A revisiting of “the hallmarks of aging” in domestic dogs: current status of the literature

Abstract

A progressive decline in biological function and fitness is, generally, how aging is defined. However, in 2013, a description on the “hallmarks of aging” in mammals was published, and within it, it described biological processes that are known to alter the aging phenotype. These include genomic instability, telomere attrition, epigenetic alterations, loss of proteostasis, deregulated nutrient sensing, mitochondrial dysfunction, cellular senescence, stem cell exhaustion, altered intercellular communication (inflammation), and changes within the microbiome. This mini-review provides a detailed account of the progress on each of these hallmarks of aging in the domestic dog within the last 5 years. Additionally, when there are gaps in the literature between other mammalian species and dogs, I highlight the aging biomarkers that may be missing for dogs as aging models. I also argue for the importance of dog aging studies to include several breeds of dogs at differing ages and for age corrections for breeds with differing mean lifespans throughout.

Introduction

Aging can be described as a holistic biological decline, which decreases survival and reproduction [120]. It is progressive, endogenously derived, irreversible, and expands across multiple processes and systems within an organism [25]. López-Otín et al. [93] highlighted the hallmarks of mammalian aging as genomic instability, telomere attrition, epigenetic alterations, loss of proteostasis, deregulated nutrient sensing, mitochondrial dysfunction, cellular senescence, stem cell exhaustion, and altered intercellular communication. In 2022, these hallmarks of aging were revisited, and microbiome disturbance, splicing dysregulation, and inflammation were explicitly included in the list of hallmarks of aging [124].

Domestic dogs are the most morphologically and physiologically diverse group of mammals that we know of. And, functionally, this diversity is also associated with vastly different aging rates and longevity across body sizes of dogs. For example, smaller dogs tend to live significantly longer than larger dogs across all breeds, unlike the positive inter-species relationship found between body mass and lifespan across mammals [68]. Smaller dogs also have lower cancer risks [30, 85, 99, 125] implying slower rates of aging. Here, I summarize the recent data available on the hallmarks of aging solely on domestic dog aging, without comparison to other aging models (including humans). This current mini-review is meant to compliment the work already reviewed by Ruple et al. [120] and update the excellent review of the hallmarks of aging in dogs by Sándor and Kubinyi [121]. I will describe each hallmark of aging, in turn, below, and highlight the recent advances in each within the dog aging literature and where the field is lacking information.

Methods

Each of the biomarkers of the hallmarks of aging described in López-Otín et al. [93] and Schmauck-Medina et al. [124] was searched, in turn, with “and dogs” and “and canines” using googlescholar.com. For each of the papers that came up in the search per biomarker, the abstracts were read. I included papers in this mini-review that included large sample sizes and dogs of various ages so as to account for an aging change in the dog population of the biomarker described. After this step, each paper that was deemed within the criteria above was then searched using the “cited by” function in googlescholar.com to find other related literature. After compiling papers for each biomarker per hallmark of aging, the reference section of each paper was examined to find any additional references that may have been overlooked.

Genomic instability

The accumulation of genetic damage is a ubiquitous aspect of the aging process [93]. Endogenous or exogenous damage that progressively leads to the accumulation of genetic mutations can come in the form of point mutations, translocations, chromosomal gains and losses, telomere attrition, and gene disruption caused by the integration of viruses or transposons. To prevent this damage, DNA repair mechanisms are present within mammalian cells [94]. Aging can rapidly progress when DNA damage is not repaired or if the accumulation of damage overwhelms the repair mechanisms [93]. Thus, genomic instability includes pathways associated with repair/damage to nuclear DNA, mitochondrial DNA, and nuclear architecture [93].

Damage to DNA can also come in the form of copy number variants (CNVs) with increases in frequency distribution having harmful phenotypic effects [26]. Whereas there is no published work on CNVs during aging in dogs, [14], looking at 50 individuals of 17 different breeds, found that CNVs are good candidates to identify breed-specific phenotypes. Furthermore, another study looking at 300 individuals across 100 dog breeds created a map of structural variation in dogs to explore the association between CNVs and phenotypes. Using this map, the authors detected more than 110 clear instances of CNVs that are at an increased frequency in a subset of modern dog clades, suggesting candidacy for recent selection. Also, because CNVs had a lower incidence compared to other kinds of genomic variation, CNVs may be strong phenotypic drivers [127].

Mitochondrial DNA mutations also contribute to the aging process [109] and may alter patterns in oxidative stress (below). Direct measurements of mitochondrial DNA mutations during aging cannot be found in the dog aging literature. Because it has long been proposed that the longer growth phase of large breed dogs [68] can accrue damage early on [48, 84], the age at which mitochondrial mutations first appear might be important for aging and age-related diseases in dogs [140]. Additionally, there is no work looking at nuclear architecture, specifically alternations to the nuclear lamina or spindle assembly checkpoint fidelity in aging dogs [93]. Of particular interest would be work that links telomere dynamics (below) with the production of progerin [93]. Additionally, the literature is also sorely lacking in data including nucleotide alterations in excision repair, nonhomologous end-joining, mismatch repair, and translesion synthesis in aging dogs [93].

Telomere dynamics

Telomeres are protective caps at the end of chromosomes that are linked to genomic stability. As cells divide with increasing organismal age or in times of increased oxidative stress, these caps get shorter [117]. When the caps get to be critically short, DNA damage begins to accrue, and cell senescence cascades are initiated [38]. In the last 5 years, there has been no new work published on this hallmark of aging in domestic dogs,thus, the details by Sándor and Kubinyi [121] are still current. In brief, using peripheral blood mononuclear cell (PBMC) or leukocyte, it was demonstrated that telomere length is a strong predictor of average life span in 15 different breeds of dogs, and thus, can play a role in longevity for this species [38]. Notably, telomeric restriction fragments demonstrate a significant effect of breed [96], as do Fick et al. [38], thus, careful consideration and corrections for breed and donor age should be used for this hallmark of aging in dogs.

It is noteworthy to consider the fact that most of the work published on this hallmark of aging in dogs has been done using white blood cells, which may differ from other tissues [104]. Thus, it would be advantageous to compare telomere dynamics in other tissues.

Epigenetic alterations

Epigenetic alterations include changes in DNA methylation patterns, posttranslational modification of histones, and chromatin remodeling [93, 121]. As highlighted by Sándor and Kubinyi [121], there is now strong evidence that epigenetic regulation is dog breed-specific [6, 81]. Specifically, Thompson et al. [133] first highlighted that changes in DNA methylation status in age-sensitive DNA regions show a strong correlation with chronological age in dogs and wolves. Then, a dual-species human–dog clock relating relative age to cytosine methylation was developed [63]. Furthermore, an epigenetic estimator of average time to death, potentially estimating mortality risk based on blood methylation profiles, was also included in Horvath et al.’s study [63]. Others have developed an oligo-capture system to characterize the canine DNA methylome and develop an epigenetic clock that was trained in one species, dog, and one breed, Labrador, but leads to a high correlation in another species [142].

A family of proteins known as sirtuins has been shown to affect longevity and lifespan, and, as enzymes, they can deacetylate histones and, thus, help maintain chromatin structure [93, 121]. Sirtuins and other histone-modifying DNA methyltransferases have still been barely studied in dogs, as highlighted before [121]. SIRT1, in particular, is a NAD⁺-dependent histone deacetylase with several protein substrates [146] and a key epigenetic regulator [107]. SIRT1 is localized to the nucleus and deacetylates multiple non-histone proteins in response to stress [52, 91, 122]. It also serves to regulate DNA repair, apoptosis, mitochondrial biogenesis, and cell stress responses, to name a few [146]. Mammalian SIRT1 has also been shown to affect downstream signaling, such as insulin-growth factor-1 (IGF-1) signaling, which has long been linked to lifespan regulation, and increased energy availability [91].

Generally, in mammals, as cells age, SIRT levels decrease and DNA damage increases as well, though without apoptosis which may lead to an increase in mutations [101]. Using primary fibroblast cells of puppy and old large breed and small breed dogs, SIRT1 decreases with increasing age similarly in small and large breed dogs [68], and DNA damage also increases with age in primary fibroblasts of large breed old dogs [69]. Together, these data suggest that small-breed dogs may have DNA repair mechanisms that are different than those of large-breed dogs, a previously understudied aspect of dog aging.

Any studies considering specific histone modifications (i.e., H4K16ac, H4K20me3, H3K4me3, H3K9me, and H3K27me3) or specific biomarkers of chromatin remodeling such as HP1α or NuRD are largely lacking in the dog aging literature. A discussion of epigenetic clocks and how they can be used to determine chronological age in domestic dogs can be found within Ruple et al.’s study [120].

Loss of proteostasis and autophagy

Proteostasis involves an array of cellular mechanisms that ensure proteins are folded correctly, for example, the heat shock family of proteins [56]. A loss of proteostasis then implies the misfolding of proteins that occurs with disease or increasing age [113]. Misfolded proteins can be refolded by heat-shock proteins (HSP) or deemed for destruction by the autophagic pathways including the recognition of misfolded proteins by chaperones like Hsc70, the ubiquitin–proteasome, and lysosomal pathways (macroautophagy). If misfolded proteins are not able to refold properly, it can lead to their accumulation and potential cell toxicity [93]. MTMR14 has been previously found to be more active in older dogs compared with young ones. Decreased autophagic activity in neurons at advanced ages may be a consequence of increasing MTMR14 accumulation during the lifespan [83]. In aging Labrador retrievers, serum baseline levels of Hsp-70 decreased with age [1], suggesting a limitation towards proper refolding mechanisms with age in that one breed. However, multi-breed comparisons are currently lacking in this aspect of aging in the dog model.

In the dog brain, we know that deposition of amyloid beta (βA), which in humans can be linked to the development of Alzheimer’s disease, is also deposited in the cerebral blood vessels of 70% of elderly dogs examined as compared with a younger dog population [105]. Using another population of 47 dogs, I also found that βA plaques were deposited in an age-dependent manner in dogs [21]. Furthermore, amyloid beta (Aβ42) levels were measured in three parts of the canine brain: prefrontal and temporal cortexes, as well as the hippocampus and the cerebrospinal fluid. Aβ42 accumulation was also found to be age-dependent and correlated to the Canine Cognitive Dysfunction Scale score [137].

Most of the work in the dog aging literature related to loss of proteostasis has been done with respect to cognitive function in dogs, and thus, measured in brain tissue [18, 119]. That is to imply that we, currently, know very little about the rates of loss of proteostasis and other associated pathways of protein folding regulation in other dog tissues, particularly muscle tissue, which should be of interest as it is associated with mobility decline with age [108]. Sarcopenic pathways in dogs of different ages and with different body masses are profoundly lacking currently. And, again, it should be noted that the dog aging literature is deficient in studies considering detailed macroautophagic pathways, Hsc70, HSF-1, and proteasomal degradation [93, 121].

Deregulated nutrient sensing

Oftentimes, the aging literature refers to the growth hormone/insulin growth factor-1 (GH/IGF-1) pathway as the most conserved aging pathway across species [25, 93]. The somatotrophic axis in mammals is composed of GH, produced by the anterior pituitary, and IGF-1 which is produced in response to GH in many cell types. This pathway informs cells of the presence of glucose. Insulin is involved in cell growth and glucose homeostasis [111], whereas IGF-1 is thought to influence cell growth, differentiation, and survival [29]. Insulin and IGF-1 concentration seem to be negatively correlated with lifespan in several mouse strains selected for disparate lifespans, like dogs [41, 134], where mice selected for dwarfism due to a lack of GH production and a reduction of IGF-1 in plasma had longer lives than their control counterparts [29].

A single IGF-1 haplotype seems to substantially contribute to size variation in domestic dogs [131], and serum IGF-1 concentration in small dogs is reduced relative to that found in large and giant breeds, providing a potential link to their longer lives [36, 51]. Additionally, IGF-1 has been found to decrease with age [51]. For example, in medium- and large-size dogs, the concentrations of GH and IGF-1 were significantly higher in younger individuals compared to older individuals in both size classes, and GH and IGF-1 concentrations were higher in larger dogs compared with middle-sized dogs [90]. Some authors have also found breed-related differences in GH concentrations at a young age, where giant breeds had higher GH concentrations compared with medium breeds [37].

Nutrient sensing can also be linked to cellular metabolic rates and metabolomics. Aerobic cellular metabolic rates from primary fibroblasts have shown that basal rates of oxygen consumption decrease and proton leak increases with age of the dog donor despite the dog’s size class [69], and long-lived breeds seem to have more mitochondrial uncoupling compared with shorter-lived breeds [106]. However, large-breed dogs seem to demonstrate a predominant glycolytic phenotype across their whole lives compared with smaller breeds [69]. This pattern may be breed-specific [114]. There is a link between increased IGF-1 concentration and cancer rates in that IGF-1 promotes tumorigenesis [7], and the predominant glycolytic phenotype in large breed dogs may predispose large breed dogs to a more cancerous phenotype.

The metabolome is defined as the collection of metabolites in a cell or organism [100]. To better understand the complexity of changes during aging, a metabolome-wide approach can be applied to pinpoint key metabolic steps that may determine lifespan in dogs [60, 61]. In dogs, metabolomic analysis has mainly been at the level of urine and plasma and includes a life-long project on Labrador retrievers stratified as control animals versus those calorically restricted (CR) for life. Creatine increases were linked to muscle wasting with increased age [141], and lactate concentration in urine increased in older dogs [141], demonstrating a shift into a glycolytic phenotype similar to studies using primary fibroblast cells highlighted above [69]. Using primary fibroblasts, Brookes and Jimenez [20] found that the lower half of glycolysis is depressed in larger older dogs, which may represent a metabolic deficiency that may be linked to glucose oxidation rather than lactate production [69]. These data suggest a decrease in pyruvate kinase activity in large old dogs and are consistent with a linear regression analysis of the enzymatic activity of this enzyme in blood plasma being lower in large dogs [144]. And, similarly, using a metabolomic approach, others have found a negative correlation between glycolysis and age using whole-blood samples from dogs [60].

The aging phenotype in dogs has been previously associated with lower levels of glycine, aspartate, creatine, and citrate. Additionally, lower levels of lipoprotein fatty acyl groups were also observed [116]. There are also metabolic differences with respect to breed, such that large breeds can be separated from others [12, 138]. Urine seems to more consistently give predictive measures of age, whereas serum analysis yields different types of metabolites [116]. Thus far, the only association with weight in the domestic dog has been a negative correlation with tryptophan metabolism [60]. Though, using lipidomics, it was demonstrated that sphingomyelins were significantly higher in large, short-lived dogs, independent of age, and triglycerides were higher in older dogs of all sizes [61]. Also using metabolomics in primary fibroblast cells of different dog breeds, Nicholatos et al. [106] found that longer-lived (smaller) breeds have lower concentrations of acylcarnitines and higher fatty acid oxidation, similar to the findings of Brookes and Jimenez [20]. L-carnitine has also been demonstrated to be a validated metabolite relevant to aging in humans [89]. However, most of the human aging literature suggests a correlation between lipid metabolism and aging [89]. Additionally, Hoffman et al. [60] demonstrated a positive correlation between saturated fatty acid β-oxidation and age in whole blood samples of dogs. Thus, a detailed, multi-breed approach to lipidomics of different tissue types may prove a fruitful approach towards determining aging differences in the dog.

Though IGF2 has gained momentum across mammalian species [11], it has yet to be measured with respect to aging and across different breeds of dogs.

Mitochondrial dysfunction

Mitochondria play a central role in organismal homeostasis, which becomes important during aging, leading to potential dysregulation. Approximately 5% of oxygen waiting to receive an electron is never fully reduced to water and can become a reactive oxygen species (ROS) [67]. At low levels, ROS are important signaling molecules for gene regulation, cell signaling, and apoptosis [35, 103, 128]. At high levels, these radicals scavenge electrons from surrounding molecules, potentially causing damage to lipids, proteins, and even DNA [39, 67, 103]. Some have hypothesized that large breed dogs, due to increases in ROS production, may be burdened with more oxidative damage at an earlier age, leading to a disease phenotype sooner than their small breed counterparts [48, 84]. However, ROS production, taken in isolation, cannot paint the full oxidative picture, as all cells have an antioxidant system that works to quench any errant ROS and can thwart oxidative damage [67]. This pathway, taken together, is called oxidative stress (OS). Broadly defined, OS is the balance between prooxidants produced during aerobic metabolism, mainly by mitochondria, and antioxidants, which are present in the form of enzymatic and non-enzymatic molecules capable of thwarting prooxidants before cellular damage occurs [5, 54]. Oxidative damage accrues when ROS production overwhelms the antioxidant system [35, 103, 128]. And oxidative damage can happen to many biologically relevant molecules, such as proteins, DNA, and lipids [67]. Lipids are among the molecules most affected, and two of the most prevalent prooxidants that can initiate damage to lipid membranes are hydroxyl radicals (OH·) and hydroperoxyl radicals (OOH·) [5]. Enzymatic antioxidants, such as glutathione peroxidase (GPx, superoxide dismutase (SOD, and catalase (CAT, function by catalyzing the oxidation of less biologically insulting molecules. Other antioxidant molecules such as vitamins E and C, act as chain-breaking antioxidants; they scavenge for ROS, remove them once they are formed, and further halt the propagation of peroxidation [54]. The following consideration is of utmost importance: The relationship between metabolism (whole-animal or mass-specific) and ROS production is not straightforward, though many assume that an increase in oxygen consumption (i.e., an increase in metabolic rate) should yield an increase in ROS production [64]. Oxygen consumption can either be coupled with ATP production or heat depending on whether the ATP synthase or mitochondrial uncoupling proteins are driving respiration [34], Hou et al. 2020); thus, it should never be assumed that oxygen consumption and ROS production are linear. Reduced mitochondrial efficiency can be linked to ROS production or other cellular mechanisms, including but not limited to those included in “genome instability” and/or “inflammation” [93]. Additionally, mitochondria from different tissues in a single individual may work differently and demonstrate different rates of ROS production and oxygen consumption [68, 69, 72].

In the blood of aging Labrador retrievers, the total antioxidant potential did not change with age, but DNA damage increased with age [16]. In aging beagles, there is an increase in serum and brain lipid peroxidation damage with age, an increase in brain protein carbonyls with age, and a decrease in glutathione function with age [58], suggesting that it is a lack of antioxidant function that leads to the accumulation of damage. However, others have measured SOD activity in blood and found a positive correlation with age, though using a fairly small sample size [136]. In leukocytes from beagles, complex III cytochrome c oxidoreductase activity increased with age, and DNA damage (8-OHdg) and protein carbonyls decreased with age [110]. In Labradors, there was an increase in 8-OHdg in the blood with age [1], highlighting potential differences across breeds. Additionally, mixed-breed dogs of similar age and size showed that blood lipid peroxidation levels were higher in male dogs compared with female dogs [135]. Using separated blood to measure total antioxidant capacity, lipid damage, and the activities/concentrations of CAT, GPx, and SOD from N = 66 large breed dogs, N = 16 medium-size breed dogs, and N = 139 small breed dogs of varying ages, it was shown that lipid damage increases with age in the domestic dog and decreases with body size [70], and small dogs have lower antioxidant concentrations [98]. This is a paradoxical finding, as the free radical theory of aging states that shorter-lived (larger breeds) should accrue more oxidative damage. Though, it remains unclear whether this accumulation is due to increased rates of lipid damage or decreased rates of clearing said damage [70].

Using primary fibroblast cells from N = 221 small and large breeds of young and old age, reduced glutathione (GSH), reactive species (RS) production, mitochondrial content, lipid oxidative damage, and DNA oxidative damage (8-OHdg) were measured [69]. It was shown that older dogs, despite their size class, had a significantly higher GSH concentration compared with younger dogs [69]. However, this study also found a significant increase in RS production in female puppies,thus, a GSH increase later in life could be a way to help offset damage from increases in ROS production early on [69]. Surprisingly, we found no differences in average ROS production across size and age classes in primary fibroblasts of dogs, though it is common to assume that with age [48, 84]. But, longer-lived breeds may have a reduction in RS production when they are puppies [69].

Though the description above entails the most commonly measured aspects of oxidative stress, this process can affect or be affected by many other aspects of metabolism. For example, as metabolites of glucose react with amino acids in proteins and lipids through the Maillard reactions, an irreversible by-product known as advanced glycation end-product (AGE) is formed as the end stage of carbohydrate metabolism [10]. First, non-enzymatic glycation of protein results in an Amadori rearrangement which is a reversible reaction [10]. However, the Amadori product reaches equilibrium over weeks, but can still undergo poorly defined rearrangements and irreversibly form AGEs [87]. AGEs can accumulate with age particularly in long-lived proteins such as collagens and crystallins and their formation, especially when it enters the circulation, renders irreversible damage to all biological macromolecules [10, 22]. The Maillard theory of aging proposes that the slow and continuous accumulation of AGEs may be a factor in aging rates [22]. The majority of AGEs that accumulate with age are glycoxidation products, formed by glycation and oxidation reactions from glucose or ascorbate [9]. Through a complex series of reactions, oxidative stress may be involved in AGE formation, and AGEs may, in turn, induce oxidative stress [10, 102]. However, BSA-AGE concentration was measured in the blood of domestic dogs, but no correlations were found between BSA-AGE and the body size or age of the dogs [73].

The dog aging literature is vastly lacking information regarding metabolic alterations downstream from IGF-1, PGC1-α, mitochondrial mutations with age, mitochondriogenesis, electron transport chain (ETC) complex activities, mitochondrial morphology, particularly related to fusion and fission dynamics with age, and/or uncoupling protein structure and function.

Cellular senescence

Cellular senescence is a term originally used to describe the limited replicative capacity of human fibroblast cells [57]. It has since been generally defined as a stable arrest of the cell cycle [93, 121]. Biologically, it can be measured in different ways, with DNA damage being one of the most commonly used biomarkers, though not all tissues in an aging organism will demonstrate similar patterns of aging/DNA damage [93]. DNA damage marker γH2AX, together with cell cycle inhibitor p21, have been used as senescence markers in dogs [55]. In 2019, Sándor and Kubinyi wrote that little was known about this hallmark of aging in dogs. To date, there has been some progress in this field, though, not much. Consideration of cellular senescence in dogs should be taken with a multi-breed approach and across different ages of each breed to further probe this hallmark of aging. Because collecting tissue from client-owned dogs is exceedingly difficult, the use of primary cell lines could be exploited here (e.g., [69, 71, 74, 75, 78, 95, 106].

Using primary fibroblast from 1- and 7-year-old Gyeongju Donggyeong dogs, a rare Korean breed, older dogs demonstrated increased β-galactosidase activity and increases in expression of p21, p53, and p16 transcripts compared with younger dogs [80], all of which are biomarkers of cellular senescence. In liver tissue, p21 correlated with age in 15 healthy dogs [82]. However, using a larger sample size of 51 dogs, γH2AX and p21 in liver tissue did not seem to demonstrate correlations with age in healthy dogs [55]. In eye tissue (retina, cornea, and lens), γH2AX and p21 did not demonstrate a correlation with age either, though sample sizes for both young and old dogs were fairly small (N = 10 young, N = 9 older dogs) [97]. The problem with small sample sizes when considering questions of aging in dogs is that there cannot be strong consideration of breed-specific health issues.

Stem cell exhaustion

In the developing and adult organisms, stem cells are still relied upon for the maintenance of tissues; thus, the abundance and replicative capacity of these cells are the rate liming step towards fixing accumulated damage [121]. A decrease in the function of hemopoietic tissue is said to be linked to immunosenescence, for example [93]. Most of the recent work with respect to this hallmark of aging in dogs relates to mesenchymal stem cells, their growth, potential for differentiation into other cells, and potential use for therapy (i.e., [15]. However, other mesenchymal stem cell populations have not been explored, even in basic terms. Additionally, the Yamanaka factors, such as Oct3/4, Sox2, Klf4m, or c-Myc, which are transcription factors that play a role in creating induced pluripotent stem cells, have not been examined across breeds and ages of dogs [65].

For example, because muscle is a post-mitotic, multinucleated tissue (syncytium), any postnatal muscle growth in mammals likely occurs via hypertrophic growth, which is an increase in muscle cell size. That is, new nuclei are drawn into the fiber itself from a population of multipotential mesenchymal stem cell population with the ability to undergo myogenesis or alternative transdifferentiation programs called satellite cells (SCs) found in the basement membrane of muscle fibers, which seem to be limited in number and/or proliferative potential for an animal’s lifespan [45]. SCs can proliferate into existing fibers to maintain a nearly constant nuclear-to-cytoplasmic ratio [66] and aid in the repair of the fiber [45]. Increases in IGF-1 expression in the pectoralis are associated with SC proliferation and differentiation and may lead to hypertrophy [2] and differing rates of protein synthesis [28]. Thus, IGF-1 signaling may be a temperature-sensitive determinant for muscle growth, making questions surrounding muscle structure and function in domestic dogs even more interesting.

Altered cellular communication/Inflammation

López-Otín et al. [93] first included inflammation as a hallmark of aging as part of the alterations to cellular communication during aging. Schmauck-Medina et al. [124] further expanded this hallmark to include a more direct description of “inflammaging.” A detailed description of this hallmark of aging in dogs was not originally included in Sándor and Kubinyi’s study [121]; thus, this section will include dog studies preceding 2019, along with a detailed description of this hallmark of aging.

Age-related physiological changes to the immune system are considered the main culprit of comorbidities in old age and are often associated with reduced life expectancy and survival across the few species in which this phenomenon has been explored, namely, humans, mice, and rats [27, 42, 129]. As age increases in most mammals, the reduced response of the immune system also leads to a reduced ability to deter infection and an increase in rates of autoimmune and neoplastic diseases [4]. Chronic inflammation is now commonly accepted as a major underlying condition of many age-related diseases, including cardiovascular disease, arthritis, and cancer, to name a few [24, 31], and disease progression seems to be a causative factor linked to the inflammatory process [23]. Inflammaging refers to the increase in markers of pro- and anti-inflammatory activity that occurs with mammal age [43]. It is further defined as low-grade, chronic, and systemic inflammation with increasing age in the absence of infection [43]. Functionally, the aging process, mainly due to the accumulation of cellular damage, activates inflammatory molecules, including cytokines, inevitably resulting in systemic inflammation [8], especially from the upregulation of pro-inflammatory cytokines such as interleukins(IL)-1β and -6 [42, 44]. All of these molecules form part of the immune system’s innate inducible defense, and their release is normally triggered during immune challenges [88, 148].

Chronic inflammation entails several cytokines, molecular pathways, effector cells, and tissue responses that appear to be shared across multiple age-related diseases, the details of which are beyond the scope of this mini-review [43]. Humans demonstrate an increase in IL-6 and tumor necrosis factor α (TNF-α) levels with increased age and a decrease in IGF-1 levels with increasing age [145]. IL-6 concentrations can predict mortality and survival in centenarians with arthritis [8, 143, 145]. Chronic inflammation due to high levels of TNF-α and IL-1β is critical in the development of cardiovascular diseases and is a significant component of human and murine aging. Proinflammatory cytokines IL-1β, TNF-α, and IL-6 act both in paracrine and autocrine manners by developing insulin resistance, and they can interfere with insulin signaling, lipid, and protein synthesis, potentially inducing metabolic disorders, including diabetes [92, 130]. Information regarding pro- and anti-inflammatory cytokine modulation in aging dogs is not vast within the literature,however, using primary fibroblast cells from young and old dogs of small and large breeds, IL-6 concentrations were measured before and after treatment with lipopolysaccharide (LPS, which is an immunostimulatory agent that initiates synthesis and release of proinflammatory cytokines, including IL-6 [3]. Younger dogs of both size classes demonstrated a drastic amplification of IL-6 after LPS treatment [75], unlike geriatric mice and humans [47, 132]. In contrast, background/control IL-6 data from isolated primary fibroblasts of young and old dogs does follow the patterns of geriatric humans and mice when considering just the large dog size class, such that fibroblasts from geriatric large dogs have an increase in background IL-6 compared with cells from their younger counterparts [75].

At basal levels, using serum from dogs of different sizes and ages, a significant decrease in IL-6 concentrations in young dogs was found, similar to humans [77]. However, only young dogs have decreased IL-6 concentrations, with adult dogs having similar IL-6 concentrations to senior and geriatric dogs, implying differences in aging rates between humans and dogs [77]. And, there was a marginally significant interaction between sex and spayed or neutered status and IL-1β concentrations, with intact females having the lowest IL-1β concentrations compared with intact males and spayed and neutered dogs [77].

Other aspects of immunosenescence have been explored in the dog aging literature. Most studies to date on immunosenescence in the dog use one breed across different ages to generalize patterns of the whole species (summarized in [32] and [112]. For example, white blood cell counts and lymphocyte subset distribution in a population of aging Labrador retrievers found a decline in the absolute number of lymphocytes, including T cells and CD4 and CD8 cells [17, 50]. Lymphocyte shift with age happened most dramatically in females compared with male Labradors, where B cells decreased and T cells increased with age [49, 50, 59]. In another population of Labrador retrievers, serum levels of immunoglobulin M, the first antibody to appear during an exposure, and 8-hydroxy-2-deoxyguanosine (8-OHdG), a type of oxidative damage (described above), increased with age [1]. In beagles, there were negative correlations between dogs’ ages and the number of peripheral blood mononuclear cells, T cells, and B cells. Specifically, the number of naive CD4 + CD45RA + T cells and CD8 + CD45RA + T cells significantly decreased with age [46]. In fact, within the first 15 months of age, it has been demonstrated that the maturation of lymphocyte subsets that occur during early life in beagles occurs [139]. Another population of beagles demonstrated a significant decline with age in the percent of bacterial killing by neutrophils [53]. In German shepherds, natural killer-cell activity and the serum cytokine-like activities of IL-1 were significantly higher in old female dogs. Additionally, white blood cell counts, peripheral blood lymphocytes, lymphocyte proliferative activity, and interleukin-2 (IL-2) serum concentrations were significantly lower in older German shepherds. But γ-globulin concentrations were significantly higher in the elderly dogs [129].

Immunosenescence seems to be linked to decreases in thymic output, which leads to a decrease in T cell output to the lymphocyte pool and reduced diversity in the T cell repertoire, leading to a reduced ability to deal with novel pathogens [126]. A way to quantify thymic output across ontogeny is to measure the signal joint T cell receptor excision circle (sj-TREC) in the blood. In large breeds, shorter-lived dogs such as Labrador retrievers, Burnese mountain dogs, Great Danes, and Dogue de Bordeaux, there was a decrease in sj-TREC values with age. There was an association between mean breed lifespan and age of beginning of decline, with shorter-lived breeds having an earlier-in-life decline compared with longer-lived breeds such as Jack Russell terriers and Yorkshire terriers [62].

Microbiome

That the population of microbes in animals’ body is linked to that animals’ lifestyle and potential well-being is now an active area of research, and correlations between the microbiome and aging are starting to appear [19, 121, 124]. The gut microbiome is the most commonly studied; however, exploration of the skin microbiome and the blood microbiome is also starting to increase [123].

In the dog gastrointestinal tract five, microbial phyla seem to dominate: Firmicutes, Fusobacteria, Bacteroidetes, Proteobacteria, and Actinobacteria [33, 118, 147]. Generally, there seems to be a negative association between body weight and Proteobacteria relative abundance in dogs [33], however, this may be due to breed-related differences in the gut microbiome. Fecal microbiome analysis in dogs has shown that there may not be breed differences in microbial evenness when 8 different breeds of dogs are compared. However, the types of bacteria did differ across breeds with, Maltese showing abundant Fusobacteria and poodles (no body weight shown to determine which poodle—toy, miniature, or standard showing Firmicutes and Actinobacteria with greater abundance [147]. In another study looking at the gut microbiome of three small breeds (Maltese, poodle, and miniature schnauzer), it was found that the relative abundance of Firmicutes was significantly lower in the Maltese than that in the other two breeds, whereas the abundance of Fusobacteria was significantly higher in the Maltese than in the Miniature Schnauzer [115]. At the genus level, the relative abundance of Streptococcus, Fusobacterium, Turicibacter, Succinivibrio, and Anaerobiospirillum differed significantly among the three dog breeds [115]. There was an age-dependent increase in Fusobacterium perfoetens across breeds [86, 147], which may be linked to intestinal diseases, including colorectal cancers and tumors [40].

Summary

While the field of canine aging has greatly progressed in some areas in the last 5 years, there are other areas that remain largely not explored. Table 1 highlights the biomarkers of aging that are missing or are currently not explored in the dog aging model. Because the domestic dog is such a diverse species, studies considering dog aging must push to include large sample sizes including multiple breeds and ages within those breeds. Importantly, a 2-year-old Chihuahua is not the same biological age as a 2-year-old Great Dane due to their differing aging rates. That is to say, because of the differences in aging rates between small and large dog breeds, a Chihuahua and a Great Dane have different aging trajectories that cannot be listed as similar. Thus, corrections to aging rates should be considered [70].

Table 1.

Using the “hallmarks of aging,” this table highlights biomarkers within each hallmark that are missing or not currently exhaustively explored in the dog aging literature. This list is not comprehensive, but instead presents options for the field to consider so that aging in dogs can be holistically investigated in the future

Hallmark of aging	Missing in the dog aging literature
Genomic instability	• Age at which mitochondrial mutations appear across breeds • Information regarding nucleotide alterations in excision repair, nonhomologous end-joining, mismatch repair, and translesion synthesis in aging dogs
Telomere dynamics	• Telomere dynamics in tissues other than white blood cells, including multiple breeds across different ages
Epigenetic alterations	• More information about sirtuin function with age across multiple breeds and tissues • Any studies considering specific histone modifications (i.e., H4K16ac, H4K20me3, H3K4me3, H3K9me, and H3K27me3) or specific biomarkers of chromatin remodeling such as HP1α or NuRD across multiple breeds and ages
Loss of proteostasis and autophagy	• Multi-breed across ages comparison of HSPs • Sarcopenia across multiple breeds and ages • Detailed macroautophagic pathways, Hsc70, HSF-1, and proteasomal degradation across multiple breeds and ages
Deregulated nutrient sensing	• IGF2 across multiple breeds and ages • Lipidomics, including measuring the activities of lipid metabolism related enzymes across multiple breeds and ages
Mitochondrial dysfunction	• Uncoupling activity across breeds and ages • Metabolic alterations downstream from IGF-1, PGC1-α, mitochondrial mutations with age, mitochondriogenesis, ETC complex activities, mitochondrial morphology particularly related to fusion, and fission dynamics with age • Cellular resistance across multiple breeds and ages
Cellular senescence	• DNA damage in metabolically active tissues across ages and multiple breeds • Senescence-associated β-galactosidase (SABG) to identify senescence in metabolically active tissues across multiple breeds and ages • P53 function across different tissues, ages, and in multiple breeds
Stem cell exhaustion	• Function, activity, and exhaustion of satellite cells (SCs) within muscle tissue across ages and breeds • Exhaustion of hematopoietic stem cells (HSCs)
Inflammation/altered cellular communication	• Immunosenescence across breeds and ages
Microbiome	• Multi-breed across age approach is lacking in measuring gut microbiome • Blood and skin microbiome are lacking • Nasal microbiome

Abbreviations

CNV

Copy number variant

PBMC

Peripheral blood mononuclear cell or leukocyte

SIRT1

Sirtuin 1

IGF-1

Insulin growth factor-1

H4K16ac, H4K20me3, H3K4me3, H3K9me, H3K27me3

Different types of histone modifications

HSP

Heat shock protein

Hsp-70

Heat shock protein 70

βA

Beta amyloid

Hsc70

Heat shock cognate 70

HSF-1

Heat shock factor 1

GH

Growth hormone

CR

Calorie restriction/restricted

ROS

Reactive oxygen species

OS

Oxidative stress

GPx

Glutathione peroxidase

SOD

Super oxide dismutase

CAT

Catalase

8-OHdg

A type of oxidative DNA damage

GSH

Reduced glutathione

AGE

Advanced glycation end-product

PGC1-α

Peroxisome proliferator-activated receptor-gamma coactivator 1 alpha

ETC

Electron transport chain

γH2AX

DNA damage marker

SCs

Satellite cells

TNF-α

Tumor necrosis factor α

IL-1β

Interleukins(IL)-1β

IL-6

Interleukins(IL)-6

LPS

Lipopolysaccharide

sj-TREC

Signal joint T cell receptor excision circle

Author contribution

AGJ reviewed the literature and wrote the manuscript.

Data availability

There is no raw data included within this manuscript.

Materials availability

For any inquiries, please email ajimenez@colgate.edu.

Declarations

Competing interests

The author declares no competing interests.