Cellular metabolic pathways of aging in dogs: could p53 and SIRT1 be at play?

Abstract

Aging and cancer seem to be closely associated, such that cancer is generally considered a disease of the elderly in both humans and dogs. Additionally, cancer is a metabolic shift in itself towards aerobic glycolysis. Larger dog breeds with shorter lifespans, and increased glycolytic cellular metabolic rates, die of cancer more often than smaller breeds. The tumor suppressor p53 factor is a key suppressor oncogene, and the p53 pathway arrests cellular proliferation and prevents DNA mutations from accumulating during cellular stress. The p53 pathway is also associated with the control of cellular metabolism to prevent cellular metabolic shifts common to cancerous phenotypes. SIRT1 deacetylates the p53 tumor suppressor protein, downregulating p53 via effects on stability and activity during stress. Here, we used primary fibroblast cells from small and large puppies and old dogs. Using UV radiation to upregulate the p53 system (100 J/m²), control cells and UV-treated cells were used to measure aerobic and glycolytic metabolic rates using a Seahorse XFe96 oxygen flux analyzer. We also quantified p53 expression and SIRT1 concentration in canine primary fibroblasts before and after UV treatment. We demonstrate that, due to a higher p53 nuclear to cytoplasmic ratio in large breed dogs after UV treatment, p53 could have a more regulatory effect on large breed dogs’ metabolism compared with smaller breeds. Thus, there may be a link between p53 upregulation and inhibition of glycolysis in large breed dogs during times of cellular stress compared with small breed dogs. However, SIRT1 concentrations decrease with age in domestic dogs of both size classes, suggesting a possible release of inhibition of p53 through the SIRT1 pathway with age. This may lead to increased incidences of cancer, especially due to the more pronounced upregulation of p53 with cellular stress.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11357-023-00942-y.

Introduction

Large breed dogs, along with shorter lifespans, tend to have an increased propensity for cancer [22, 26, 44]. Sixteen percent of dogs sampled in the UK died of cancer, with large breed dogs being overrepresented in those with a cancer diagnosis [32]. In bullmastiff breeds, 47% of all deaths were due to cancer [3]. Tumor likelihood increases with age, and malignant tumors are more likely than benign tumors in dogs, where larger breeds have a higher prevalence of malignant tumor formation ([12, 46]. The fact that many large breeds have higher incidences of specific types of cancer suggests that there are inherent and heritable physiological traits that are placing these breeds into a more cancerous phenotype [44], though smaller dogs also see some incidences of cancer [13]. The emerging hallmarks of cancer include an energetic shift change in cells and a decrease in tumor suppressor genes, like p53 [19], an unexplored pathway in the cellular physiology of dogs that may be linked to higher incidences of cancer in larger breeds of dogs.

The cancerous phenotype is characterized by a cellular shift to increased aerobic glycolysis and lactate production regardless of oxygen availability [43]. This phenotype is called the Warburg effect [19, 36]. This metabolic switch to aerobic glycolysis seems counterintuitive, considering that the glycolytic process is significantly less energetically efficient compared with oxidative phosphorylation in the mitochondria [30]. However, cancer cells are known to make up for this energetic efficiency by first upregulating glucose transporters to increase the amount of glucose in cells [19]. At the cell level, using primary dermal fibroblast cells from puppies and senior dogs of small and large breeds, previous work has repeatedly found that larger breed dogs have significantly higher glycolytic phenotypes compared with smaller breeds [23, 24]. Additionally, in older dogs, shorter-lived breeds have significantly higher glycolytic phenotypes [23]. However, more current work suggests that the increase in glycolytic phenotype in large breed dogs is not associated with an increase in lactate formation but does demonstrate a deficiency in the lower half of glycolysis in large breed dogs [6]. This cellular phenotype may be related to increased cancer rates in larger breed dogs.

Cancer rates in dogs seem to be associated with an increase in DNA mutations which lead to changes in the expression patterns of oncogenes and tumor suppression genes [12]. One of the most common mutations associated with human cancers is a p53 gene mutation [41, 54], present in more than 50% of human tumors [12, 30], and some canine tumors as well [2]. The tumor suppressor p53 factor is a key suppressor oncogene, and the p53 pathway arrests cellular proliferation and prevents DNA mutations from accumulating during cellular stress [12, 45, 54]. Additionally, if DNA damage cannot be repaired after the upregulation of p53, then p53 upregulates apoptosis pathways for the damaged cell [25]. Lifespan can also be associated with p53 activity since preventing tumors in early life through this pathway is important for longevity [12]. The efficiency of the p53 system seems to decline with age, which provides a mechanistic link to higher cancer rates in the elderly [12]. p53 basal concentration in normal, non-cancerous tissues could inform the likelihood of developing cancer and inform variations in higher rates of cancer for some dog breeds [14].

Other proteins help regulate p53 function within cells. For example, SIRT1 is a NAD⁺-dependent histone deacetylase with several protein substrates [53] and a key epigenetic regulator [48]. SIRT1 is localized to the nucleus and deacetylates multiple non-histone proteins including p53 in response to stress [17, 29, 42]. It also serves to regulate DNA repair, apoptosis, mitochondrial biogenesis, and cell stress responses, to name a few [53]. Mammalian SIRT1 has also been shown to affect downstream signaling, such as insulin-growth factor (IGF) signaling, which has long been linked to lifespan regulation, and to increase energy availability [29]. In response to DNA damage, p53 protein levels are upregulated, though its activities are brought on by its post-translational phosphorylation and acetylation which increase protein stability (Dai and Gu [10]). This increased protein stability with phosphorylation also increases p53’s affinity for p300/CBP acetylase, and, thus, deacetylation by SIRT1 can modulate the function of p53 [49]. If SIRT1 activity is linked to cancer prevention in dogs, it may serve to limit proliferation and decrease DNA damage [42]. In primary cells, SIRT1 promotes cell survival by initiating cell cycle arrest rather than apoptosis after DNA damage through deacetylation of p53 [42]. Expression of wild-type SIRT1 in human cells reduces the transcriptional activity of p53 [49]. The interplay between SIRT1 and p53 and their reciprocal regulation of expression may be an important regulator of decreased replicative senescence when upregulated [42].

The relative genetic homogeneity of dog breeds allows for ease in genetic studies; however, this same quality increases the likelihood of germline mutations being passed on easily. In fact, large breed dogs tend to have higher coefficients of inbreeding compared with smaller breeds [52]. Thus, even at basal levels, some of the cellular metabolic differences we see in dogs, especially large breeds, may be associated with p53 and SIRT1 function. Exploring this pathway further may lead to a greater understanding of its role in dog metabolism, aging, and mechanistic links to cancer rates and potentially provide new avenues for cancer prevention and treatment in dogs. Here, at the cell level, we tested for changes in aerobic and anaerobic cellular metabolism, p53 expression, and SIRT1 concentrations in primary fibroblasts of small and large breeds both young and old before and after UV treatment.

Materials and methods

Isolation of dog primary fibroblasts

We isolated primary fibroblast cells from large breed puppies (N = 47), small breed puppies (N = 40), large breed senior dogs (N = 36), and small breed senior dogs (N = 13), for a total of N = 136 individuals for the cellular metabolism experiments. The small breed size class was composed of breeds with an adult body mass of 15 kg or less. The large breed size class included breeds or mixes with an adult body mass of 20 kg or more. These size classes are based on American Kennel Club (AKC) standards for each breed and are described in Jimenez et al. [23]. Table 1 provides a summary of the data collected about the dogs in the study, i.e., information about breed, sex, body mass, age, and reason for euthanasia.

Table 1.

Information about breeds, sample sizes, sex, body mass, and reason for euthanasia for dogs included in this study

Breed	Sample size	Sex/fixed	Mean weight (kg)	Mean age	Size class	Age class	Reason for euthanasia
American pitbull terrier	1	Male neutered	38.55532	10.41667 yrs	L	Old	Cancer
Australian/blue heeler mix	1	Female spayed	30.7535376	6.91667 yrs	L	Old	Cancer
Beagle mix	1	Female spayed	24.94756	14.5 yrs	L	Old	Old age issues
Bernese mountain dog	2	1 female, 1 male neutered	48.62506	8.5 yrs	L	Old	Cancer
Boxer	2	Female unspayed	30.61746	10.125 yrs	L	Old	Cancer, old age, dementia
Boxer mix	2	Female spayed, male	27.442316	10.874835 yrs	L	Old	Mobility decline, cancer
Boxer/American bulldog	1	Female spayed	43.544832	11.5 yrs	L	Old	Unknown
Boxer/shepherd	1	Male	27.669112	13 yrs	L	Old	Cancer
Collie mix	1	Male	31.297848	12.41667 yrs	L	Old	Cancer
Dalmatian	1	Female spayed	39.008912	10.25 yrs	L	Old	Kidney failure
Dalmatian mix	1	Male neutered	36.28736	13 yrs	L	Old	Cancer
English bulldog	1	Female spayed	24.040376	11.0833 yrs	L	Old	Heart failure
German shepherd	2	Female spayed, male	36.514156	12.625 yrs	L	Old	Mobility decline, cancer
German shepherd mix	1	Female	24.040376	15.66667 yrs	L	Old	Cancer
Golden doodle	1	Spayed female	30.6628192	9.3333 yrs	L	Old	Cancer
Golden retriever	4	2 female spayed, 2 males	34.0647592	12.25 yrs	L	Old	Mobility decline, chronic severe bacterial otitis
Husky	1	Male neutered	37.648136	6.91667 yrs	L	Old	Cancer
Husky mix	1	Male	19.958048	15.5833 yrs	L	Old	Mobility decline
Labrador mix	1	Female	20.41164	14.16667 yrs	L	Old	Mobility decline
Labrador retriever	4	2 female spayed, male, unneutered male	36.740952	12.395825 yrs	L	Old	Cancer, pulmonary neoplasia, mobility decline
Labrador/poodle mix	1	Spayed female	48.987936	13.25 yrs	L	Old	Mobility decline
Mixed breed	1	Male	36.28736	15 yrs	L	Old	Mobility decline
Pitbull mix	1	Female spayed	37.194544	10.5 yrs	L	Old	Liver failure
Rottweiler mix	1	Unneutered male	31.75144	13.5 yrs	L	Old	Mobility decline
Springer spaniel	2	2 males	24.8114824	13.124835 yrs	L	Old	Cancer, splenic neoplasia
Australian cattle dog	3	2 females, 1 male	0.44452016	3 days	L	Puppy
Beauceron	3	2 females, male	5.443104	49 days	L	Puppy
Bernese mountain dog	4	3 females, 1 male	N/A	2 days	L	Puppy
Bouvier des Flanders	3	1 female, 2 males	0.695507733	4 days	L	Puppy
Brittany spaniel	4	1 female, 3 males	0.285621288	3 days	L	Puppy
Cane corso	4	3 females, 1 male	0.771815275	4 days	L	Puppy
Doberman	3	2 females, 1 male	0.5443104	4 days	L	Puppy
Great dane	4	1 female, 3 males	0.680388	2 days	L	Puppy
Large musterlander	4	2 females, 2 males	0.528009438	2 days	L	Puppy
Old English sheepdog	1	Female	0.2721552	1 day	L	Puppy
Polish lowland sheepdog	3	1 female, 2 males	N/A	3 days	L	Puppy
Samoyed	4	1 female, 3 males	0.396893	1 day	L	Puppy
Spanish waterdog	3	1 female, 2 males	0.219236193	3 days	L	Puppy
Standard schnauzer	4	3 females, 1 male	0.283495	3 days	L	Puppy
American eskimo dog	1	Female spayed	8.0739376	15.667 yrs	S	Old	Mobility decline
Beagle	1	Female	8.164656	16 yrs	S	Old	Mobility decline
Beagle mix	1	Female	11.24908	16.667 yrs	S	Old	Mobility decline
Chihuahua	1	1 unneutered male	5.0802304	11.91667 yrs	S	Old	Heart failure
Maltese	2	2 female spayed	4.3998424	15.5415 yrs	S	Old	Liver failure, old age
Poodle/Pekignese mix	1	Female spayed	6.2595696	13.91667 yrs	S	Old	Diabetic or kidney disease and dementia
Pug	1	Female spayed	10.4326	14.5 yrs	S	Old	Cancer
Shetland sheepdog	3	3 males	11.73291307	12.4443333333333 yrs	S	Old	Abdominal neoplasia, complications with Cushing’s disease, cancer
West highland white terrier	1	1 male	9.979024	15.1667 yrs	S	Old	Cancer
Yorkshire terrier	1	1 female	2.26796	17 yrs	S	Old	Chronic kidney disease
Boston terrier	4	2 females, 2 males	0.340194	5 days	S	Puppy
Chihuahua	5	2 females, 3 males	0.1632931	1 day	S	Puppy
French bulldog	4	1 female, 3 males	0.3572037	3 days	S	Puppy
Miniature Australian shepherd	4	2 females, 2 males	0.26535132	3 days	S	Puppy
Miniature poodle	3	1 female, 2 males	0.1984465	5 days	S	Puppy
Miniature schnauzer	3	2 females, 1 male	N/A	3 days	S	Puppy
Parson Russell terrier	4	3 females, 1 male	0.170097	4 days	S	Puppy
Pembroke welsh corgi	4	4 females	0.447213363	4 days	S	Puppy
Soft coated wheaten terrier	3	1 female, 2 males	0.316569417	2.66666666666667 days	S	Puppy
West highland white terrier	3	N/A	0.2055339	4 days	S	Puppy
Yorkshire terrier	3	3 females	0.1572452	4 days	S	Puppy

Puppy samples were obtained from routine tail docks, ear clips, and dewclaw removals performed at veterinarian offices in the USA. Senior dog samples were collected from ear clips immediately after euthanasia. Owners consented to the collection of tissue for research. The samples were placed in cold transfer media (Dulbecco’s modified Eagle medium [DMEM], with 4.5 g/L glucose, sodium pyruvate, and 4 mM L-glutamine supplemented with 10% heat-inactivated fetal bovine serum and antibiotics [100 U/mL pen/strep], containing 10 mM HEPES) and transferred to Colgate University on ice.

To isolate primary fibroblast cells, skin samples were sterilized in 70% ethanol and 10% bleach. Once any fat and bone were removed, the skin was minced and incubated in sterile 0.5% collagenase type 2 (Worthington Chemicals, Cat. No. LS004176) overnight in an atmosphere of 37 °C, 5% CO₂, and 5% O₂. After incubation, the collagenase mixture was filtered through a 20-μM sterile mesh and centrifuged at 1000 rpm for 5 min. The resulting supernatant was removed, and the pellet was resuspended in 4 mL of mammal media (Dulbecco’s modified Eagle medium [DMEM], with 4.5 g/L glucose, sodium pyruvate, and 4 mM L-glutamine supplemented with 10% heat-inactivated fetal bovine serum and antibiotics [100 U/mL pen/strep]). Cells were grown in Corning T-25 culture flasks at 37 °C in an atmosphere of 5% O₂ and 5% CO₂. When cells reached 90% confluence, they were trypsinized (0.25%) and cryopreserved at 10⁶ cells/mL in DMEM supplemented with 40% fetal bovine serum and dimethylsulfoxide (DMSO) at a final concentration of 10%. We stored cells in liquid N₂ prior to any experiments.

To resuspend cell lines, individual aliquots were thawed by continuously swirling the frozen aliquot in a 37 °C water bath until only a small amount of ice remained. We resuspended the pellet by pipetting 6 mL of chilled mammal media and plating the resuspension in a T-25 culture flask at 37 °C in an atmosphere of 5% O₂ and 5% CO₂. Once confluent, cells from each individual dog were removed from the T-25 flask by using 0.05% trypsin, followed by pelleting by centrifugation. Then, cells were manually counted before being plated on seahorse plates for metabolic experiments and on plus slides for immunohistochemistry of p53. All seahorse and p53 immunohistochemistry experiments highlighted below were done at passage 2 (P2), and SIRT1 concentration was measured on cells at passage 3 (P3).

All the procedures within this study were approved by Colgate University’s Institutional Care and Use Committees under protocol number 1819–13.

Treatments

We manually counted cells and plated 10,000 cells/well. We allowed cells to attach overnight at P2 before running seahorse experiments. We plated separate control oxygen consumption rate (OCR-aerobic profile) and extracellular acidification rates (ECAR-glycolytic profile) plates. Additionally, we plated another OCR and another ECAR plate for UV treatments. Control plates were run first. Then, we removed the media and UV-treated the second set of plates with 100 J/m² of UV using a VWR UV crosslinker. We allowed cells to recover from UV treatment for 1 h in seahorse running media (as described below) at 37 °C in a non-CO₂ incubator prior to running OCR and ECAR experiments.

Measuring cellular metabolism

OCRs were determined using XFe96 FluxPaks from Agilent Technologies. We measured OCRs after cells were equilibrated to running media, which contains 10 mM glucose, 1 mM sodium pyruvate, and 2 mM glutamine, pH = 7.4, for 1 h in a non-CO₂ incubator at 37 °C. Baseline measurements of OCRs were made three times prior to injecting a final well concentration of 2 μM oligomycin, which inhibits ATP synthesis by blocking the proton channel of the F_o portion of the ATP synthase. This baseline is used to distinguish the percentage of O₂ consumption devoted to ATP synthesis and the O₂ consumption required to overcome the natural proton leak across the inner mitochondrial membrane plus any non-mitochondrial O₂ consumption. We then injected a final well concentration of 0.125 μM carbonyl cyanide-4-(trifluoromethoxy) phenylhydrazone (FCCP), an uncoupling agent that disrupts ATP synthesis by collapsing the proton gradient across the mitochondrial membrane, leading to uncoupled consumption of energy and O₂ without generating ATP, providing a theoretical maximal respiratory rate. Finally, we injected a final well concentration of 0.5 μM antimycin A, a complex III inhibitor, and rotenone, a complex I inhibitor. This combination stops mitochondrial respiration and enables non-mitochondrial respiration to be evaluated (Gerenser et al. 2009; Brand and Nichols 2011; Hill et al. 2013). After the measurements were completed, we used a 1:200 concentration of CyQUANT dye to quantify the final counts of cells in each well (Jones et al. 2001) and normalized all rates to 20,000 cells.

ECAR was measured in units of mpH, which is the pH change in the media surrounding the cells due to proton flux in glycolysis. Measurements of ECAR were performed after the cells were equilibrated to running media for 1 h in a non-CO2 incubator at 37 °C. Running media contained no glucose and 2 mM L-glutamine in all experiments, pH = 7.4. Baseline rates were measured three times prior to any injections. We first injected a final well concentration of 10 mM glucose into the media surrounding cells, which provides a measure of glycolytic rate, and then injected a final well concentration of 2 μM oligomycin, giving an estimate of glycolytic capacity in cells. Finally, we injected a final well concentration of 50 mM 2-DG, a glucose analog that inhibits glycolysis, providing an estimate of non-glycolytic acidification (Hill et al. 2003). After measurements were completed, we used a 1:200 concentration of CyQUANT dye to quantify the final counts of cells in each well (Jones et al. 2001) and normalized all rates to 20,000 cells.

Immunohistochemistry of p53

We used only a subset of all of the cell lines listed above to stain for p53 (N = 33 large breed old dogs; N = 15 large breed puppies; N = 9 small breed old dogs; N = 13 small breed puppies). Two slides per individual cell line were grown on plus slides for 3 days in full media (as described above). One slide was used as a control, and another slide was UV-treated as above. Media was removed from slides prior to UV treatment, as above. Both control and UV-treated slides were then incubated at 37 °C in a 5% O₂ and 5% CO₂ incubator with 1 mL of FBS-free media for 1 h to define the cytoplasm of cells (data not shown) [20, 21]. Cells were then fixed using 4% paraformaldehyde in PBS pH 7.4 for 10 min at room temperature. The cells were washed three times with ice-cold PBS. To permeabilize the cells, slides were incubated for 10 min with PBS containing 0.25% Triton X-100. Slides were then washed in PBS three times for 5 min to remove excess Triton-X. To block unspecific binding of the antibodies, slides were incubated with 1% BSA and 22.52 mg/mL glycine in PBST (PBS + 0.1% Tween 20) for 30 min.

A subset of slides was incubated with a 1:100 concentration of Alexa Fluor® 488 anti-p53 antibody [SP161] (Abcam cat no. ab270123) in 1% BSA in PBST in a humidified chamber overnight at 4 °C. Another subset of the slides was incubated in primary anti-p53 antibody [PAb 240] (ab26) [54] at a concentration of 1 µg/mL in a humidified chamber overnight at 4 °C. The next morning, the solution was decanted, and the cells were washed three times in PBS, 5 min each wash. For cells stained with [PAb 240] (ab26), they were then incubated with the secondary antibody, goat anti-mouse IgG H&L (Alexa Fluor® 488) (ab150113) at a concentration of 1:200 in 1% BSA for 1 h at room temperature in the dark. All slides were then washed. The first wash included 300 nM of the nuclear probe 4′,6-diamidino-2-phenylindole (DAPI, Molecular Probes). The decision to use two antibodies was made due to the fact that (SP161) had not been validated in dog samples, but (PAb 240) had [54]. Since both stains showed similar staining patterns across our samples, the data collected for each were combined into one dataset.

All slides were cover slipped with glycerin. Slides were examined using a Zeiss 710 laser filter confocal microscope. Three pictures were taken per slide, one in the center of the slide and the other two on opposite sides. We used Zen 3.5 (blue edition) to analyze the mean intensity of the p53 nuclear stain as well as the mean intensity of this stain in the cytoplasm to estimate staining intensity between the nucleus and cytoplasm (referred to as the p53 nuclear to cytoplasmic ratio). We quantified p53 staining intensity in 90 nuclei and cytoplasm per picture, a total of 270 cells per treatment per dog. Sample images can be found in Figure S55.

SIRT1 ELISA

We used only a subset of all of the cell lines listed above to measure SIRT1 concentration (N = 35 large breed, old control and N = 33 large breed, old UV; N = 30 large breed, puppy control and N = 30 large breed, puppy UV; N = 11 small breed, old control and N = 11 small breed, old UV; N = 27 small breed, puppy control and N = 27 small breed, puppy UV).

We passaged the cells used for seahorse experiments to P2 and allowed them to grow to confluence in two T-25 flasks. Once confluent, one flask of each dog cell line was used as a control and another as an experimental flask. The media was removed from all flasks. Treatment flasks were exposed to 100 J/m² UV, as described above. All flasks received 4 mL of FBS-free media and were allowed to recover in an incubator at 37 °C, 5% CO₂, and 5% O₂ for 1 h. After 1 h, all cells were removed by trypsinization with 0.05% trypsin (P3), pelleted, and immediately frozen at − 80 °C until further processing.

We thawed the cells on ice, lysed the cells, and used the resultant lysate on a mouse SIRT1 ELISA kit (ab206983) as the immunogen sequence homology with canine is > 99%. We followed the manufacturer’s protocol to measure SIRT1 concentration and used a Tecan Infinite m200 plate reader at 450 nm to read plates. We also measured total protein concentration using a protein determination kit (Cayman Chemicals Cat. No. 704002) on the extracted cell supernatant and normalized all SIRT1 values to total protein concentration.

Statistics

Where necessary to meet the assumptions of an ANOVA, the data were transformed in a manner informed by a Box-Cox power transformation (Box and Cox, 1964). This meant a log transform for all variables except SIRT1/total protein concentration (ug/mL) and p53 nuclear to cytoplasmic ratio (a.u.). First, we analyzed all data in a single MANOVA to test for differences in the collection of dependent variables across body size (small or large, as stated above), age (young or old), treatment (control or UV), their two-way interactions, and their three-way interactions. Follow-up analyses were conducted using three-way ANOVAs on each dependent variable. The models are fit in R [38], and the mixed effects models are fit using the lmerTest package [27] for R. Best-subset model selection according to the Bayesian information criterion (BIC) [47] was completed via complete enumeration [35] for each model. BIC penalizes false positives more than false negatives when there are eight or more observations. We adjusted the p-values in the MANOVA analyses and within each model reported using the approach in Benjamini and Hochberg [4] rather than using the overly conservative Bonferonni adjustment [11]. This approach allows for the control of the expected proportion of false discoveries instead of the probability of making at least one false discovery, thus preserving power. Post hoc testing involved estimating marginal means for specified values of independent variables and comparing these means via statistical contrasts. We performed these post hoc tests via the emmeans package for R [40]. Results were considered significant if the corrected p-value was less than 0.05.

Limitations

Due to the nature of working with pet dogs, our sample collection was limited in several ways. First, the number of puppies per breed or group was limited to those breeds that are altered as a breed requirement (tail docks and dewclaw removal). The number of older euthanized dogs was limited due to the owners’ willingness to provide us with ear clips at the time of euthanasia. Secondly, we were not given complete medical charts for any dogs included in this study; thus, we have no records of diet or exercise, which may be variants in OCR and ECAR. At times, veterinarians did not provide sex or weight for puppies, either. We do know that none of the dogs included in this study were taking any medications, and none were obese, as we provided these requirements as collection criteria for veterinarians.

Results

Full details of statistical models and all MANOVA/ANOVA results can be found at https://doi.org/10.6084/m9.figshare.22220698.v1.

The repeated measures MANOVA showed significant differences for the multivariate analysis in age class (F = 3.72, p = 0.0067; Table S1). Thus, we probed each dependent variable separately using repeated measures ANOVA.

Results showed significant unique effects of age class on SIRT1/total protein concentration (F = 47.90, p < 0.0001; Fig. 1; Table S4). We found a significant effect of sex on p53 nuclear density (F = 9.2741, p = 0.0225; Table S11) and on p53 cytoplasmic density (F = 9.8259, p = 0.0170; Table S12). The ratio between p53 nuclear intensity and p53 cytoplasmic intensity demonstrated significant treatment by size class interaction (F = 7.88, p = 0.0525; Fig. 1; Supplementary Table 13). Additionally, we found a significant effect of treatment on non-glycolytic acidification (F = 9.6205, p = 0.0190; Table S2). But the treatment effect did not reach traditional levels of statistical significance for glycolysis, SIRT1/total protein concentration, total glycolytic capacity, basal OCR, proton leak, maximal respiration, ATP production, non-mitochondrial respiration, p53 nuclear intensity, or p53 cytoplasmic intensity through the unique or interaction effects (Fig. 1; Table S3-12).

Fig. 1.

Repeated measures ANOVA showed significant unique effects of age class on SIRT1/total protein concentration (L; F = 46.70, p < 0.0001). The ratio between p53 nuclear intensity and p53 cytoplastic intensity demonstrated significant treatment by size class interaction (K; F = 8.66, p = 0.0313). Additionally, the treatment effect did not reach traditional levels of statistical significance for glycolysis (A), total glycolytic capacity (B), non-glycolytic acidification (C), basal OCR (D), proton leak (E), maximal respiration (F), non-mitochondrial respiration (G), and ATP production (H) through the unique or interaction effects. Full-transformed models were fitted for glycolytic capacity, maximal respiration, non-mitochondrial respiration, and ATP production. We applied a log transformation to all variables. We found no significant treatment effects for glycolytic capacity, maximal respiration, non-mitochondrial respiration, or ATP production

It should be noted, however, that after UV treatment small breed puppies showed a mean log₂-ratio of − 0.5029, indicating a decrease in glycolysis, and a mean log₂-ratio of − 0.7398, indicating a decrease in glycolytic capacity. Large breed puppies, in contrast, showed a very small decrease, with a log₂-ratio of − 0.0271 indicating a decrease in glycolysis and a log₂-ratio of − 0.0343 indicating a decrease in glycolytic capacity after UV treatment. Large breed puppies showed increased glycolysis (log₂-ratio = 0.6384) and increased glycolytic capacity (log₂-ratio = 0.5756) compared with small breed puppies at control, as we have shown before [23, 24], and large breed old dogs showed increased glycolysis (log₂-ratio = 2.1099) and increased glycolytic capacity (log₂-ratio = 1.9861) compared with small breed old dogs at control, as we have shown before [23, 24]. The lack of statistical significance with size and age class within glycolytic phenotypes is likely due to a much smaller sample size within this study compared with the previous two,however, the metabolic signature that we have seen before still remains in place.

Full transformed model

Full models were fitted for glycolytic capacity, maximal respiration, non-mitochondrial respiration, ATP production, and p53 nuclear to cytoplasmic ratio. Similar to the ANOVA models, we performed transformations where necessary to ameliorate concerns with the assumptions of the linear mixed effects models. This meant a log transform for all variables except SIRT1/total protein concentration (ug/mL) and p53 nuclear to cytoplasmic ratio (a.u.).

We found no significant treatment effects for glycolytic capacity (Table S18), maximal respiration (Table S34), non-mitochondrial respiration (Table S41), or ATP production (Supplementary Table 48). However, there was a significant treatment effect for the p53 nuclear to cytoplasmic ratio (p = 0.0307; Table S60) and a size class x treatment interaction (p = 0.0092; Table S60), where after UV treatment, small breed puppies showed a decrease and large breed puppies showed an increase in the p53 nuclear to cytoplasmic ratio compared to control values (Fig. 2). We also found a significant effect of sex on p53 nuclear density (p = 0.0128; Table S55) and on p53 cytoplasmic density (p = 0.0105; Table S58).

Fig. 2.

A p53 nuclear intensity, B or p53 cytoplasmic intensity, C nuclear to cytoplasmic ratio, and D SIRT1/total protein concentration through unique or interaction effects. Full-transformed models were fitted for the p53 nuclear to cytoplasmic ratio. We applied a log transformation for all variables except SIRT1/total protein concentration (µg/mL) and p53 nuclear to cytoplasmic ratio (a.u.). We found a significant treatment effect for the p53 nuclear to cytoplasmic ratio (p = 0.0307) and a size class x treatment interaction (p = 0.0092), where after UV treatment, small breed puppies showed a decrease and large breed puppies showed an increase in the p53 nuclear to cytoplasmic ratio compared to control values

Post hoc analysis of the full transformed model

After UV treatment, primary fibroblasts from small breed puppies showed a significantly lower p53 nuclear to cytoplasmic ratio (p = 0.0375; Table S63), whereas primary fibroblasts from large breed old dogs showed a significantly higher p53 nuclear to cytoplasmic ratio (p = 0.0427; Table S63). Additionally, we found that treatment effects were larger for large puppies than small puppies (p = 0.0056; Table S64; Fig. S51), and the treatment effects were larger for large old dogs than small puppies (p = 0.0046; Table S64; Fig. S51).

Best subset models

When selecting independent variables using best subset regression according to BIC, we identified the best model following hierarchy without including redundant or unnecessary terms. According to these models, we can expect a 49.63% increase in non-glycolytic acidification with treatment (p = 0.0004; Table S26) and a 33.11% decrease in p53 nuclear density (p = 0.0031; Table S56) and a 36.29% decrease in p53 cytoplasmic concentration (p = 0.0021; Table S59) in male dogs compared with females, on average. Furthermore, there was a decrease in SIRT1 concentration in older dogs (p < 0.0001; Table S67) compared with puppies.

Discussion

The underlying genetics of several different dog breeds have led to high incidences and elevated risks of specific cancer types [44] as well as higher rates of malignant tumors in larger breeds of dogs [46]. Additionally, there is evidence of germ line p53 mutations in dogs [2, 50]. The oncogene p53 can switch on a cascade of transcription factors during cellular stress to prevent it from accumulating mutations that may lead to cancer, or it can become overexpressed, which also leads to cancerous phenotypes. Under certain conditions, p53, which is mainly found in the nucleus, can be transported out of the nucleus through nuclear pores into the cytoplasm [16] and can help regulate several steps in cellular metabolism [37].

The tumor suppressor p53 has been shown to influence metabolism in response to stress signals associated with DNA damage, hypoxia, and oxidative stress, to name a few [5]. For example, p53 can increase mitochondrial respiration by inducing the expression of cytochrome c oxidase complex regulators, a key step in aerobic oxidative phosphorylation [5], so cells that lack functional p53 have lower mitochondrial respiration than those with functional p53. Additionally, cells lacking functional p53 have an increased tendency to shift towards glycolysis, thereby contributing to the Warburg effect [5, 36]. The p53 pathway can also induce the expression of regulator proteins that inhibit glycolysis by lowering the level of rate-limiting intermediates, such as fructose-2,6-bisphosphate, and activating rate-limiting aerobic enzymes so as to block the Warburg effect [41]. Thus, the loss of p53 expression enhances aerobic glycolysis, and, in turn, upregulation of p53 inhibits glycolysis, as we see in our data in large older dogs.

Here, we link previously reported increases in glycolytic profiles in primary fibroblasts of large breed dogs compared to smaller breeds [23, 24] and report functional linkages between cellular metabolic function and p53 expression and SIRT1 concentration in dogs of different sizes and ages. Our data show that, on average, though not significantly different, primary fibroblasts from older large breed dogs had higher control glycolysis compared with after UV treatment (log₂-ratio = 0.271), in opposition to primary fibroblasts from older small breed dogs (log₂-ratio =  − 0.0884), small breed puppies (log₂-ratio =  − 0.503), and large breed puppies (log₂-ratio =  − 0.0271), which showed higher glycolytic rates after UV. It should be noted that after UV treatment, the p53 nuclear to cytoplasmic ratio increases in primary fibroblasts of large breed old dogs (log₂-ratio = 0.191) and large breed old puppies (log₂-ratio = 0.211). In contrast, we see a smaller increase in primary fibroblasts in small breed old dogs (log₂-ratio = 0.0686) and a decrease in small breed puppies (log₂-ratio =  − 0.387). Thus, primary fibroblasts from large breed old dogs upregulate p53 in times of cellular stress to yield a decrease in glycolysis compared with smaller breeds. Taken together, p53 could be inhibiting glycolysis, particularly in primary fibroblasts from large breed older dogs but not in primary fibroblasts from small breed older dogs. Following upregulation of p53 due to DNA damage, TIGAR is activated, causing the inhibition of glycolysis and the shuttling of metabolic products into the PPP to produce nucleotides for DNA repair [37]. The small increase in p53 nuclear to cytoplasmic ratio after UV in small old dogs was associated with a decrease in aerobic cellular respiration values with UV treatment—decreased basal OCR (log₂-ratio =  − 0.7424), maximal respiration (log₂-ratio =  − 0.6231), and ATP production (log₂-ratio =  − 0.8414)—suggesting either a different mechanism of action during times of cellular stress or a lack of functional p53, which has lower mitochondrial respiration [5]. Comparatively, primary fibroblasts from small and large breed puppies and large breed old dogs demonstrated increases in aerobic cellular respiration values with UV treatment.

By activating TIGAR, a p53-inducible gene, p53 can modulate glycolysis. Increases in TIGAR expression are associated with decreases in fructose-2,6-biphosphate, slowing glycolysis and shifting glucose processing through the pentose phosphate pathway (PPP) [30, 37]. Increased flux through the PPP increases the synthesis of nucleotides for cell cycle repair and replenishes NADPH [30, 37, 41]. NADPH, which can act as a powerful antioxidant, is needed in conjunction with reduced glutathione (GSH) for scavenging reactive oxygen species (ROS) [5, 30], which, when uncontrolled within cells, can cause lipid, protein, and DNA damage [23]. Using primary fibroblast cells and a metabolomics approach, at basal conditions, large breed young dogs had more GSH compared with older, larger dogs [6, 23]. Additionally, NADPH concentrations are significantly higher in older dogs of both sizes compared with younger dogs [6]. Potentially implying that the combination of increased NADPH and GSH found in large breed dogs could be manipulated through the p53 pathway, as we demonstrate here in primary fibroblasts from large breed old dogs with a higher p53 nuclear to cytoplasmic ratio after UV treatment or a more responsive p53 system in times of stress compared with smaller breeds.

Experimental evidence regarding the function of SIRT1 during with an exogenous stressor, such as UV, is sparse [8], especially the mechanism associated with the potential role of SIRT1 as a protector from senescence [8]. As cells age, SIRT levels decrease, and DNA damage increases as well, though without apoptosis which may lead to an increase in mutations [34]. SIRT1 functions to limit replicative senescence in mouse embryonic fibroblasts [7] as well as in human primary fibroblasts [1, 33]. Additionally, SIRT1 decreases with age in human primary fibroblasts [28], similar to our study. However, others have seen a downregulation of SIRT1 in human aging models with UV treatment [8, 28], which we did not observe in dog aging models. Increased concentration of SIRT1 in small breed puppies may inhibit p53 after UV treatment [39] and prevent p53 translocation to the nucleus [15]. However, increased concentrations of SIRT1 in large breed puppies do not seem to inhibit p53 functioning, as demonstrated by the higher p53 nuclear to cytoplasmic ratio after UV treatment in large breed dogs. Taken together, this implies that at a young age, small breed and large breed dogs may demonstrate different mechanisms for p53 control.

Both p53 and SIRT1 require NAD + as a cofactor. When p53 binds to NAD + , the resulting conformational change prevents its binding to DNA, thus decreasing p53-mediated translational activity (McClure et al. 2004; [48]). Large breed dogs seem to have higher concentrations of NAD⁺ compared with smaller breeds, and older dogs of both size classes also have higher concentrations of NAD⁺ compared with younger dogs [6]. Thus, NAD⁺ is not a rate-limiting step in SIRT1 activity, and p53 does not seem to be inhibited by SIRT1 in smaller dogs due to its decrease in concentration in those groups. In larger, older dogs, the higher concentration of NAD⁺ and the decrease in SIRT1 may allow for increased acetylation of p53, thus inhibiting its binding to DNA and decreasing its potential for repair. However, p53 inhibiting glycolysis remains in play in large breed dogs. Decreases in SIRT1 activity with respect to p53 functioning may lead to the degeneration of tissue and eventually chronic diseases [48]. Some studies have highlighted the direct effect of SIRT1 function on p53 [9, 31], though others have shown only modest effects of SIRT1 on p53 [18, 51], suggesting that there may be other mechanisms at play between SIRT1 and p53.

Here, we show that p53 may inhibit glycolysis in primary fibroblasts from large breed older dogs in times of cellular stress, more so than in primary fibroblasts from small breed dogs. We demonstrate that, due to a higher p53 nuclear to cytoplasmic ratio in primary fibroblasts from large breeds after UV treatment, p53 could have a more regulatory effect on large breed dogs’ metabolism compared with smaller breeds. However, SIRT1 concentrations decrease with age among domestic dogs’ primary fibroblasts of both size classes, suggesting a release of inhibition of p53 through the SIRT1 pathway with age. Still, increases in NAD⁺ in large breed dogs may limit p53 DNA binding ability, yielding a lack of DNA repair mechanism in large breed dogs, as seen before [23], which may lead to increased incidences of cancer, especially due to the more pronounced upregulation of p53 with cellular stress.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

AGJ designed the experiment, collected and grew primary fibroblasts, performed seahorse, immunohistochemistry and ELISA experiments, and wrote a first draft. KDP and MB did image analysis on immunohistochemistry. SL and WC cleaned the data, performed statistical analyses, developed tables and figures, helped with the interpretation of the data, and helped with the writing of the materials and methods and results sections.

Funding

This research was funded by a Colgate University Research Council Picker fellowship to AGJ. The Seahorse XFe96 oxygen flux analyzer was purchased via a National Science Foundation Major Research Instrument grant (NSF MRI 1725841 to AGJ).

Data availability

All raw data can be found at https://figshare.com/articles/dataset/Cellular_metabolic_pathways_of_aging_in_dogs_could_p53_and_SIRT1_be_at_play_/22329277.

Materials availability

For any inquiries, please email ajimenez@colgate.edu. For statistical analyses and code, please email wcipolli@colgate.edu.

Declarations

Competing interests

The authors declare no competing interests.