertal IGF-1 levels) on carcinogenesis later in life (Panici et al. 2010; Ungvari et al. 2011b). In this context, there is evidence that IGF-1 may regulate the expression of DNMTs (Armstrong et al. 2014). For example, in long-lived Ames dwarf mice, in which the peripubertal IGF-1 surge is missing, decreased expression and activity of DNMTs are associated with global changes in DNA methylome (Armstrong et al. 2014). Caloric restriction is also associated with significant epigenetic changes (Daniel and Tollefsbol 2015), improved DNA repair capacity (Heydari et al. 2007), and marked resistance to DMBA-induced carcinogenesis (Klurfeld et al. 1989). To better understand the developmental reprogramming of cancer susceptibility and longevity in geroscience research, studies on mouse models of longevity involving integrated assessment of DNA methylation, DNA repair efficiency, and susceptibility to chemically induced carcinogenesis will be highly informative.

Perspectives

Mechanisms of aging, including increased DNA damage and impaired DNA repair, also contribute to carcinogenesis. Combining concepts, animal models, and research tools of biogerontology and cancer research provide tremendous opportunities for discovery of new drugs that target both aging and cancer (Sharp and Richardson 2011). In particular, studies on murine models of longevity are expected to reveal critical mechanisms by which resilience to genotoxic stressors, susceptibility to carcinogenesis, and aging processes are regulated. Understanding these mechanisms will provide novel targets for intervention that both delay aging and prevent carcinogenesis (Anisimov 2001). The association of peripubertal IGF-1 levels with cancer susceptibility provides a starting point identifying and for developing new interventions. As DMBA-induced mouse mammary carcinogenesis models are valid equivalents of human breast cancer (Abba et al. 2016), the translational relevance of this approach is high.

In geroscience studies (Bennis 2017; Blodgett et al. 2017; Contreras et al. 2018; Deepa et al. 2017; Fang et al. 2017; Fulop et al. 2018; Habermehl et al. 2018; Kaeberlein 2018; LaRoche et al. 2018; Lee et al. 2018; Lewis et al. 2018; Logan et al. 2018; Meschiari et al. 2017; Nacarelli et al. 2018; Reglodi et al. 2018; Scerbak et al. 2018; Snider et al. 2018; Tenk et al. 2017; Tucsek et al. 2017; Ungvari et al. 2017; Urfer et al. 2017; Vedovelli et al. 2017), there is a growing need for new assays to assess organismal stress resilience, including resistance to genotoxic stresses, to predict health span and lifespan and to predict the efficiency of novel anti-aging treatments. Testing DMBA-induced mammary carcinogenesis in mice is a useful assay to measure resilience to genotoxic stress. As resistance to chemically induced carcinogenesis represents an aggregate effect, cell-based DNA repair assays can serve as complementary, mechanistic approaches.

Funding information

This work was supported by grants from the American Heart Association (to ST), the National Institute on Aging (R01-AG055395 to ZU, R01-AG047879 to AC, R01-AG038747), the National Heart Lung Blood Institute (R01HL132553), the National Institute of Neurological Disorders and Stroke (NINDS; R01-NS056218 to AC, R01-NS100782 to ZU), the NIA-supported Geroscience Training Program in Oklahoma (T32AG052363), the NIA-supported Oklahoma Nathan Shock Center (to ZU and AC; 3P30AG050911-02S1), NIH-supported Oklahoma Shared Clinical and Translational Resources (to AY, NIGMS U54GM104938), the National Cancer Institute (R01-CA196200 and R01-CA200126 to DMB), the Oklahoma Center for the Advancement of Science and Technology (to AC, ZU, AY), the Presbyterian Health Foundation (to ZU, AC, AY), and the Reynolds Foundation (to ZU and AC).