|Home | About | Journals | Submit | Contact Us | Français|
TAp63 prevents premature aging suggesting a link to genes that regulate longevity. Further characterization of TAp63−/− mice revealed that these mice develop obesity, insulin resistance, and glucose intolerance, similar to those seen in mice lacking two key metabolic regulators, Silent information regulator T1 (Sirt1) and AMPK. While the roles of Sirt1 and AMPK in metabolism have been well studied, their upstream regulators are not well understood. We found that TAp63 is important in regulating energy metabolism by accumulating in response to metabolic stress and transcriptionally activating Sirt1, AMPKα2, and LKB1 resulting in increased fatty acid synthesis and decreased fatty acid oxidation. Moreover, we found that TAp63 lowers blood glucose levels in response to metformin. Restoration of Sirt1, AMPKα2, and LKB1 in TAp63−/− mice rescued some of the metabolic defects of the TAp63−/− mice. Our study defines a role for TAp63 in metabolism and weight control.
Obesity is a global health problem, and obesity related diseases such as diabetes and heart disease are a critical threat to human longevity (Banks et al., 2008; Flegal et al., 2005). The sirtuin family of genes is an important regulator of longevity in multiple organisms from yeast to mice (Haigis and Guarente, 2006; Kume et al., 2010; Lin et al., 2000; Rogina and Helfand, 2004). Sirt1 is a member of seven sirtuin members in mammals, and has received a flurry of attention based on its ability to regulate synthesis, storage, and utilization of lipids (Rodgers and Puigserver, 2007; Takemori et al., 2011). Transgenic and knock out mice of Sirt1 have phenotypes that have revealed critical clues about its function in aging, metabolism, and cancer. For example, Sirt1 transgenic mice are protected from glucose intolerance, have a more active metabolism, and leaner bodies (Banks et al., 2008). Tissue specific deletion in liver and brain of Sirt1 in mice resulted in increased body weight, fatty liver, and glucose intolerance associated with aging (Cohen et al., 2009; Purushotham et al., 2009). Sirt1 also plays an important role in cancer and has been shown to regulate genes involved in senescence, DNA repair and tumor suppression (Wang et al., 2008).
Recent reports have indicated an interdependence of AMP-activated protein kinase (AMPK) and Sirt1 (Canto et al., 2010) in response to fasting and exercise. AMPK regulates energy homeostasis through its ability to balance catabolic and anabolic activity to regulate lipid and glucose metabolism (Hardie and Frenguelli, 2007; Towler and Hardie, 2007). This function of AMPK is critical for its response to metabolic stress. Importantly, AMPK activation is required for the therapeutic benefits of the drug metformin in the management of type 2 diabetes (Shaw et al., 2004; Shaw et al., 2005). The activation of AMPK occurs through phosphorylation by the tumor suppressor and upstream kinase, LKB1(Woods et al., 2003).
Sirt1, AMPK, and LKB1 are key metabolic regulators. Sirt1 mediates metabolic and the physical beneficial effects of caloric restriction (CR), such as the extension of lifespan, while AMPK stimulates fatty acid oxidation (Boily et al., 2008; Canto et al., 2010), and LKB1 is a regulator of AMPK. The mechanistic regulation of these functions of Sirt1 and AMPK is primarily through their ability to post translationally modify transcription factors, such as FOXO1, p53, and acetyl-CoA carboxylase (ACC1), thereby affecting their downstream functions (Banks et al., 2008; Canto et al., 2009; Rodgers et al., 2005; Vaziri et al., 2001). Sirt1 directly deacetylates p53 and AMPK phosphorylates ACC1. The phosphorylation of ACC1 is critical for the production of malonyl-CoA as a substrate for fatty acid biosynthesis. While it is clear that Sirt1, AMPK, and LKB1 play crucial roles in energy metabolism in response to different energy stresses, how they are regulated in response to metabolic stress remains unclear.
Recently, we have demonstrated that the p53 family member and p63 isoform, TAp63, plays critical roles in aging and in the suppression of tumorigenesis and metastasis (Flores et al., 2005; Su et al., 2010; Su et al., 2009). The vast majority of research on p63 has been performed using mouse models deficient for all isoforms of p63 (Mills et al., 1999; Yang et al., 1999). Using these mouse models, p63 has been found to be critical for epidermal morphogenesis (Mills et al., 1999; Yang et al., 1999). The generation of isoform specific knock out mice for p53 family members has unveiled unique functions for these genes (Guo et al., 2009; Su et al., 2010; Su et al., 2009; Suh et al., 2006). We have generated TAp63−/− mice and have found that TAp63−/− mice age prematurely, and TAp63 is critical for the maintenance of adult stem cells in quiescence (Su et al., 2009). Loss of TAp63 triggers a senescence program in TAp63 deficient tissues, resulting in premature aging and reduced lifespan in mice (Guo et al., 2009; Su et al., 2009). We also found that TAp63+/− and TAp63−/− are highly tumor prone and develop metastatic disease (Su et al., 2010). Here we show that in addition to these phenotypes, aging TAp63−/− mice develop multiple phenotypes consistent with type 2 diabetes, including glucose intolerance, insulin resistance, and liver steatosis (fatty liver). These mice phenocopy those with tissue specific deletions of Sirt1 as well as those deficient for AMPKα (Cohen et al., 2009; Purushotham et al., 2009; Viollet et al., 2003). Moreover, we found that TAp63 upregulates Sirt1 after calorie restriction. We found that the mechanism for these phenotypes is the requirement for TAp63-mediated transcriptional activation of Sirt1, AMPKα2, and LKB1. Consistent with regulation of AMPKα2 and LKB1 by TAp63, we found that TAp63 is critical for lowering blood glucose levels in response to metformin, a drug used to treat type 2 diabetes. These findings reveal roles for TAp63 in metabolism and indicate that TAp63 is a potential therapeutic target for metabolic disorders.
We observed that as TAp63−/− mice age they become obese by 8 months of age (Fig. 1A). By 12 months of age, we observed that some mice weighed as much as 91 grams and were immobile and unable to access food and water (Fig. 1B). We found that TAp63−/− mice had increased body fat that was present underneath the skin and intercalated into multiple organs (Fig. 1C). To understand the role of TAp63 in obesity, we aged a cohort of 15 mice for 18 months and found that by eight months of age, the TAp63−/− mice weighed 30% more than their wild-type littermates (Fig. 1D). This difference increased to 40% by 12 months of age (Fig. 1D).
To systematically assess weight gain in the TAp63−/− mice, we maintained 15 TAp63−/− and 15 wild-type mice on a high fat diet for 16 weeks beginning at 4 weeks of age. A similar number of wild-type and TAp63−/− mice were fed a control (con) or CR diet. We found that as early as 2 weeks after being shifted to a high fat diet, the TAp63−/− mice exhibited significant weight gain as compared to wild-type mice (Fig. 1E). TAp63−/− mice fed control or CR diets had body weights that were comparable to their wild-type counterparts throughout the experiment (Fig. 1E) indicating that TAp63 is critical for regulating weight when mice are challenged with high fat food (HF). This was not due to increased food intake, since we found no significant difference in the amount of food consumed by TAp63−/− mice compared to wild-type mice on all three diets (Fig. 1F, 1G, and data not shown).
To determine whether the weight gain in the TAp63−/− mice could be attributed to differences in activity, we compared the horizontal activity of five wild-type and five TAp63−/− mice over a twenty-four hour period and found the activity of the TAp63−/− mice is significantly lower than their wild-type littermates (Fig. 1H). We also measured levels of oxygen consumed and carbon dioxide produced and found that TAp63−/− mice consume less oxygen (Fig. 1I) and produce less carbon dioxide (Fig. 1J) than their wild-type littermates. Moreover, we found that the respiratory quotient (RQ) is higher in TAp63−/− mice during the light cycle relative to wild-type mice, indicating a relative resistance to fatty acid metabolism (Fig. 1K). Lastly, we found that heat production in TAp63−/− mice is significantly lower than in wild-type mice (Fig. 1L). Taken together, these data indicate that TAp63−/− mice have impaired energy metabolism.
To determine whether TAp63−/− mice had defects in energy metabolism, we measured glucose tolerance and serum levels of triglycerides, cholesterol, and insulin. To determine whether TAp63−/− mice have decreased glucose tolerance, we challenged 10 TAp63−/− and 10 wild-type mice at twelve months of age with 2 grams of glucose per kilogram of body weight. We found that TAp63−/− mice had a significantly decreased glucose tolerance as compared to wild-type mice (Fig. 2A and B). To determine whether this was an age-associated phenomenon, we challenged 15 TAp63−/− mice with a high fat diet for 16 weeks beginning at 4 weeks of age and measured the glucose tolerance of these mice after 16 weeks on the high fat diet (Fig. 2C and D). We found that young TAp63−/− mice on a high fat diet also developed glucose intolerance (Fig. 2C and D). Diabetes and insulin resistance is associated with defects in glucose uptake. To understand whether TAp63 deficient mice are defective in glucose uptake, we measured the ability of TAp63−/− mouse embryo fibroblasts (MEFs) to uptake glucose. Indeed, TAp63−/− MEFs had a dramatically reduced ability to uptake glucose (Fig. 2E). Glucose intolerance and defects in glucose uptake are associated with insulin resistance and type 2 diabetes. To determine whether TAp63−/− mice develop insulin resistance, we measured levels of glucose in wild-type and TAp63−/− mice subsequent to insulin injection (Fig. 2F) and found that TAp63−/− mice has significantly higher levels of glucose (Fig. 2G) indicating resistance to insulin. We also found that TAp63−/− mice fed a control or a high fat diet have significantly higher serum insulin compared with their wild-type littermates at 5 months of age (Fig. 2H). We have shown previously that TAp63−/− mice exhibit signs of premature aging in some tissues (Su et al., 2010; Su et al., 2009). To determine whether the metabolic defects observed in the TAp63−/− mice is due to premature senescence, we assayed for markers of senescence in MEFs and livers of TAp63−/− mice. We found that TAp63−/− MEFs at passage 2 had increased levels of senescence associated β-galactosidase (Suppl. Fig. 1A). Therefore, we used passage 1 MEFs in all of our studies. Moreover, we found that livers from TAp63−/− mice did not show evidence of senescence and expressed wild-type levels of p16Ink4a, p19Arf, and PML indicating that premature senescence was not the cause of the metabolic defects of the TAp63−/− mice (Suppl. Fig. 1B). Taken together, these data are consistent with type 2 diabetes and other mouse models with this disease (Memon et al., 1994).
Lipid accumulation and increased plasma levels of triglycerides are associated with insulin resistance and type 2 diabetes (Ginsberg et al., 2005). TAp63−/− mice fed a control diet are glucose intolerant and insulin resistant; therefore, we examined the levels of lipids in these mice. Blood was drawn from 15 fasting TAp63−/− and wild-type mice at 5 months of age that had been maintained on a control diet. We found that the levels of cholesterol, triglycerides, adiponectin, and leptin were significantly higher in TAp63−/− than in wild-type mice (Fig. 2I–K). These data indicate that TAp63 is important for lipid metabolism and that loss of TAp63 leads to obesity and diabetes in mice.
High levels of lipids and tryglycerides in the blood are associated with fat accumulation in the liver and liver steatosis (fatty liver disease). To determine whether TAp63−/− mice develop liver steatosis, we examined hematoxylin and eosin (H&E) stained livers from TAp63−/− and wild-type mice (Fig. 3A–H). We found that livers from 12 month old TAp63−/− mice fed control chow had many areas of fat deposition as compared to livers from wild-type mice of the same age (Figs. 3A and E). Even livers from young TAp63−/− mice at 5 months of age that were fed control, CR, and high fat diets for 16 weeks beginning at 4 weeks of age exhibited a higher accumulation of fat than their wild-type counterparts (compare Figs. 3B & F, Figs. 3C & G, and Figs. 3D & H). This was further demonstrated by Oil Red O staining showing accumulation of lipids in the livers of TAp63−/− mice (compare Figs. 3I & M, Figs. 3J & N, Figs. 3K & O, and Figs. 3L & P). These data indicate that lipids accumulate in the tissues of TAp63−/− mice.
Increased lipid deposition in the blood, liver, and other tissues can be caused by increased fatty acid synthesis, decreased utilization of lipids, or both. To determine whether levels of important regulators of lipid homeostasis were affected in the absence of TAp63, we measured mRNA levels derived from livers of 5 month old TAp63 deficient mice fed a control diet for fatty acid synthase (FAS), an enzyme that catalyzes fatty acid synthesis, leading to increased incorporation of free fatty acids into triglycerides (Fig. 3Q) and carnitine palmitoyltransferase-I (CPT-I) (Fig. 3R), a rate-limiting enzyme for fatty acid oxidation. We found that the mRNA levels were altered in livers from TAp63−/− mice compared to wild-type mice, suggesting that the increase in the levels of lipids in TAp63 deficient mice is due to deregulation of these metabolic enzymes. Moreover, we examined a panel of genes involved in lipid storage and metabolism and found that expression of these mRNAs are deregulated in the absence of TAp63 (Fig. 3S).
To determine the contribution of CPT1 deregulation to lipid metabolism and mitochondrial dysfunction in TAp63 deficient mice, we measured the oxygen consumption rate in control and TAp63−/− MEFs under basal conditions and in response to palmitate, the substrate of CPT-I. TAp63−/− MEFs exhibited low basal mitochondrial oxygen consumption (Fig. 3T) and an impaired ability to metabolize palmitate (Fig 3U). These data indicate that TAp63−/− mice have defects in both fatty acid accumulation and fatty acid oxidation. On the basis of these findings, we conclude that fatty acid synthesis is increased and fatty oxidation is decreased in TAp63−/− mice.
The altered expression of enzymes regulated by silent information regulator T1 (Sirt1) and/or AMPKα in TAp63 deficient mice (Fig. 3Q–S) led us to ask whether TAp63 regulates lipid metabolism through Sirt1, AMPK, or LKB1. We measured the protein and mRNA levels of these energy sensors and metabolism regulators in tissues from TAp63 deficient mice at 4 weeks and 5 months of age. We found that MEFs (Fig. 4A) and muscle from 4 week-old TAp63−/− mice have lower levels of Sirt1 than their wild-type counterparts while liver fat and skin showed no significant difference (Fig. 4B). Importantly, we found that levels of Sirt1 are unchanged in the skin of TAp63−/− mice (Fig. 4B) suggesting that TAp63 regulates Sirt1 expression in metabolic tissues. In contrast, by 5 months of age, we could see that Sirt1 is lower in muscle, fat, and liver This difference could also be seen in tissues from TAp63−/− mice fed a control (con) or calorie restricted (CR) diet (Fig. 4C–E). When TAp63−/− mice were fed a high fat diet, Sirt1 levels were only found to be lower in fat (Fig. 4C–E).
Levels of Sirt1 are upregulated in response to CR through a mechanism that is not clearly understood and levels of Sirt1 decline with age (Brooks and Gu, 2009). To determine whether TAp63 is required for Sirt1 upregulation in response to CR, we measured the levels of Sirt1 in the livers of 5 month old wild-type and TAp63−/− mice that had been on a CR diet for 16 weeks beginning at 4 weeks of age. At this age, levels of Sirt1 were lower in muscle and fat of TAp63−/− mice compared to wild-type mice (Fig. 4C–E). These data indicate a role for TAp63 in Sirt1 upregulation in response to CR in certain tissues, like muscle and fat. Sirt1 is known to de-acetylate p53. To ask whether p53 is de-acetylated by Sirt1 in the absence of TAp63 where levels of Sirt1 are low, we performed western blot analysis for acetylated p53 in the muscle, fat, and livers of TAp63−/− mice fed a con, CR, or HF diet. We found that in all cases there is an increase in p53 acetylation (Fig. 4C) indicating that Sirt1 in not functional in the absence of TAp63.
To determine if the difference occurred at the transcriptional level, we assayed the levels of Sirt1 mRNA in wild-type and TAp63−/− MEFs starved for 6 hours. We found that while levels of Sirt1 are upregulated in wild-type MEFs, the levels of Sirt1 remain unchanged in TAp63−/− MEFs suggesting that TAp63 is required for upregulation of Sirt1 in response to glucose starvation (Fig. 4F). Moreover, we found that Sirt1 levels are upregulated in wild-type mice that were fasted for 18 hours, while levels of Sirt1 do not change after this treatment in livers isolated from TAp63−/− mice (Fig. 4G). These data indicate that TAp63 is necessary for the transcriptional upregulation of Sirt1 in the CR response.
AMPK is another critical regulator of lipid and glucose metabolism. Activation of AMPK results in an increase in ATP production by increasing fatty acid oxidation and glucose uptake. Dysfunction of hepatic AMPK induces lipid accumulation and hyperlipidemia, which is associated with diabetes (Zang et al., 2006). Given the importance of the LKB1/AMPK in the regulation of fat oxidation and gluconeogenesis and the accumulation of lipids in the tissues from TAp63−/− mice, we examined the protein levels of AMPKα and LKB1 in livers from wild-type and TAp63−/− mice. We found that in 4 week old TAp63−/− mice expression of AMPKα is significantly lower in muscle and fat compared to wild-type littermates (Figure 4H). Expression of LKB1 is lower only in the muscle of TAp63−/− mice at 4 weeks of age (Figure 4H). To determine whether this was an age-associated phenomenon, we assessed AMPKα and LKB1 expression in 5 month old wild-type and TAp63−/− mice fed control, CR, or high fat diets. We found that AMPKα is low in muscle, fat, and liver of TAp63−/− mice fed a control or CR diet (Figure 4I–K). Five month old TAp63−/− mice fed a high fat diet expressed low levels of AMPKα in fat and liver only and not in muscle (Figure 4I–K). We found that LKB1 expression is lower in muscle, fat, and liver of 5 month old TAp63−/− mice fed a control diet (Fig. 4I–K). When these mice were fed a CR or high fat diet, significant differences in LKB1 expression was detected in fat and liver only and not in muscle (Fig. 4I–K). Taken together, these data indicate that TAp63 regulates expression of AMPKα and LKB1 in a tissue and context specific manner and seems to play a more significant role in fat and liver.
We next asked whether levels of the catalytic subunits of AMPK, AMPKα1 and AMPKα2, were decreased in TAp63−/− mice (Fig. 4L&M). We found that the level of AMPKα2 was significantly lower in young TAp63−/− mice fed a control or high fat diet (Fig. 4L). These data suggest that AMPKα2 is the subunit that is transcriptionally regulated by TAp63 in the liver.
To determine whether metabolic defects led to defects in the development of metabolic tissues in TAp63−/− mice, we performed histological analysis and electron microscopy of cross sections of fat and muscle (Suppl. Fig. 2). We found no significant differences between these tissues in wild-type and TAp63−/− mice (Suppl. Fig. 2).
Previous studies have shown that AMPKα is phosphorylated and activated by the tumor suppressor and upstream kinase, LKB1, in response to metformin (Shaw et al., 2004; Shaw et al., 2005; Woods et al., 2003). To determine whether TAp63 is required for the response to metformin and whether AMPKα is phosphorylated in the absence of TAp63, we tested the activation of AMPKα and TAp63 in livers from 5 month old wild-type and TAp63−/− mice treated with metformin. We have previously shown that TAp63γ can be activated and accumulates in response to multiple stresses, including DNA damage and wound healing (Su et al., 2009). Importantly, we found that TAp63γ accumulates in response to metformin in livers of wild-type mice at 5 months of age (Fig. 5A). Additionally, we found that LKB1 is upregulated with a concomitant phosphorylation of AMPKα in wild-type livers as reported previously (Canto et al., 2010; Shaw et al., 2004; Shaw et al., 2005; Woods et al., 2003) while no such regulation occurred in the absence of TAp63 (Fig. 5A). We also measured phosphorylation of a downstream target of AMPK, acetyl-CoA carboxylase, ACC1, and found that levels of phosphorylated ACC1 are significantly lower in livers of TAp63−/− mice (Fig. 5B) indicating that AMPKα is functionally inactive in the absence of TAp63. To further investigate the kinetics of the regulation of the LKB1-AMPK pathway by TAp63, we measure protein expression of TAp63, LKB1, and phosphorylated AMPKα in AML-12 hepatocytes after treatment with metformin for 30 minutes, 2 hours, 6 hours, and 24 hours. We found that TAp63γ accumulates 30 minutes after treatment with metformin with a concomitant phophorylation of AMPKα (Fig. 5C). Importantly, we also found that levels of TAp63γ accumulates in vivo, in mouse livers, one hour after metformin treatment with peak expression at 2 hours. High expression of TAp63γ correlated with peak expression of LKB1 and phosphorylation of AMPKα in wild-type mice (Fig. 5D). We did note that in 5 month old TAp63−/− mice there was not a significant increase in LKB1 (Figure 5A) while in young 4 week old TAp63−/− mice there was still an increase in LKB1 (Figure 5D) indicating an age related factor in TAp63’s ability to regulate LKB1 in response to metformin. These data indicate that TAp63 is required for LKB1 activation and downstream phosphorylation of AMPKα in response to metformin in older mice.
Because TAp63 deficiency leads to a blunted activation of the LKB1-AMPK pathway, we next asked whether metformin could lower blood glucose levels in TAp63−/− mice. We treated 8 wild-type and TAp63−/− mice at 4 weeks of age with 250 mg/kg body weight of metformin. We found significantly higher levels of glucose in the blood of TAp63−/− mice compared to wild-type mice (Fig. 5E), indicating a poor response to metformin in the absence of TAp63. Taken together, these data indicate that TAp63γ responds to metformin and is critical for the downstream responses that regulate glucose levels and utilization.
We observed low levels of Sirt1, AMPKα, and LKB1 in the absence of TAp63 suggesting that they may be transcriptionally regulated by TAp63. To determine if this is the case we performed chromatin immunoprecipitation (ChIP) and Luciferase reporter gene assays. We identified a putative p63 response element within the Sirt1 promoter located 2016 nucleotides upstream of the start site (Supplemental Table 1). Indeed, in a ChIP assay, we found that TAp63 binds to this response element but not a non-specific binding site located 2000 nucleotides upstream of the p63 binding site (Fig. 6A). We next performed a Luciferase reporter gene assay and found that the TAp63β and γ isoforms can activate the Sirt1-luciferase reporter gene 4 to 5 times over vector alone (Fig. 6B and C). p53 and TAp63α were unable to activate this reporter gene (Fig. 6C). To confirm the specificity of the identified p63 binding site on the Sirt1 promoter, we generated a Sirt1-luciferase reporter gene with a mutated p63 binding site (Fig. 6B). The p63 isoforms were unable to activate this reporter gene and an additional luciferase reporter gene with a nonspecific p63 binding site (Fig. 6B and C). These data indicate that TAp63 binds to the Sirt1 promoter and can transcriptionally activate it and that this regulation may be critical for Sirt1 to activate the CR response.
Importantly, we also found that AMPKα2 mRNA and protein levels are significantly lower in metabolic tissues from TAp63−/− mice compared to their wild-type counterparts (Fig. 4H–M). To determine whether TAp63 is a transcriptional regulator of AMPKα2, we preformed ChIP and luciferase analyses. We identified a TAp63 binding site 1235 nucleotides upstream of the start site (Supplemental Table 1). Interestingly, we could only detect binding of TAp63 at the AMPKα2 promoter in cells treated with 1mM metformin, an activator of AMPK and a drug used to treat type 2 diabetes, suggesting that TAp63 may activate the AMPK pathway in response to metformin. We also found that TAp63β and γ isoforms can activate an AMPKα2-luciferase reporter gene containing the identified binding site (Fig. 6D–F). AMPKα2 - luciferase reporter gene with mutations in the p63 binding site or a non-specific binding site resulted in a complete abrogation of transactivation (Fig. 6E and F). These data indicate that TAp63 is a transcriptional activator of AMPKα2 and that this regulation may be critical for the regulation of energy metabolism.
LKB1 protein expression was also found to be significantly lower in TAp63−/− mice (Fig. 4H–K). To determine whether TAp63 is a transcriptional regulator of LKB1, we again performed ChIP (Fig. 6G) and luciferase analyses (Fig. 6H & I) and found a TAp63 binding site within intron 1 of LKB1 (Supplemental Table 1). This site was activated by TAp63α and TAp63β in luciferase reporter assays (Fig. 6I). Taken together, these data indicate that TAp63 can bind to the Sirt1, AMPKα2, and LKB1 promoters and can transcriptionally activate them.
We also examined the ability for TAp63 to bind to the Sirt1, AMPKα2, and LKB1 promoters in hepatocytes (AML-12 cells) where we found that levels of TAp63 are upregulated in response to metformin (Fig. 6J) or glucose starvation (Fig. 6K). In AML-12 cells, we found that TAp63γ is upregulated in response to metformin (Fig. 7J) resulting in binding to the Sirt1 (Fig. 6L), AMPKα2 (Fig. 6M), and LKB1 promoters (Fig. 6N). Additionally, we found that TAp63γ is upregulated in AML-12 hepatocytes in response to glucose starvation (Fig. 6K) resulting in binding to the Sirt1 (Fig. 6O), AMPKα2 (Fig. 6P), and LKB1 promoters (Fig. 6Q). These data indicate that metabolic stress such as starvation or addition of metformin results in upregulation of TAp63γ and downstream transcriptional regulation of these key metabolic regulators.
We have shown that TAp63 transcriptionally activates Sirt1 (Fig. 6A–C, L, & O). Sirt1 is an important regulator in the response to CR by regulating the expression of PEPCK and GLS2 (a gene that encodes a mitochondrial glutaminase catalyzing the hydrolysis of glutamine to glutamate), two critical enzymes for mice to adapt to CR. Given the low levels of Sirt1 in tissues from TAp63 deficient mice, we tested the CR response in TAp63−/− mice. To do this, we fed 15 TAp63−/− and wild-type mice a CR diet beginning at 4 weeks of age. After 16 weeks of being fed a CR diet, we measured fasting levels of triglycerides and glucose in blood from these mice (Supp. Fig. 3A & B). Importantly, we found that triglycerides were elevated in TAp63−/− mice (Supp. Fig. 3A) and that glucose levels were lower in TAp63−/− mice than in wild-type mice (Supp. Fig. 3B). These data are consistent with our previous data indicating defects in fat utilization, fatty acid oxidation, and glucose utilization in the TAp63−/− mice (Figs. 1–3). Taken together these data demonstrate that the metabolic phenotype of the TAp63−/− mice is similar to that of the Sirt1−/− mice and that TAp63 upregulates Sirt1 in a tissue specific manner in response to CR.
To determine whether the metabolic defects observed in the TAp63−/− MEFs could be rescued, we transfected wild-type and TAp63−/− MEFs with vectors expressing Sirt1, AMPKα2, LKB1, or Sirt1 and LKB1 in combination (Supp. Fig. 3C). To ask whether expression of these genes could rescue the defects in the mitochondria of TAp63−/− MEFs, we measured mitochondrial membrane potential using, the JC-1 assay. We found that expression of Sirt1 or AMPKα2 partially rescues the mitochondrial defect while expression of both Sirt1 and AMPKα2 completely rescue the mitochondrial defect of the TAp63−/− MEFs (Supp. Fig. 3D). We also asked whether the rate of basal oxygen consumption in TAp63−/− MEFs could be restored to wild-type levels by expression of Sirt1, AMPKα2, or LKB1. Indeed, we found that expression of Sirt1 in TAp63−/− MEFs partially rescues oxygen consumption of TAp63−/− MEFs while expression of AMPKα2 completely rescues the defect in the TAp63−/− MEFs (Supp. Fig. 3E). These data indicate that re-expression of Sirt1, AMPKα2, or LKB1 can rescue the defects of the TAp63−/− MEFs and indicate that regulation of these genes by TAp63 leads to the metabolic defects in TAp63 deficient cells.
To ask whether expression of Sirt1, AMPKα2, or LKB1 could rescue the defects of the TAp63−/− mice in vivo, we infected TAp63 deficient mice at 4 weeks of age with adenoviruses expressing these genes via tail vein injection (Fig. 7). Expression of AMPKα2 in the livers of TAp63−/− mice (Fig. 7A) resulted in a rescue of expression of CPT1, a key regulator of fatty acid metabolism (Fig. 7B & C). We did not see significant rescue in respiration, glucose tolerance, or fatty liver disease in the mice (Supp. Fig. 4). This may be due to the age of the mice or that the method of delivery of virus to the liver is insufficient to rescue all of the phenotypes, i.e. expression of AMPKα2 needs to be expressed in muscle and fat or other tissues in the TAp63−/− mice to fully rescue the metabolic defects. Expression of LKB1 rescued the expression of AMPKα and phosphorylation of AMPKα in the livers of TAp63−/− mice (Fig. 7D). We also found that STRAD is expressed at low levels in TAp63−/− mice, on of the other components of the LKB1-MO25-STRAD complex (Fig. 7D) while no change was seen in MO25. Importantly, TAp63−/− mice expressing LKB1 responded to metformin similar to its wild-type counterparts (Fig. 7E). These data indicate that TAp63 regulates LKB1 and is critical for the response to metformin. Lastly, we found that expression of Sirt1 in TAp63−/− mice resulted in wild-type expression of PEPCK and GLS2 in the liver (Fig. 7C). Also, the levels of triglycerides in the blood are restored to wild-type levels in TAp63−/− mice expressing Sirt1 (Fig. 7G). Taken together, these data indicate that re-expression of Sirt1, AMPKα2, or LKB1 can rescue some of the metabolic defects of the TAp63−/− mice at 4 weeks of age and indicate that regulation of these genes by TAp63 is critical for lipid and glucose metabolism.
The roles of tumor suppressor genes in metabolism are an area of intense interest and research (Bensaad et al., 2006; Jones and Thompson, 2009; Matoba et al., 2006). The altered mechanisms used by cancer cells to circumvent limiting nutrients are key to the survival and growth of a tumor. Here, we show that the tumor suppressor gene and p53 family member, TAp63, plays critical roles in regulating energy metabolism. We found that TAp63−/− mice develop obesity, glucose intolerance, and insulin resistance. The TAp63−/− mice have defects in fatty acid oxidation and display mitochondrial dysfunction. These phenotypes of the TAp63−/− mice are reminiscent of tissue-specific deletion of Sirt1 and AMPKα2−/− mouse models (Cohen et al., 2009; Purushotham et al., 2009; Viollet et al., 2003). We found that TAp63 transcriptionally activates Sirt1, AMPKα2, and LKB1. Loss of expression of these factors leads to defects in lipid utilization, fatty acid synthesis, fatty acid oxidation, and insulin resistance. Consequently, the TAp63−/− mice exhibit symptoms of premature aging (Su et al., 2009), obesity, and type 2 diabetes. Importantly, we were able to rescue some of the metabolic defects of the TAp63−/− mice by expressing Sirt1, AMPKα, or LKB1 in the liver of TAp63−/− mice at 4 weeks of age. These data reveal a role for TAp63 in regulating energy metabolism and identify TAp63 as master upstream regulator of lipid and glucose metabolism. Given the extensive interaction between the p53 family members, these data have important implications for understanding how the family of p53 tumor suppressor genes regulates metabolism in cancer cells.
The roles of p63 have been extensively studied in epidermal morphogenesis (Mills et al., 1999; Yang et al., 1999), epidermal homeostasis (Su et al., 2009), and tumor suppression (Flores et al., 2005; Guo et al., 2009; Su et al., 2010). While the vast majority of p63 studies have been performed using mice that are deficient for all isoforms of p63 (Mills et al., 1999; Yang et al., 1999), our TAp63 isoform specific knock out mice have unveiled functions for this isoform that have not been previously appreciated in studies using the p63−/− mice. Here we show that TAp63 is a critical transcriptional regulator of genes involved in lipid and glucose metabolism. Given the extensive interaction of the p53 family members (Flores et al., 2005; Flores et al., 2002; Lang et al., 2004; Olive et al., 2004), our findings have important implications for the p53 family in regulating metabolism in cancer. Previous studies have shown that p53 plays critical roles in regulating metabolism. The primary mechanism of action of p53 in this regard is through transcriptional activation of downstream target genes. These include TIGAR (TP53-induced glycolysis regulator), SCO2 (synthesis of cytochrome c oxidase), (Bensaad et al., 2006; Matoba et al., 2006), sestrin 1 and 2 (Budanov and Karin, 2008) and AMPKβ (Feng et al, 2007). Cells within a tumor are nutrient deprived and use aerobic glycolysis as their major source of energy, the so-called Warburg effect. The switch to this mode of metabolism is critical for the survival of cells proliferating in a tumor with limited blood and oxygen. Here, we found that TAp63 transcriptionally regulates Sirt1, AMPKα2, and LKB1 to regulate the utilization of fat and glucose. This has important implications for tumor cells with mutant p53. Mutant p53 has been shown to bind to TAp63 and inhibit its transcriptional activity (Gaiddon et al., 2001; Lang et al., 2004; Olive et al., 2004). Mice lacking TAp63 are highly prone to metastatic tumors (Su et al., 2010) and to metabolic disorders, as we have shown here. The ability of mutant p53 to inactivate TAp63 suggests there may be interplay between p53 family members in regulating cellular metabolism in cancer. Therefore, our work has opened venues of investigation for the p53/p63/p73 field.
Consistent with a role for TAp63 in metabolic stress, we found that TAp63 responds to a drug used to treat type 2 diabetes, metformin. The levels of TAp63γ are elevated in response to metformin (Fig. 5C). In turn, levels of LKB1 increase, which regulates not only energy metabolism through its ability to phosphorylate AMPKα but is also a potent tumor suppressor gene (Shaw et al., 2004; Shaw et al., 2005; Woods et al., 2003). Therefore, we found two modes by which TAp63 regulates AMPKα. One is through regulating the mRNA levels of AMPKα2 itself and the other is by regulating its upstream kinase, LKB1. The reasons for this multi-level of regulation are intriguing and areas for future research. Possible reasons for this complex regulation could be due to different cellular and metabolic stress contexts.
We have shown here that TAp63 is a critical regulator of lipid and glucose metabolism, loss of which leads to obesity and diabetes. Not only is TAp63 a potent suppressor of tumorigenesis and metastasis (Su et al., 2010), it is also a central regulator and integrator of the metabolic response to CR and starvation. These findings provide exciting avenues of research for the p53 family and cancer fields.
Forty-five TAp63−/− and wild-type (WT) male mice on an enriched C57BL/6 background at 4 weeks of age were randomly assigned to three groups of 15 mice each. Each group was fed a control diet (con) (D12329, Research Diets, New Brunswick, NJ) (11% kcal from fat), high fat diet (HF) (D12492, Research Diets, New Brunswick, NJ) (60% kcal from fat), or calorie restricted diet (CR) (60% of the average daily food intake of the control group) for 16 weeks. Each mouse was housed individually. All procedures were approved by the IACUC at the University of Texas - MD Anderson Cancer Center.
Intraperitoneal glucose tolerance test (IPGTT) was performed by injecting D-glucose (2g/kg body weight) intraperitoneally into mice that were fasted for 18 hours. Insulin tolerance test (ITT) was performed by injecting mice with insulin (0.75 U/kg, Humulin-N from Ely Lilly) in ~ 0.1 ml 0.9% NaCl intraperitoneally. Blood was collected from tail bleeds every 20–30 minutes over a 2 hour period and whole blood glucose was measured using a Precision Xtra advanced diabetes management system (MediSense). Areas under the curves for the IPGTT and ITT were calculated using PRISM5 software (GraphPad).
Horizontal activity, oxygen consumption rate (VO2), Carbon dioxide release (VCO2), respiratory exchange ratio (RER) and heat production were measured under a consistent environmental temperature and light cycle using an indirect calorimetry system (TSE systems). After 3 days of acclimation to the metabolic chamber, VO2 was measured in individual mice at 5 min intervals for 72 hours. Data are normalized with respect to lean body weight. RER is the ratio of VCO2 to VO2. Horizontal activity was measured on x, y, and z axes by using infrared beams to count the beam breaks during a specified measurement period. Fat and lean tissue mass was determined in living, non-anesthetized mice by using a magnetic resonance imaging (MRI) at the Baylor College of Medicine Metabolism Core.
Mice were fasted for 18 hours prior to collecting blood. Serum was analyzed using a Insulin Mouse Ultrosensitive ELA kit (cat# 80-INSMSU-E01, ALPCO), Mouse Adiponectin ELISA kit (cat# CYT286, Chemicon), Mouse Leptin ELISA Kit (Millipore .a Lactate Reagent Kit), and the COBAS INTEGRA 400 plus System (Roche) to measure serum triglycerides and total cholesterol.
Frozen sections of livers from wild-type and TAp63−/− mice were fixed with 10% formalin. Oil Red O (Polysciences, Inc.) staining was performed using 60 % isopropanol saturated with Oil Red O dye for one hour at room temperature.
Keratinocytes and AML-12 hepatocytes used to assess TAp63 binding at the AMPKα2 promoter were treated with metformin 1mM for 6 hours. In some ChIP experiments, AML12 cells starved of serum and glucose for 6 hours. Anti-pan-p63 antibody (4A4, Abcam for Keratinocytes, 4A4, Santa Cruz for AML12), or IgG were used for immunoprecipitation. Putative p63 binding sites were scanned within 5000 bp upstream of the 5′ UTR and intron 1 for the consensus p53/p63 binding site (el-Deiry et al., 1992; Su et al., 2010; Yang et al., 2006). All primers used for ChIP-PCR are listed in Supplemental Table 2. ABI Step One Plus real-time PCR and SYBR green PCR master mix (Applied Biosystems) were used for quantitative real time PCR.
2×109 plaque-forming units (PFU) active rat Ad-AMPKα2 (Eton Bioscience), 1×109 (PFU) human Ad-LKB1 (Vector BioLabs) or 2×109 PFU human Ad-Sirt1 (Vector BioLabs) were injected through the tail vein of TAp63−/− mice. Forty-eight hours after injection, livers were collect for western blot analysis and qRT-PCR. Metformin sensitivity was measured after metformin injection (250mg/kg) using mice infected with human Ad-LKB1.
All data are represented as mean ± SEM. Data were analyzed using one-way ANOVA test or Student’s t test for comparison between two groups. A p value of 0.05 was considered significant.
This work was supported by grants to E.R.F. from the American Cancer Society (RSG-07-082-01-MGO), the Mel Klein Foundation, and the Hildegardo E. and Olga M. Flores Foundation. This work was supported in part by NCI-R01 (R01CA160394), NCI-R01 (R01CA134796), and CPRIT (RP120124) to E.R.F., NCI-Cancer Center Core Grant (CA-16672)(U.T. M.D. Anderson Cancer Center) and a Career Development Award from the Genitourinary Cancer SPORE (NCI CA091846). E.R.F. is a scholar of the Leukemia and Lymphoma Society of America, the Rita Allen Foundation, and the V Foundation for Cancer Research. D.C. was funded by a CPRIT training grant (RP101502). We would like to thank L. C. B. Chan, M.D. and P. Saha, Ph.D. for scientific discussion and technical advice and the Mouse Metabolism Core at Baylor College of Medicine funded by NIH P30 DK079638.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.