Pathological Relationship between Intracellular Superoxide Metabolism and p53 Signaling in Mice

Intracellular superoxide dismutases (SODs) maintain tissue homeostasis via superoxide metabolism. We previously reported that intracellular reactive oxygen species (ROS), including superoxide accumulation caused by cytoplasmic SOD (SOD1) or mitochondrial SOD (SOD2) insufficiency, induced p53 activation in cells. SOD1 loss also induced several age-related pathological changes associated with increased oxidative molecules in mice. To evaluate the contribution of p53 activation for SOD1 knockout (KO) (Sod1−/−) mice, we generated SOD1 and p53 KO (double-knockout (DKO)) mice. DKO fibroblasts showed increased cell viability with decreased apoptosis compared with Sod1−/− fibroblasts. In vivo experiments revealed that p53 insufficiency was not a great contributor to aging-like tissue changes but accelerated tumorigenesis in Sod1−/− mice. Furthermore, p53 loss failed to improve dilated cardiomyopathy or the survival in heart-specific SOD2 conditional KO mice. These data indicated that p53 regulated ROS-mediated apoptotic cell death and tumorigenesis but not ROS-mediated tissue degeneration in SOD-deficient models.


Introduction
Age-related pathological changes are caused by several genetic and environmental factors. To analyze the age-related changes in vivo and in vitro, researchers have used several genetic and pharmacological manipulations for the induction of redox imbalance [1,2]. Superoxide dismutase (SOD) enzymes play a major role in the intracellular antioxidant system by catalyzing the conversion of superoxide radicals (O 2 •− ) to hydrogen peroxide and O 2 [3]. In mammals, copper/zinc-SOD (SOD1) exists in the cytoplasm, while manganese-SOD (SOD2) is distributed in the mitochondrial matrix to regulate intracellular redox balance in cells. Since SOD expression and activity are significantly decreased in aged osteoporotic, end-stage osteoarthritic, and Alzheimer's disease individuals [4][5][6], redox imbalance caused by SOD decline is considered an important mechanism underlying the induction of age-related pathological changes.
In contrast to the above findings, SOD2-deficient mice showed dilated cardiomyopathy (DCM), steatosis, and metabolic acidosis, which resulted in neonatal lethality [21]. To analyze the SOD2-deficient phenotypes in adults, we generated and established tissuespecific SOD2 knockout (KO) mice [2,22]. Several tissue-specific SOD2-deficient mice showed DCM-type heart failure [23], disturbance of exercise activity [24], spongiform encephalopathy [25], bone loss [26], and cartilage degeneration [27]. Consequently, we proposed that tissue-specific SOD2 KO mice were a useful model of age-related pathological changes caused by mitochondrial dysfunction. In addition, pressure overload, hypoxic stress, and genotoxic stress-induced p53 upregulation resulted in cardiomyocyte death [28]. Heart failure such as DCM and right ventricular hypertrophy model mice also showed the accumulation of p53 and cardiomyocyte apoptosis [29][30][31][32]. Since suppression of p53 via pharmacological as well as genetic approaches ameliorated heart failure [29,33,34], p53 is considered a key molecule involved in heart failure. However, while SOD2 loss induced DCM in mice, the involvement of p53 in DCM-type heart failure caused by SOD2 loss remains unclear.
In the present study, in order to clarify the contribution of p53 to the Sod1or Sod2deficient phenotypes in vivo, we generated two types of double-knockout (DKO) mice: Sod1 and p53 DKO mice, as well as heart-specific Sod2 and p53-deficient mice. We also discussed the influence of p53 deficiency on the phenotypes of SOD1 or heart-specific SOD2 KO mice.

p53 Insufficiency Effectively Suppressed Apoptotic Cell Death In Vitro
Previously, we reported that the intrinsic O 2 •− accumulation by SOD1 loss promoted p53 activation and apoptotic cell death in vitro [20]. In addition, antioxidant reagents effectively attenuated Sod1-deficient phenotypes accompanied by p53 upregulation in fibroblasts and skin tissues [20]. To clarify the pathological relationship of p53 upregulation in Sod1-deficient phenotypes in mice, we generated Sod1 and p53 DKO mice. First, we performed in vitro fertilization of Sod1 KO oocytes with p53 KO frozen sperm to obtain double-heterozygous mice. Next, we intercrossed double-heterozygous males and females to generate DKO mice. Unexpectedly, we obtained only a very small number of DKO mice from cross-fertilization via natural mating as well as in vitro fertilization (Table 1). These data indicated that the birth rate of DKO mice was not Mendelian. Next, we assessed the efficiency of p53 loss in Sod1 −/− fibroblasts. Although Sod1 −/− fibroblasts died within 3 days under 20% O 2 conditions, p53 loss improved the cell number decline among Sod1 −/− cells ( Figure 1A). However, DKO cells showed low cell proliferation as well as significant incrementation of dihydroethidium (DHE)-and CM-H 2 DCFDA (DCF)-positive ROS, including O 2 •− accumulation, with values similar to those seen in Sod1 −/− cells ( Figure 1B-D). Interestingly, DKO cells showed significantly fewer apoptotic cells than Sod1 −/− cells ( Figure 1E). In contrast, p53 −/− cells exhibited no harmful phenotypic effect, including with regard to the cell proliferation and ROS accumulation ( Figure 1B-D). These data indicated that p53 impairment suppressed apoptotic cell death, which resulted in an increase in the cell survival among DKO cells.
( Figure 1B-D). These data indicated that p53 impairment suppressed apoptotic cell death, which resulted in an increase in the cell survival among DKO cells.

p53 Loss Failed to Attenuate the SOD1-Deficient Phenotypes in Mice
To evaluate the effect of p53 deficiency in Sod1 −/− mice, we expanded the intercrossing and analyzed organ phenotypes of DKO mice. Sod1 −/− mice revealed body weight reduction, muscle atrophy, and liver weight gain [1], but these were not significant differences compared with wild-type (WT) mice in this analysis (Figure 2A-C). SOD1 loss significantly induced skin thinning and decrease of red blood cell number but not splenomegaly ( Figure 2D-F). On the other hand, p53 −/− mice showed no significant differences in all parameters ( Figure 2A-F). DKO mice showed significant reductions of muscle weight, skin thickness, and red blood cell number ( Figure 2B,D,F). Interestingly, DKO mice also exhibited exacerbation of splenomegaly compared with Sod1 −/− mice ( Figure 2E). Importantly, p53 haploinsufficiency also failed to improve Sod1 −/− phenotypes (Figure 2A-F). Furthermore, Sod1 +/− , p53 −/− mice were extremely similar to WT and p53 −/− mice (Figure 2A-F). These data indicate that p53 insufficiency did not seriously influence the organ phenotypes of Sod1 −/− mice.

p53 Loss Failed to Attenuate the SOD1-Deficient Phenotypes in Mice
To evaluate the effect of p53 deficiency in Sod1 −/− mice, we expanded the intercrossing and analyzed organ phenotypes of DKO mice. Sod1 −/− mice revealed body weight reduction, muscle atrophy, and liver weight gain [1], but these were not significant differences compared with wild-type (WT) mice in this analysis ( (Figure 2A-F). These data indicate that p53 insufficiency did not seriously influence the organ phenotypes of Sod1 −/− mice.

SOD1 and p53 DKO Mice Showed Early Tumor Progression
About half of p53 KO mice reportedly show tumor progression by six months of age [35]. In contrast, Sod1 KO mice have been reported to reveal no tumor phenotypes until six months of age [14]. We therefore monitored the tumor progression phenotypes in DKO mice until four months of age. A large number of DKO mice showed remarkable spontaneous tumor progression in the appearance of discriminative by four months of age (Table 2 and Figure 3A). Whereas p53 KO mice mostly showed thymic lymphoma or sarcomas [35], DKO mice developed multifarious tumor throughout the whole body, including in the cervix, abdomen, limbs, and testis ( Figure 3A,B). Importantly, Sod1 −/− , p53 +/− as well as Sod1 +/− , p53 −/− mice displayed no tumor progression by four months of age ( Table 2), suggesting that heterozygotic loss of p53 or Sod1 was sufficient to achieve the suppression of tumor development in DKO mice. These data indicated that systemic

SOD1 and p53 DKO Mice Showed Early Tumor Progression
About half of p53 KO mice reportedly show tumor progression by six months of age [35]. In contrast, Sod1 KO mice have been reported to reveal no tumor phenotypes until six months of age [14]. We therefore monitored the tumor progression phenotypes in DKO mice until four months of age. A large number of DKO mice showed remarkable spontaneous tumor progression in the appearance of discriminative by four months of age (Table 2 and Figure 3A). Whereas p53 KO mice mostly showed thymic lymphoma or sarcomas [35], DKO mice developed multifarious tumor throughout the whole body, including in the cervix, abdomen, limbs, and testis ( Figure 3A,B). Importantly, Sod1 −/− , p53 +/− as well as Sod1 +/− , p53 −/− mice displayed no tumor progression by four months of age (Table 2), suggesting that heterozygotic loss of p53 or Sod1 was sufficient to achieve the suppression of tumor development in DKO mice. These data indicated that systemic oxidative damage caused by complete SOD1 loss accelerated the tumor initiation and/or development in the whole body of p53 −/− mice. Table 2. The number of tumor or death to each mouse until 4 months of age.

Genotype
Number   We previously found that heart-specific Sod2-deficient (Sod2 H/H ) mice showed a short lifespan associated with DCM [23]. Accumulating evidence has suggested that heart fail-

p53 Insufficiency Had No Effect on the Heart Failure of Heart-Specific Sod2-Deficient Mice
We previously found that heart-specific Sod2-deficient (Sod2 H/H ) mice showed a short lifespan associated with DCM [23]. Accumulating evidence has suggested that heart failure involves the p53 signaling pathway [28]. In vitro studies revealed that Sod2 loss increased mitochondrial ROS and p53 activation in mouse embryonic fibroblasts (Watanabe et al., personal communication). In this context, to clarify the contribution of p53 to heart failure in Sod2 H/H mice, we generated heart-specific Sod2and p53-deficient mice (Sod2 H/H , p53 H/H ). Sod2 H/H , p53 H/H mice had a similarly short lifespan to Sod2 H/H mice ( Figure 4A). Furthermore, DCM caused by heart-specific Sod2 loss was also recognized in Sod2 H/H , p53 H/H mice ( Figure 4B,C). p53 H/H mice showed a normal lifespan and heart tissue structures with strong similarity from those of WT mice including Sod2 f/f and p53 f/f mice ( Figure 4A-C). Importantly, the pathogenesis of cardiac fibrosis was also not markedly different between Sod2 H/H and Sod2 H/H , p53 H/H mice ( Figure 4C). These data indicated that the induction and the progression of DCM phenotypes by Sod2 in mice were not influenced by the loss of the p53 molecule.

Discussion
It is very well known that p53 is involved in several signaling pathways, including the DNA damage response (DDR) leading to cellular senescence induction, cell cycle arrest, DNA repair, autophagy, and cell death [36,37]. Previously, we reported that SOD1 loss induced the marked intracellular accumulation of ROS (about 40-fold) accompanied by p53 activation in vitro [1,20], suggesting a close relationship between the induction of SOD1-decifient phenotypes and p53 activation in vitro. p53 regulates the cell fate, such as the transcriptional induction of antioxidant-, cell cycle arrest-, and apoptosis-related genes, according to the intracellular redox state [34,38]. Low levels of ROS accumulation induced p53-mediated cytoprotective property and suppressed apoptosis [39]. In addition, moderate ROS activated cell cycle checkpoint genes, which resulted in cell cycle ar-

Discussion
It is very well known that p53 is involved in several signaling pathways, including the DNA damage response (DDR) leading to cellular senescence induction, cell cycle arrest, DNA repair, autophagy, and cell death [36,37]. Previously, we reported that SOD1 loss induced the marked intracellular accumulation of ROS (about 40-fold) accompanied by p53 activation in vitro [1,20], suggesting a close relationship between the induction of SOD1-decifient phenotypes and p53 activation in vitro. p53 regulates the cell fate, such as the transcriptional induction of antioxidant-, cell cycle arrest-, and apoptosis-related genes, according to the intracellular redox state [34,38]. Low levels of ROS accumulation induced p53-mediated cytoprotective property and suppressed apoptosis [39]. In addition, moderate ROS activated cell cycle checkpoint genes, which resulted in cell cycle arrest for DNA repair [40]. In contrast, excessive ROS stress was shown to lead to apoptosis [39]. Whereas p53 loss remarkably increased the survival of Sod1 −/− fibroblasts to a point similar to that of WT cells, p53 insufficiency did not influence the intracellular ROS accumulation or cell proliferation caused by SOD1 loss (Figure 1). p53 deficiency is well known to be incapable of promoting apoptotic cell death caused by ROS in fibroblasts [41,42]. Our data also indicated that p53 deficiency effectively suppressed apoptosis induction via DDR in Sod1 −/− cells. Therefore, DKO cells can survive despite oxidative damage under normal atmospheric conditions (Figure 1). These results indicated that p53 mainly regulated apoptotic cell death rather than cell cycle arrest when excessive intracellular ROS accumulated in Sod1 −/− cells.
Since SOD1 enzyme includes copper and zinc ions, SOD1 also acts as a chelator of copper and zinc ions. Sod1 deficiency might induce an increase in free copper and zinc ions in cytoplasm. Overdose of copper induced apoptotic cell death in granule cells, resulting in degeneration and neuronal loss in the central nervous system [48]. Furthermore, excessive zinc induced the disturbance of redox balance, gene expression, bone metabolism, and alternation of the p53 protein structure [49][50][51]. Several age-related chronic diseases also showed increased serum levels of copper and zinc ions [52,53]. In this context, ion homeostasis failure caused by SOD1 loss might induce age-related pathological changes in vivo and in vitro. Recently, several studies reported that SOD1 protein induces post-translational modifications and regulates the expression of antioxidant genes as a transcriptional factor [54]. In addition to the loss of antioxidant activity, the loss of the metal chelating ability and transcriptional function might markedly affect Sod1-deficient phenotypes. Further studies are needed to clarify the molecular mechanisms involved in Sod1-defcient phenotypes.
In contrast to the above findings, our results showed that Sod1 deficiency exacerbated tumor progression in p53 −/− mice ( Table 2). A high percentage (68%) of Sod1 −/− mice revealed nodular hyperplasia or hepatocellular carcinoma by 20 months of age [14], whereas DKO mice showed the early detection of tumor formation (by 4 months) with a high probability (50%) ( Table 2 and Figure 3). Because p53 protects tumorigenesis from ROSmediated DNA damage [55], the significant increase in oxidative damage induced by SOD1 insufficiency may accelerate tumorigenesis in p53 −/− mice. Furthermore, E2-promoter binding factor (E2F) transcriptional factor interacted with retinoblastoma susceptibility genes to regulate cellular proliferation and tumorigenesis [56,57]. Double mutant for the E2F family of transcription factors, including E2F1 and E2F2, resulted in γH2AX accumulation accompanied by p53 activation, which consequently caused apoptotic cell death in the pancreas [58]. Disruption of p53 in E2F1 and E2F2 double-knockout mice caused the suppression of apoptosis induction, which resulted in the progression of thymic lymphomas and a shortened lifespan [58]. This suggests that p53-dependent apoptosis induced by SOD1 or E2F1/E2F2 deficiencies is a key mechanism underlying tumor suppression. Accordingly, p66Shc generates hydrogen peroxide, and p66Shc loss decreases ROS production. In this context, p66Shc insufficiency significantly increased the lifespan and suppressed tumor progression in p53 KO mice [59].
Many reports have shown that heart failure as age-or pathology-related phenotypes is mediated by p53 upregulation [28]. In addition, p53 suppression by pharmacological and genetic techniques ameliorates the phenotypes in heart failure models [28]. In this context, we generated Sod2 H/H , p53 H/H mice to attenuate the DCM phenotypes of Sod2 H/H mice. Unexpectedly, p53 loss failed to improve the short lifespan and DCM phenotypes in Sod2 H/H mice (Figure 4). In general, hypoxic and genotoxic stress induces cardiomyocyte apoptosis through p53 activation, resulting in heart failure [29,30]. However, since Sod2 H/H mice did not show the induction of apoptotic cell death in the heart [23], p53 insufficiency may not mitigate heart failure in Sod2 H/H , p53 H/H mice. We previously reported that antioxidant reagents, such as EUK-8, MnTBAP, manganese porphyrins, and apple procyanidins, improved DCM accompanied by a reduction in mitochondrial ROS accumulation in Sod2 H/H mice [23,[60][61][62]. Recently, Guo et al. reported that a loss-of-function mutation in extracellular SOD (SOD3) induced chronic kidney disease accompanied by systolic hypertension and cardiac hypertrophy in a Dahl/salt-sensitive strain of rats [63]. In addition, SOD3 KO mice also showed hypoxia-induced pulmonary vascular disease [64,65]. These reports suggest that not only intracellular but also extracellular ROS affect cardiac hypertrophy and cardiovascular diseases.
Taken together, our data indicate that p53 plays a minimal role in the pathogenesis of SOD1 or heart-specific SOD2 deficiency in mice. Since p53 mainly functions in apoptosis induction and tumor suppression, it has little involvement in apoptosis-independent tissue disorders, including adult Sod1 −/− mice and Sod2 H/H heart. In contrast, the suppression of apoptosis by p53 loss accelerated tumor initiation/progression in Sod1 −/− mice. Likewise, SOD1 deficiency accelerated tumor progression in p53 −/− mice, indicating that apoptosis induction by p53 as well as intracellular O 2 •− metabolism by SOD1 strongly contributed to tumor suppression. In conclusion, SOD1-deficient mice and tissue-specific SOD2-deficiet mice were useful model mice for an aging study without tumor progression.

Histology
For histological morphology, skin specimens from the back tissue, heart, and tumor were dissected and fixed in a 20% formalin neutral buffer solution (FUJIFILM Wako, Osaka, Japan) overnight while embedded in paraffin and then sectioned on a microtome at 4 µm according to standard techniques. Hematoxylin and eosin staining for the skin morphology and heart as well as Azan staining for total collagen deposition were performed as described previously [15,23,67]. The thickness of the skin tissue was determined using the BZ-X Analyzer software program (Keyence, Osaka, Japan).

Cell Culture
The skin tissue specimens were dissected from 5-day-old neonates. The primary dermal fibroblasts were isolated by dissociation in 0.2% collagenase type 2 (Worthington Biochemical Corporation, Lakewood, NJ, USA) at 37 • C for 60 min. The cells were cultured in minimum essential medium Eagle, alpha modification (α-MEM; Life Technologies Corporation, Carlsbad, CA, USA) supplemented with 20% fetal bovine serum (FBS, Thermo Fisher Scientific, Waltham, MA, USA), 100 unit/mL penicillin, and 0.1 mg/mL streptomycin at 37 • C in a humidified incubator (ASTEC, Fukuoka, Japan) with 5% CO 2 and 1% O 2 to expand and maintain Sod1 −/− fibroblasts. During experiments, the cells were cultured under 20% O 2 conditions. Cell viability was measured by the cell proliferation enzymelinked immunosorbent assay bromodeoxyuridine (BrdU; Roche Diagnostics K.K., Basel, Switzerland) according to the manufacturer's instructions. The relative BrdU incorporate values were calculated by a triplicate analysis.

Flow Cytometry
The accumulation of intracellular ROS was detected using DHE and DCF (Life Technologies Corporation). The cells were incubated with 10 µM DHE or 10 µM DCF for 30 min at 37 • C. Following incubation, the cells were trypsinized and resuspended in phosphate-buffered saline. Apoptosis was measured using a fluorescein isothiocyanate (FITC) Annexin V Apoptosis Detection Kit I (BD Biosciences, San Diego, CA, USA) according to the manufacturer's instructions. The fluorescence intensities were assessed using a flow cytometer (BD FACSCanto II; BD Biosciences, San Diego, CA, USA).

Statistical Analyses
Statistical evaluations were performed using a two-way analysis of variance with the GraphPad Prism9 software program (GraphPad Software, San Diego, CA, USA). Differences between the data were considered to be significant when the p-values were less than 0.05. The data are represented as the means ± the standard deviation (SD).
Author Contributions: K.W. and T.S. designed the research. K.W. and T.S. wrote the manuscript. K.W., S.S., Y.O. and T.T. performed research. K.W. analyzed the data. K.W., S.S., Y.O., T.T. and T.S. discussed the hypothesis and interpreted the data. K.W. and T.S. edited the article. T.S. coordinated and directed the project. All authors have read and agreed to the published version of the manuscript.