Potential conflict of interest: Nothing to report.
Age declines liver functions, leading to the development of age-associated diseases. A member of the sirtuins family, SIRT1, is involved in the control of glucose homeostasis and fat metabolism. Because aging livers have alterations in glucose and fat metabolism, we examined a possible role of SIRT1 in these alterations. We found that aged livers have a reduced expression of SIRT1 and have lost proper control of the regulation of SIRT1 after partial hepatectomy (PH). Down-regulation of SIRT1 in the liver of old mice is mediated by CCAAT/Enhancer Binding Protein/histone deacetylase 1 (C/EBPβ-HDAC1) complexes, which bind to and repress the SIRT1 promoter. In the livers of young mice, SIRT1 is activated after PH and supports high levels of glucose and triglycerides during liver regeneration. In old mice, however, C/EBPβ-HDAC1–mediated repression of the SIRT1 promoter blocks activation of SIRT1, leading to low levels of glucose and triglycerides during liver regeneration. Down-regulation of SIRT1 in the livers of young mice resulted in alterations similar to those observed in the livers of old mice, whereas the normalization of SIRT1 in the livers of old mice corrects the levels of glucose and triglycerides after PH. The normalization of SIRT1 in old mice also improves liver regeneration via the elimination of the C/EBPα-Brm complex. These studies showed a critical role of the reduction of SIRT1 in age-associated liver dysfunctions and provide a potential tool for the correction of liver functions in old patients after surgical resections. (HEPATOLOGY 2011;)
A member of histone deacetylases, SIRT1 (sirtuin 1), is expressed in many tissues, including the liver, and controls activities of proteins by regulation of their promoters and by direct deacetylation of proteins such as p53.1 SIRT1 regulates physiological processes, including apoptosis, fat metabolism, glucose homeostasis, and neurodegeneration.2 SIRT1 also provides a functional link between metabolism, chromatin structure, and aging.1, 3 Knockdown of SIRT1 in the liver causes mild hypoglycemia, increased systemic glucose and insulin sensitivity, and decreased glucose production.4 It has been shown that the activity of SIRT1 is reduced in aged tissues.5
The liver is one of the critical tissues that regulate glucose and fat metabolism. Two members of the CCAAT/Enhancer Binding Protein (C/EBP) family, C/EBPα and C/EBPβ, are expressed in the liver at high levels and play critical roles in the regulation of liver biology.6, 7 C/EBPα and C/EBPβ proteins are transcription factors that interact with promoters of many genes and activate the expression of corresponding proteins.8 The generation of knockout models for these proteins showed that they regulate a variety of biological processes in the liver and body homeostasis.7, 9 During the process of aging, many liver functions are altered, leading to the development of age-associated diseases.10 Activities of C/EBPα and C/EBPβ are changed in aged livers. Aging increases the amounts of C/EBPα-HDAC1-Brm complex, which also contains E2F411 and amounts of C/EBPβ-HDAC1 (histone deacetylase 1) complexes.12, 13 Our recent papers with C/EBPα-S193D knock-in mice have demonstrated the critical role of C/EBPα-HDAC1 complexes in age-like dysfunctions of the liver.14
Antibodies to C/EBPβ (C19), C/EBPα (14AA), and HDAC1 (H-51) were from Santa Cruz Biotechnology (Santa Cruz, CA), and Anti-Sir2 polyclonal antibody was from Millipore (Billerica, MA). Antibodies to acetyl-histone H3 (Lys9) and histone H3 trimethyl Lys9 were from Abcam (Cambridge, UK). Monoclonal anti-β-actin antibody was from Sigma-Aldrich (St. Louis, MO). The bromodeoxyuridine (BrdU) uptake assay kit was from Invitrogen (Carlsbad, CA). Coimmunoprecipitation (Co-IP) studies were performed as previously described.13, 14
Animals, Treatments of Old Mice with Growth Hormone, and partial hepatectomy.
Experiments with animals were approved by the Institutional Animal Care and Use Committee at Baylor College of Medicine (protocol AN-1439). In this article, we used young (4-6 months) and old (22-24 months) mice. Livers were harvested and kept at −80°C. Procedures for the growth hormone (GH) treatment of old mice and for partial hepatectomy (PH) are described in the Supporting Information.
Hep3B2 cells were transfected using the Fugene 6 transfection reagent (Roche Molecular Biochemicals, Mannheim, Germany), according to the manufacturer's protocol. Twenty-four hours after transfection, the luciferase assay was performed using the Dual-Luciferase Reporter Assay System (Promega, Madison, WI). Luciferase activity was normalized by the Renilla internal control.
Protein Isolation and Western Blotting.
Cytoplasmic and nuclear extracts were isolated from the livers of young and old mice, as described in a previous article.15 Previous studies showed that both the expression and activity of SIRT1 are regulated in a circadian manner.16 Therefore, all studies described in this article were performed with animals sacrificed at 8:00 a.m. to 10:00 a.m.
Chromatin Immunoprecipitation Assay.
The chromatin immunoprecipitation assay (ChIP) assay was performed, as described in our previous articles, using the Chip-It kit.17 The sequences of primers used in the work were as follows: SIRT1: 5′-TCCTGGCCC TGTCATTTTAG-3′ (forward) and 5′-GACCCAC CAGGTCCTCTGTA-3′ (reverse). The final volume of polymerase chain reaction (PCR) reactions was 50 uL. We usually load 5 uL of the PCR reaction on the agarose or PAAG.
Real-Time Quantitative Reverse-Transcriptase PCR.
The RNeasy Plus minikit (Qiagen, Germantown, MD) was applied to extract total RNA, according to the manufacturer's instructions. The sequences of PCR primers were as follows: SIRT1: 5′-ACTGCAGAAAC TTTTAGCCTTTCAA-3′ and 5′- GGCAATGTTCC AAAGAAGTCTGT-3′; mouse β-actin: 5;-AAGTGC TTCTAGGCGGACTGTT-3′ and 5′-TTTTCTGC GCAAGTTAGGTTTTG -3′. Additional procedures are described in the Supporting Information.
Old Mice Fail to Activate SIRT1 After PH.
Because SIRT1 is an important regulator of liver biology, we began investigations of mechanisms that regulate the expression of SIRT1 in the liver. We found that protein levels of SIRT1 and SIRT1 messenger RNA (mRNA) are reduced in aged livers by approximately 3-4-fold (Fig. 1A,B). The liver is a unique tissue that is able to completely regenerate itself after partial hepatectomy (PH).18 However, liver proliferation is impaired in old mice, and liver functions are also reduced in the course of regeneration. Therefore, we asked whether the reduction of SIRT1 in old mice might be involved in liver dysfunctions. We found that protein levels of SIRT1 are significantly increased in the livers of young mice at 36 and 48 hours after partial hepatectomy, but SIRT1 is not elevated in livers of old mice after PH. One of the important liver functions after PH is the support of blood glucose levels,19 as glucose is an essential source for the promotion of liver regeneration.20 We found that glucose and TG levels are reduced in old and young mice at 24 hours after PH; however, the reduction of glucose and TGs in old mice is much deeper (Fig. 1D,E), and young animals recover glucose and TGs at 72 hours after PH; but, old mice have a weak recovery of both.
Elevation of SIRT1 After PH in Young Mice Is Required for Support of Glucose Homeostasis and for Liver Proliferation.
We next asked whether the increase of SIRT1 after PH is required for liver regeneration. We inhibited the expression of SIRT1 by siRNA and performed an examination of liver proliferation and liver functions within 72 hours after PH. Control animals were treated with an unrelated RNA of random composition. SIRT1 was completely inhibited by short-interfering RNA (siRNA) at all stages of liver regeneration (Fig. 2A). We next examined levels of glucose and TGs in young mice treated with siRNA to SIRT1, and found that the recovery of glucose and TG was much slower in young mice with inhibited SIRT1 (Fig. 2C). We next examined the expression of cell-cycle proteins and BrdU uptake (i.e., DNA synthesis), and found that the expression of PCNA and cyclin D1 and DNA synthesis are reduced in young mice with inhibited SIRT1 (Fig. 2B,D,E). Thus, these studies showed that the inhibition of SIRT1 in young mice leads to the reduction of liver proliferation and to impaired recovery of glucose and TGs after PH.
HDAC1-C/EBPβ Complexes Repress the SIRT1 Promoter in Livers of Old Mice.
We next examined the mouse and human SIRT1 promoters for the presence of binding sites for transcription-factor activities, which are altered in the livers of old mice. These studies revealed that both mouse and human SIRT1 promoters contain several C/EBP sites, and that C/EBPααand C/EBPβ positively regulate the promoters in tissue culture systems (see Supporting Figs. 1 and 2). Therefore, we examined the hypothesis that C/EBP proteins might be positive regulators of the SIRT1 promoter in the livers of young mice; whereas the complexes of C/EBPβ with HDAC1 are negative regulators of the SIRT1 promoter in old mice. C/EBPβ alone activates the SIRT1 promoter; however, simultaneous transfections of C/EBPβ and HDAC1 inhibit the SIRT1 promoter (Fig. 3A). We next tested whether inhibition of endogenous HDAC1 would change the activity of the SIRT1 promoter. Expression of HDAC1 was inhibited in cells transfected with an empty vector and with vector expressing C/EBPβ. The siRNA-mediated down-regulation of HDAC1 leads to the higher activation of the SIRT1 promoter (Fig. 3B). We next inhibited C/EBPβ by siRNA, and found that HDAC1 is not able to repress the SIRT1 promoter in Hep3B2 cells with inhibited C/EBPβ (Fig. 3C). Thus, these studies demonstrated that HDAC1 represses the SIRT1 promoter via interactions with C/EBPβ.
We next examined whether the SIRT1 promoter might be repressed by C/EBPβ-HDAC1 complexes in the livers of old mice. The livers of old mice contain elevated levels of both HDAC1 and C/EBPβ (Fig. 3D) and higher amounts of the C/EBPβ-HDAC1 complexes (Fig. 3E, upper). The ChIP assay (Fig. 3E, bottom) showed that C/EBPα and C/EBPβ could be observed on the SIRT1 promoter in the livers of young mice, and that histone H3 is acetylated at K9. In the livers of old mice, C/EBPα was reduced, whereas C/EBPβ and HDAC1 were increased on the SIRT1 promoter. These alterations were accompanied with the increase of histone 3 trimethylation at K9, demonstrating that the C/EBPβ-HDAC1 complexes inhibited the SIRT1 promoter (Fig. 3F).
Elimination of C/EBPβ-HDAC1 Complexes Correct Levels of SIRT1.
Because inhibition of HDAC1 in the livers of old mice causes cell death,17 we eliminated the C/EBPβ-HDAC1 complex by the treatment of old mice with GH. GH reduces levels of C/EBPβ and HDAC1, whereas levels of SIRT1 are increased (Fig. 4A,B). Co-IP studies revealed that the GH-mediated reduction of HDAC1 and C/EBPβ eliminates the C/EBPβ-HDAC1 complexes (Fig. 4C). We found that GH partially restores levels of SIRT1 mRNA in the livers of old mice (Fig. 4D). The ChIP assay showed that GH treatment removes C/EBPβ-HDAC1 complexes from the SIRT1 promoter (Fig. 4E). Because GH might have some additional effects on chromatin structure, we inhibited the expression of C/EBPβ by siRNA and found that this inhibition removes C/EBPβ-HDAC1 from the SIRT1 promoter and de-represses the promoter (Fig. 4E). Thus, these data show that the elimination of the C/EBPβ-HDAC1 complexes activates the SIRT1 promoter in the livers of old mice.
The C/EBPβ-HDAC1 Complex Blocks Activation of the SIRT1 Promoter in Livers of Old Mice.
The livers of old mice failed to elevate SIRT1 protein after PH (Fig. 1C). To examine whether this inhibition would take place on the level of transcription, we examined the expression of SIRT1 mRNA in young and old livers after PH. Levels of SIRT1 mRNA are 4-5-fold increased in young livers at 36 and 48 hours after PH (Fig. 5A). However, the livers of old mice contain very low levels of the SIRT1 mRNA at all stages of liver regeneration. We next examined the hypothesis that C/EBPβ-HDAC1 complex blocks the transcription of SIRT1 mRNA in the livers of old mice after PH. Examination of the protein levels of C/EBPβ and HDAC1 showed that C/EBPβ is significantly increased after PH in the young liver at all time points after PH. HDAC1 is elevated at 24 hours after PH, but returned to normal levels at 36, 48, and 72 hours. However, the livers of old mice had much higher levels of both C/EBPβ and HDAC1 in the quiescent state and after PH (Fig. 5B). Amounts of C/EBPβ-HDAC1 complexes were also 4-5-fold higher in the livers of old mice within 72 hours after PH (Fig. 5C). The ChIP assay showed that C/EBPβ-HDAC1 complexes repress the SIRT1 promoter at the zero time point and at 36 hours after PH (Fig. 5D). In young mice, we detected an increase of C/EBPβ on the SIRT1 promoter at 36 hours. PCR reactions with different numbers of cell cycles demonstrated that the PCR signals with C/EBPβ IPs from 36 hours are detected after 28 cycles, whereas the PCR signals with the zero time point are detectable after 30-32 cycles (Fig. 5E). Therefore, we conclude that C/EBPβ activates the SIRT1 promoter after PH in young mice; however, C/EBPβ-HDAC1complexes prevent activation of the SIRT1 promoter in the livers of old mice.
Ectopic Expression of SIRT1 in Livers of Old Mice Corrects Levels of Glucose and Triglycerides and Accelerates Liver Proliferation.
To determine the causal role of the reduction of SIRT1 in the impaired glucose metabolism, we delivered the FLAG-SIRT1 into old mice. Western blotting showed that FLAG-SIRT1 protein was expressed in the livers of injected mice at all time points used for the investigations (Fig. 6A). Analysis of blood parameters showed that the correction of SIRT1 in the livers of old mice significantly increased plasma-glucose and TGs levels at 36 and 48 hours after PH, compared with levels in control old mice (Fig. 6B,C). Because the inhibition of SIRT1 in young mice reduces liver proliferation and because the aging liver is characterized by a weak proliferation after PH,10 we examined the levels of PCNA and cyclin D1 during liver regeneration and found that both proteins were activated after PH in old mice with ectopic expression of SIRT1 to higher levels than in control old mice (Fig. 6D). To determine whether DNA replication might also be affected in SIRT1-treated mice, we examined BrdU uptake. These studies showed that DNA synthesis was significantly accelerated in old mice injected with SIRT1 (Fig. 6E).
Ectopic Expression of SIRT1 in Livers of Old Mice Down-regulates C/EBPα and Releases Repression of E2F Targets.
We next determined the mechanisms by which SIRT1 accelerates liver proliferation. Inhibition of liver proliferation in old mice is mediated by C/EBPα-Brm-HDAC1-E2F4 complexes, which repress E2F-dependent promoters.11 Therefore, we first examined the expression of components of the complex. Although protein levels of Brm and HDAC1 were not altered in the livers of SIRT1-injected mice, C/EBPα was dramatically inhibited at the zero time point and at 36 hours after PH in SIRT1-injected mice (Fig. 7A). Levels of E2F4 were slightly reduced, but stayed at relatively high levels. In agreement with PCNA elevation, levels of cyclin A were also increased in SIRT1-injected mice. Examination of C/EBPα mRNA showed that SIRT1 did not affect C/EBPα mRNA (data not shown), suggesting that SIRT1 down-regulated C/EBPα at levels of translation or protein stability. We next examined C/EBPα-HDAC1-Brm complexes by the Co-IP approach, and found that ectopic expression of SIRT1 eliminated the complexes (Fig. 7B). In the livers of old mice, the C/EBPα-HDAC1-Brm complex represses E2F-dependent promoters, such as the c-myc promoter.21 Therefore, we applied the ChIP assay to measure the occupation of the c-myc promoter by the complex and found that, in control mice, the C/EBPα-Brm complex is observed on this promoter before and after PH; however, SIRT1 expression leads to the removal of the complexes from the promoter (Fig. 7C). Taken together, these studies demonstrated that the normalization of levels of SIRT1 in the livers of old mice improves liver regeneration through the down-regulation of C/EBPα and after de-repression of E2F-dependent promoters (Fig. 7D).
The biological processes in the liver are controlled by a complex network of signaling pathways, which are affected by age. Recent reports demonstrated the critical role of epigenetic alterations in age-associated liver dysfunction.10, 22 Despite intensive studies of the role of SIRT1 in liver biology, the molecular mechanisms of the regulation of SIRT1 in the liver have not been previously studied. In this article, we found that the expression of SIRT1 in the liver was controlled by C/EBP proteins. Because activities of these proteins are altered with age, we have shown that SIRT1 is reduced in the livers of old mice, and that this reduction of SIRT1 is associated with the repression of the SIRT1 promoter by C/EBPβ-HDAC1 complexes. Experiments in cultured cells clearly demonstrated that HDAC1 alone did not repress the SIRT1 promoter, and that the HDAC1-dependent repression of the SIRT1 promoter required C/EBPβ.
One of the most interesting findings of our studies was that overexpression of SIRT1 in the livers of old mice corrects liver proliferation. This observation was quite unexpected, as previous studies did not suggest that SIRT1-dependent pathways might be associated with the regulation of C/EBPα. However, our new data clearly demonstrate that ectopic expression of SIRT1 resulted in a dramatic reduction of C/EBPα protein, but not C/EBPα mRNA. Interestingly, this reduction was observed 24 hours after SIRT1 injection before PH, showing that SIRT1 might target C/EBPα in the quiescent liver. Thus, our data show that the normalization of SIRT1 in the livers of old mice corrects the expression of E2F-dependent genes, leading to the acceleration of liver proliferation (Fig. 7D). Although we do not yet know the mechanisms by which SIRT1 down-regulates C/EBPα protein in the livers of old mice, this observation shows one of the possible ways to correct liver regeneration in the elderly. Further studies are required to elucidate how SIRT1 reduces protein levels of C/EBPα.
The authors thank Emily Sullivan for her excellent help in her work with animals and Il-Hwa Hong and Vladislav Sharin for their discussion of the experimental data.