p300-mediated acetylation of TRF2 is required for maintaining functional telomeres

Abstract

The human telomeric protein TRF2 is required to protect chromosome ends by facilitating their organization into the protective capping structure. Post-translational modifications of TRF2 such as phosphorylation, ubiquitination, SUMOylation, methylation and poly(ADP-ribosyl)ation have been shown to play important roles in telomere function. Here we show that TRF2 specifically interacts with the histone acetyltransferase p300, and that p300 acetylates the lysine residue at position 293 of TRF2. We also report that p300-mediated acetylation stabilizes the TRF2 protein by inhibiting its ubiquitin-dependent proteolysis and is required for efficient telomere binding of TRF2. Furthermore, overexpression of the acetylation-deficient mutant, K293R, induces DNA-damage response foci at telomeres, thereby leading to induction of impaired cell growth, cellular senescence and altered cell cycle distribution. A small but significant number of metaphase chromosomes show no telomeric signals at chromatid ends, suggesting an aberrant telomere structure. These findings demonstrate that acetylation of TRF2 by p300 plays a crucial role in the maintenance of functional telomeres as well as in the regulation of the telomere-associated DNA-damage response, thus providing a new route for modulating telomere protection function.

INTRODUCTION

Telomeres are specialized nucleoprotein complexes, which protect the ends of eukaryotic chromosomes and have been implicated in aging and cancer (1,2). Mammalian telomeres consist of duplex tandem TTAGGG repeats with 3′ single-stranded G overhang and can form the higher order structure (such as a t-loop) that provides telomere protection by preventing chromosome ends from being recognized as DNA damage (3,4). Telomeric DNA is tightly associated with the six-subunit protein complex named shelterin (5–7). The specificity of shelterin for telomeric DNA is provided by the double-stranded binding factors, TRF1 and TRF2 (8,9), and single-stranded binding protein POT1 (10). Other shelterin core components such as TIN2, RAP1 and TPP1 are recruited through the interactions with TRF1 and TRF2 (11–13). In addition, shelterin associates with several accessory factors that are distinguished from the shelterin core components (14). The accessory factors are less abundant at telomeres and appear to be transiently associated with chromosome ends. Most of these proteins are involved in DNA transactions such as DNA repair (15,16), DNA-damage signaling (17) and chromatin structure (18). Although some accessory factors have been shown to be essential for telomere protection, the mechanisms by which these proteins communicate with different signaling pathway remain largely unknown.

TRF2 protects chromosome ends by facilitating their organization into the t-loop structure (19,20). Experimental disruption of TRF2 induces a DNA-damage response at telomeres. Damaged telomeres have been shown to become associated with DNA-damage response factors such as 53BP1, MDC1, phosphorylated forms of H2AX, ataxia-telangiectasia mutated (ATM) and the Mre11/Nbs1/Rad50 complex (21), and activate the ATM signaling cascade, leading to cell cycle arrest mediated by the p53/p21 pathway (22–24). Furthermore, the DNA-damage signal induced by TRF2 disruption is abrogated in the absence of ATM, suggesting that TRF2 prevents ATM signaling pathway (25,26). TRF2 also protects chromosome ends by recruiting the shelterin accessory factors to telomeres. The Artemis-like nuclease Apollo has the ability to localize to telomeres through an interaction with TRF2 (27,28). Despite its low abundance at telomeres, Apollo knockdown results in cellular senescence and the activation of a DNA-damage signal at telomeres. PNUTS and MCPH1 have been identified as telomere-associated proteins that directly interact with TRF2 and regulate telomere length and the telomeric DNA-damage response, respectively (29). Recently, the DEAD-box RNA helicase DDX39 has been identified as a TRF2-interacting protein and is required for genome integrity and telomere protection (30).

Post-translational modifications of TRF2 have been shown to play important roles in telomere protection (31). TRF2 is rapidly and transiently phosphorylated in response to DNA damage by an ATM-dependent pathway (32,33). The phosphorylated form of TRF2 is not bound to telomeric DNA and accumulates at DNA-damage sites, suggesting that TRF2 phosphorylation plays a role in the DNA-damage response. In the case of ubiquitination, Siah1 is the first factor identified as an E3 ubiquitin ligase for TRF2 (34). During replicative senescence, p53 is activated, which induces Siah1, and thereby represses the levels of TRF2. The p53-dependent ubiquitination and proteasomal degradation of TRF2 attributes to the E3 ubiquitin ligase activity of Siah1. The MMS21 SUMO ligase of the SMC5/6 complex SUMOylates multiple telomere-binding proteins, including TRF1 and TRF2 (35). Inhibition of TRF1 or TRF2 SUMOylation prevents the localization of telomeres in promyelocytic leukemia (PML) bodies, termed alternative lengthening of telomeres (ALT)-associated PML bodies (APBs) (36), suggesting that SUMOylation of TRF1 and TRF2 facilitates telomere elongation in ALT cells by promoting APB formation. The N-terminal basic domain of TRF2 has been shown to be methylated by protein arginine methyltransferase (37). Amino acid substitutions of arginines with lysines in the basic domain induce DNA-damage response foci at telomeres, leading to induction of cellular senescence. Finally, poly(ADP-ribose) polymerases (PARP1 and PARP2) are capable of the poly(ADP-ribosyl)ation of TRF2 and reduce TRF2 binding to telomeric DNA (38,39).

In this study, we demonstrate that TRF2 interacts with and is acetylated by p300. We have identified lysine 293 as a major acetylation site within the Hinge domain of TRF2. p300-mediated acetylation increases the stability of TRF2 by inhibiting its ubiquitin-dependent proteolysis and is required for the efficient telomere binding of TRF2. We also show that lysine-to-arginine mutation of the acetylation site (K293R) results in DNA-damage response foci at telomeres. Furthermore, overexpression of K293R induces the growth arrest, which is accompanied by cellular senescence and altered cell cycle distribution. These findings suggest that the acetylation status of TRF2 provides a molecular switch that controls the levels of TRF2 at telomeres and represents a new route for maintaining functional telomeres.

MATERIALS AND METHODS

Cell culture

The human embryonic kidney cell line HEK293, the fibrosarcoma cell line HT1080 and the prostate cancer cell line DU145 were cultured in Dulbecco’s Modified Eagle’s Medium supplemented with 10% fetal bovine serum, 100 units/ml penicillin and 100 μg/ml streptomycin in 5% CO₂ at 37°C. The expression vectors were transiently transfected using Lipofectamine-PLUS reagent according to the manufacturer’s protocol (Invitrogen, Carlsbad, CA, USA). To establish stable cell lines, the TRF2-Myc and K293R-Myc expression vectors were transfected into HT1080 and DU145 cells. As a control, cells were stably transfected with the empty vector. After selection with 4 μg/ml blasticidin (Invitrogen, Carlsbad, CA, USA), multiple independent single clones were isolated and checked for protein expression by immunoblotting analysis with anti-TRF2 or anti-Myc antibody.

Glutathione S-transferase (GST) pulldown, immunoprecipitation and immunoblotting

GST pulldown and immunoprecipitation were performed as described previously (40). Briefly, the expression vectors were transiently transfected into HEK293 cells followed by lysis. For the GST pulldown assay, the cellular supernatants were precleared with glutathione-Sepharose 4B (Amersham Biosciences, Piscataway, NJ, USA) and incubated with glutathione-Sepharose beads containing GST fusion proteins. For immunoprecipitation, the supernatants were preincubated with protein A-Sepharose (Amersham Biosciences, Piscataway, NJ, USA) and incubated with primary antibodies precoupled with protein A-Sepharose beads. The precipitated proteins were washed extensively and subjected to immunoblotting analysis. Immunoprecipitation and immunoblotting were performed using anti-TRF2 (Upstate Biotechnology, Waltham, MA, USA), anti-TRF1 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-acetylated lysine (Cell Signaling Technology, Danvers, MA, USA), anti-HA (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-Flag (Sigma, St. Louis, MO, USA), anti-V5 (Invitrogen, Carlsbad, CA, USA), anti-Myc (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-p300 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-phospho-TRF2 (T188) (Upstate Biotechnology, Waltham, MA, USA), anti-ATM (Cell Signaling Technology, Danvers, MA, USA), anti-phospho-ATM (S1981) (Cell Signaling Technology, Danvers, MA, USA) and anti-actin (Sigma, St. Louis, MO, USA) antibodies.

In vitro acetylation assay

The GST fusion proteins were expressed in Escherichia coli Rosetta (DE3) LysS cells and purified with glutathione-Sepharose 4B matrix (Amersham Biosciences, Piscataway, NJ, USA). The purified GST-TRF2 proteins (2.5 μg) were incubated with purified p300 proteins (0.5 μg) in the presence of 0.5 μM acetyl coenzyme A (Sigma, St. Louis, MO, USA) or 0.05 μCi [¹⁴C] acetyl coenzyme A (Amersham Biosciences, Piscataway, NJ, USA) in 50 μl reaction buffer [50 mM Tris–HCl pH 8.0, 10% (v/v) glycerol, 0.1 mM ethylenediaminetetraacetic acid, 1 mM dithiothreitol, 10 mM sodium butyrate] at 30°C for 1 h. p300-HA was immunoprecipitated by anti-HA antibody from lysates of transfected HEK293 cells and used as acetyltransferase. The reactions were analysed on Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE), and gels were stained with Coomassie blue. Acetylated TRF2 was detected by immunoblotting using polyclonal anti-acetylated lysine antibody or by autoradiography.

Peptide identification using liquid chromatography-tandem mass spectrometry (LC-MS/MS)

Agilent 6530 Accurate-Mass Q-TOF (Agilent, Wilmington, DE, USA) was used for the peptide identification. Peptide separation was performed with a nano Chip HPLC system (Agilent, Wilmington, DE, USA). The mobile phase A for LC separation was 0.1% formic acid in deionized water and the mobile phase B was 0.1% formic acid in acetonitrile. The chromatography gradient was designed for a linear increase from 2% B to 60% B in 10 min, 90% B in 2 min and 2% B in 3 min. The flow rate was set at 600 nl/min. Full-scan range was set to m/z 3500–1200. Maximum five precursor ions were selected for MS/MS. The dynamic exclusion time for precursor ion m/z values was 30 s. Singly charged ions were rejected for MS/MS.

Chromatin immunoprecipitation

Chromatin immunoprecipitation (ChIP) was performed as previously described (41) with the following modifications. Briefly, cells were fixed with 1% formaldehyde in phosphate-buffered saline (PBS) and lysed, followed by sonication to obtain chromatin fragments with an average size of 600 bp. Lysates were immunoprecipitated with anti-Flag antibody and supplemented with protein A-Sepharose beads. The immunocomplexes were heated at 65°C for 4 h to reverse the cross-links. Slot blots were hybridized with a 300-bp random-labeled TTAGGG probe or an Alu probe. The quantification of the percent precipitated DNA was done with ImageQuaNT software (Molecular Dynamics, Sunnyvale, CA, USA), and the percentage of each immunoprecipitation sample was calculated based on the signal relative to the corresponding total DNA signal.

Small interfering RNA transfections

The small interfering RNA (siRNA) target sequences specific for p300 were 5′-GCACAAAUGUCUAGUUCUUTT-3′. The siRNA duplexes were transfected into HEK293 cells using RNAiMax transfection reagent (Invitrogen, Carlsbad, CA, USA). The scrambled sequence (5′-AATCGCATAGCGTATGCCGTT-3′) was used as a control and did not correspond to any known gene in the databases.

Immunofluorescence and telomere FISH staining

Cells grown on glass coverslips were fixed with 2% paraformaldehyde at room temperature for 10 min and permeabilized with 0.5% Triton X-100 in PBS for 20 min. Cells were then blocked in PBS containing 2% bovine serum albumin and incubated with rabbit anti-53BP1 (Santa Cruz Biotechnology, Santa Cruz, CA, USA) overnight at 4°C. After thorough washing with PBS, cells were incubated with Alexa Fluor 488 goat anti-rabbit immunoglobulin (Molecular Probes). Telomere fluorescent in situ hybridization (FISH) staining was performed with Cy3-(CCCTAA)₃ peptide nucleic acid (PNA) probe (Panagene, Seoul, Korea). DNA was stained with 4,6-diamidino-2-phenylindole (DAPI) (Vectashield; Vector Laboratories, Burlingame, CA, USA). Immunofluorescence images were captured using a confocal laser-scanning microscope (Carl Zeiss, Jena, Germany). For telomere FISH on metaphase spreads, cells were treated with demecolcine (10 μg/ml) for 6 h, and metaphase chromosomes were prepared from growing cell cultures by standard methods (42). Telomere FISH with a Cy3-(CCCTAA)₃ PNA probe was performed using telomere PNA FISH kit (Dako, Glostrup, Denmark).

Senescence-associated β-galactosidase assay

HT1080 cell lines stably expressing the empty vector, TRF2 and K293R were plated in side culture chambers at the indicated population doublings (PDs) and washed with PBS, followed by fixation with 2% formaldehyde/0.2% glutaraldehyde for 5 min. Cells were washed with PBS twice and then incubated for 12 h with senescence-associated β-galactosidase (SA-β-Gal) staining solution (1 mg/ml X-gal, 40 mM citric acid/sodium phosphate, pH 6, 5 mM potassium ferrocyanide, 5 mM ferricyanide, 150 mM NaCl and 2 mM MgCl₂). For quantitation of senescent cells, a total of 200 cells were counted in each field, and four fields were examined.

Fluorescence-activated cell sorter analysis

HT1080 cell lines stably expressing the empty vector, TRF2 and K293R were washed with PBS and fixed for 30 min in ice-cold 70% ethanol. Cells were resuspended in PBS containing RNase A (200 μg/ml) and propidium iodide (50 μg/ml), and incubated in dark for 15 min at room temperature. Cell cycle distribution was examined by flow cytometry using a FACScan flow cytometer (BD Biosciences, San Jose, CA, USA).

Irradiation of cells

Cells were irradiated with 20 Gy using a 160-kV Faxitron X-ray machine (0.5-mm Cu filter; diameter, 33 cm; dose rate, 64.2 cGy/min) on ice. Irradiated cells were harvested after 30 min incubation at 37°C. Non-irradiated controls were incubated for corresponding times on ice.

RESULTS

TRF2 interacts with and is acetylated by p300

Post-translational modifications of TRF2 such as phosphorylation (32), ubiquitination (34), SUMOylation (35), methylation (37) and poly(ADP-ribosyl)ation (38,39) contribute to the maintenance of functional telomeres. In this work, we wanted to investigate whether TRF2 function is controlled by acetylation in mammalian cells. To determine if endogenous TRF2 is acetylated in vivo, HEK293 cells were subjected to immunoprecipitation by anti-TRF2 antibody, followed by immunoblotting with an antibody specific for acetylated lysine (Figure 1A). We were unable to detect acetylation of endogenous TRF2 in the empty vector-expressing cells. To identify a potential acetyltransferase that acetylates TRF2, the two human acetyltransferases, p300 and PCAF, were overexpressed in HEK293 cells. Although both acetyltransferases were expressed to similar levels, endogenous TRF2 was found to be strongly acetylated in p300-expressing cells, but not in PCAF-expressing cells (Figure 1A). A control IgG antibody failed to precipitate TRF2 and confirmed the specificity of TRF2 antibody. To confirm that TRF2 is acetylated by p300, HEK293 cells were co-transfected with Flag-TRF2 and p300-HA and subjected to immunoprecipitation. Flag-TRF2 was acetylated when p300-HA was expressed (Figure 1B). In contrast, deletion of the histone acetyltransferase (HAT) domain (p300ΔHAT) failed to acetylate TRF2. p53, which was known to be a substrate of p300, was used as a positive control (43) (Figure 1B). We also confirmed that expressed TRF2 was not acetylated by PCAF (Supplementary Figure S1A). These results indicate that TRF2 is acetylated by p300 in vivo, and this acetylation might play a role in the control of TRF2 function.

Figure 1.

p300 acetylates TRF2 in vivo and the two proteins form a stable complex. (A) Endogenous TRF2 is acetylated by p300. HEK293 cells were transfected with p300-HA or Flag-PCAF and subjected to immunoprecipitation with anti-TRF2 or normal IgG antibodies, followed by immunoblotting with anti-acetylated lysine antibody (Ac-Lys). (B) HEK293 cells were co-transfected with Flag-TRF2 or Flag-p53 and either p300-HA or p300ΔHAT-HA and subjected to immunoprecipitation with anti-Flag antibody, followed by immunoblotting with anti-acetylated lysine antibody. The arrowheads mark the positions of p300-HA and p300ΔHAT-HA. (C) HEK293 cells were co-transfected with Flag-TRF2 or Flag-p53 and either p300-HA or p300ΔHAT-HA and subjected to immunoprecipitation with anti-HA antibody, followed by immunoblotting with anti-Flag antibody. (D) HEK293 cells were transfected with Flag-TRF2 or Flag-TRF1 and subjected to immunoprecipitation with anti-TRF2, anti-TRF1 or anti-Flag antibodies, followed by immunoblotting with anti-p300 antibody. (E) HEK293 cells were subjected to immunoprecipitation with either anti-TRF2 or anti-TRF1 antibodies, followed by immunoblotting with anti-p300 antibody. IgG antibody was used as a negative control. (F) HEK293 cells were analysed by indirect immunofluorescence for co-localization of TRF2 with p300. Immunofluorescence was used to detect endogenous TRF2 (red) and p300 (green). DNA was stained with DAPI (blue).

To determine whether TRF2 and p300 associate in vivo, HEK293 cells were co-transfected with Flag-TRF2 and p300-HA and subjected to immunoprecipitation with anti-HA antibody, followed by immunoblotting with anti-Flag antibody. Flag-TRF2 was detected in anti-HA immunoprecipitates when p300-HA was expressed (Figure 1C). Deletion of the HAT domain (p300ΔHAT) did not impair the interaction between TRF2 and p300 (Figure 1C). p53, which was known to interact with p300, was used as a positive control (44). In contrast, no binding of TRF2 to PCAF was observed (Supplementary Figure S1B). To further verify the binding ability of TRF2 to p300, HEK293 cells were transfected with Flag-TRF2 or Flag-TRF1 and subjected to immunoprecipitation. Endogenous p300 was detected in anti-TRF2 and anti-Flag immunoprecipitates when Flag-TRF2 was expressed (Figure 1D). Endogenous p300 was also recovered in anti-TRF1 and anti-Flag immunoprecipitates when Flag-TRF1 was expressed. Furthermore, endogenous p300 was immunoprecipitated by endogenous TRF1 and TRF2 in HEK293 cells (Figure 1E), indicating that p300 interacts with both TRF1 and TRF2 in mammalian cells. We next determined the subcellular localization of TRF2 and p300 by indirect immunofluorescence. Endogenous p300 clearly localized to the nucleus, and about 28% of the p300 foci co-stained with the endogenous TRF2 foci (Figure 1F). p300 was also found in many foci that did not co-localize with TRF2 signals. Although the nature of these localization sites was not determined, our data demonstrate that TRF2 and p300 co-immunoprecipitate and co-localize in mammalian cells.

p300 interacts with the GAR domain of TRF2

To map the region in TRF2 that is responsible for p300 binding, we generated deletion constructs lacking the glycine and arginine rich (GAR) and Myb domains (Figure 2A). When HEK293 cells were co-transfected with p300-HA and various TRF2-Myc fragments, p300-HA was immunoprecipitated only by TRF2 fragments containing the GAR domain (Figure 2B). To confirm the binding ability of the GAR domain to p300, a series of TRF2 fragments were fused to GST and used in the in vitro binding assay (Figure 2C). GST-GAR bound efficiently to p300, whereas GST-Hinge had substantially reduced binding activity (Figure 2D). In contrast, GST-TRFH (TRF homology) and GST-Myb had no detectable binding activity. These results suggest that p300 may bind TRF2 through the two distinct regions, but the GAR domain has higher affinity for p300 binding than the Hinge domain.

Figure 2.

Identification of the interaction domain of TRF2 with p300. (A) Schematic representation of TRF2 and its deletion variants used in this study. (B) HEK293 cells were co-transfected with p300-HA and various TRF2-Myc and subjected to immunoprecipitation with anti-HA antibody, followed by immunoblotting with anti-Myc antibody. The TRF2-Myc proteins were detected by immunoblotting with anti-Myc antibody. (C) Schematic representation of TRF2 and its deletion variants (GAR, TRFH, Hinge and Myb). (D) The various GST-TRF2 proteins were affinity-purified and incubated with lysates prepared from cells expressing p300-HA, followed by detecting p300-HA. The purified GST fusion proteins were visualized by Coomassie staining. The arrowhead indicates the degradation product of GST-Hinge. Molecular mass makers are shown in kilodaltons.

Identification of the acetylation site on TRF2

To identify the acetylation site on TRF2, HEK293 cells were co-transfected with p300-HA and TRF2 deletion constructs and analysed the acetylation domain by immunoprecipitation with anti-acetylated lysine antibody. All deletion constructs were efficiently acetylated by p300 (Figure 3A), suggesting that the acetylation site could be contained in the TRFH and/or Hinge domains. To determine which of the two domains is acetylated, we performed an in vitro acetylation assay using various GST-TRF2 proteins. As shown in Figure 3B, p300 was able to acetylate the Hinge domain but not the GAR, TRFH and Myb domains. When the GST-TRF2 proteins were subjected to the in vitro acetylation assay in the presence of [¹⁴C] acetyl coenzyme A, the identical results were obtained (Figure 3C). These results suggest that, although the Hinge domain has lower affinity for p300 binding than the GAR domain (see Figure 2D), the interaction between p300 and Hinge is sufficient to acetylate TRF2. However, we do not exclude the possibility that p300 acetylates the Hinge domain of TRF2 through the interaction with the GAR domain.

Figure 3.

Identification of the acetylation site on TRF2. (A) HEK293 cells were co-transfected with p300-HA and various TRF2-Myc and subjected to immunoprecipitation with anti-Myc antibody, followed by immunoblotting with anti-acetylated lysine antibody. The various TRF2-Myc proteins were detected by immunoblotting with anti-Myc antibody. (B) The in vitro TRF2 acetylation assay was carried out using the various GST-TRF2 proteins. The acetylated proteins were detected by immunoblotting with anti-acetylated lysine antibody. The purified GST fusion proteins were visualized by Coomassie staining. Molecular mass makers are shown in kilodaltons. The arrowheads indicate the degradation product of GST-Hinge. (C) The various GST-TRF2 proteins were subjected to the in vitro acetylation assay using GST-p300 in the presence of [¹⁴C] acetyl coenzyme A. The acetylated proteins were detected by autoradiography. The purified GST fusion proteins were visualized by Coomassie staining. Molecular mass makers are shown in kilodaltons. The arrowheads indicate the degradation product of GST-Hinge. (D) MS/MS spectra corresponding to the acetylated peptide (AAFKacTLSGAQDSEAAFAK) from TRF2. HEK293 cells were co-transfected with Flag-TRF2 and p300-HA and subjected to immunoprecipitation with anti-Flag antibody. Immunoprecipitated TRF2 proteins were recovered from the gel, and analysed by LC-MS/MS. In gel, digestion was performed as described previously (45). The peak at 126.09 m/z (immonium ion) is due to the presence of the acetylated lysine residue. (E) HEK293 cells were co-transfected with p300-HA and various point mutant constructs of Flag-TRF2 and subjected to immunoprecipitation with anti-Flag antibody, followed by immunoblotting with anti-acetylated lysine antibody. (F) HEK293 cells were co-transfected with p300-HA and either Flag-TRF2 or Flag-K293R and subjected to immunoprecipitation with anti-HA or anti-Flag antibodies, followed by immunoblotting with anti-Flag or anti-HA antibodies.

There are 19 lysines within the Hinge domain in human TRF2. To determine the exact lysine residue on TRF2, HEK293 cells were co-transfected with Flag-TRF2 and p300-HA and subjected to immunoprecipitation with anti-Flag antibody, followed by SDS-PAGE and immunoblotting. The bands containing Flag-TRF2 were isolated and analysed by LC-MS/MS. From mass spectrometry analysis, lysine 293 was identified as the site of acetylation (Figure 3D; Supplementary Figure S2). To verify that lysine 293 is the acetylation site in vivo, lysine-to-arginine substitutions were constructed in lysine clusters within the Hinge domain and their effects on the overall levels of TRF2 acetylation were tested. The substitution of lysine 293 to arginine (K293R) completely abolished acetylation of TRF2, whereas none of the substitution mutants at other lysine residues had significant impact on acetylation (Figure 3E; Supplementary Figure S3). We also performed an in vitro acetylation assay with TRF2 and K293R. GST-TRF2, but not GST-K293R, was acetylated by overexpression of p300 (compare the first two lanes in Figure 3B), further demonstrating that lysine 293 serves as a major acetylation site of p300. We also found that the K293R mutation did not impair the ability of TRF2 to bind p300 (Figure 3F).

p300 regulates the stability of TRF2

To investigate the functional significance of p300-mediated TRF2 acetylation, we examined whether the amount of p300 affects the cellular abundance of TRF2. We transiently expressed p300-HA or p300ΔHAT-HA in HEK293 cells and examined the levels of endogenous TRF2. Whereas endogenous TRF2 was readily detectable in the empty vector-expressing cells, the levels of TRF2 was substantially increased in p300-expressing cells but was not altered in p300ΔHAT-expressing cells (Figure 4A). We also found that endogenous TRF1 was increased by overexpression of p300. In conjunction with the results of Figure 1D and E, these data demonstrate that p300 can interact with and stabilize both TRF1 and TRF2. However, the levels of other shelterin core proteins such as TIN2, POT1 and RAP1 were not influenced by overexpression of p300 (Figure 4A). To further confirm the role of p300 in a more physiological setting, the expression of endogenous p300 was depleted using siRNA duplex. p300 knockdown resulted in a clear reduction in the levels of endogenous TRF2 (Figure 4B). p300 knockdown also decreased the levels of endogenous TRF1 but did not affect the levels of other shelterin core proteins. Because p300 is a well known transcriptional coactivator (46), we examined whether reduction in the levels of both TRF1 and TRF2 is due to a p300 knockdown-related decrease in the transcription levels. The impact of p300 knockdown on gene expression of TRF1 and TRF2 was evaluated using reverse transcription–PCR analysis. No significant differences in the steady-state levels of TRF1 and TRF2 mRNAs were observed in cells expressing p300 siRNA and control siRNA (Supplementary Figure S4).

Figure 4.

p300 regulates the levels of endogenous TRF2. (A) HEK293 cells expressing p300-HA or p300ΔHAT-HA were subjected to immunoblotting as indicated. (B) HEK293 cells expressing p300 siRNA (sip300) or scrambled control siRNA (siControl) were subjected to immunoblotting as indicated. (C) HEK293 cells expressing p300-HA or p300 siRNA were treated with 100 μg/ml cycloheximide and together with, or without, 10 μM MG132 for the indicated times, followed by immunoblotting with anti-TRF2 or anti-actin antibodies. (D) Graphical representation of the relative TRF2 levels normalized against the β-actin loading control. The TRF2 expression levels were quantified with the average and standard deviation from three independent experiments. (E) HEK293 cells were co-transfected with Flag-TRF2 or Flag-K293R, TIN2-V5 and p300-HA, and subjected to immunoprecipitation with anti-V5 or anti-Flag antibodies, followed by immunoblotting with anti-Flag or anti-V5 antibodies. The p300 expression was detected with anti-p300 or anti-HA antibodies.

We further verified that the amount of p300 is involved in regulating the half-life of endogenous TRF2. The stability of TRF2 was monitored by immunoblotting with anti-TRF2 antibody in cells expressing p300-HA and p300 siRNA after cycloheximide treatment to inhibit new protein synthesis. Overexpression of p300-HA significantly extended the half-life of TRF2 compared with the empty vector-expressing cells (Figure 4C and D). In contrast, p300 knockdown reduced the half-life of TRF2. In all cell lines, turnover of TRF2 was blocked by the MG132 treatment, demonstrating that degradation of TRF2 is mediated by the proteasome.

It has been previously reported that TIN2 binds TRF1 and TRF2 simultaneously and stabilizes the TRF2 complex on telomeres (47,48). Because p300 interacts with both TRF1 and TRF2 and regulates their protein levels, we examined whether p300-mediated acetylation affects TRF2 interaction with TIN2. HEK293 cells were co-transfected with Flag-TRF2 or Flag-K293R, TIN2-V5 and p300-HA, and subjected to immunoprecipitation (Figure 4E). The amounts of Flag-TRF2 recovered in anti-V5 immunoprecipitates were not influenced by overexpression of p300-HA. Conversely, the amounts of TIN2-V5 detected in anti-Flag immunoprecipitates were not changed by overexpression of p300-HA. These results suggest that p300-mediated acetylation of TRF2 does not affect the interaction between TRF2 and TIN2.

Acetylation inhibits ubiquitin-dependent proteolysis of TRF2

Acetylation often interferes with ubiquitin-dependent proteolysis by competing for the same lysine residues, and competition between these two modifications might influence the stability of the substrates by altering protein structure (49). Because TRF2 has been shown to be ubiquitinated (34), p300-mediated acetylation might serve as a molecular switch to regulate ubiquitin-dependent proteolysis of TRF2. To investigate this possibility, HEK293 cells were transfected with Flag-TRF2 or Flag-K293R and treated with cycloheximide. Their protein levels were assessed in the absence or presence of MG132. The results indicate that K293R was present at significantly lower levels than wild-type TRF2 when de novo synthesis of proteins was inhibited (Figure 5A and B). The levels of both proteins were rescued when transfected cells were treated with MG132.

Figure 5.

Acetylation inhibits ubiquitin-dependent proteolysis of TRF2. (A) HEK293 cells expressing Flag-TRF2 or Flag-K293R were treated with 100 μg/ml cycloheximide and together with, or without, 10 μM MG132 for the indicated times, followed by immunoblotting with anti-Flag or anti-actin antibodies. (B) Graphical representation of the relative TRF2 levels normalized against the β-actin loading control. The TRF2 expression levels were quantified with the average and standard deviation from three independent experiments. (C) HEK293 cells were co-transfected with HA-ubiquitin and either Flag-TRF2 or Flag-K293R and treated with, or without, 10 μM MG132 for 2 h as specified. Immunoprecipitation was carried out with anti-Flag antibody before probing with anti-HA antibody. (D) ChIP of telomeric DNA by ectopically expressed TRF2. HEK293 cells were co-transfected with Flag-TRF2 or Flag-K293R and together with p300-HA or p300 siRNA and subjected to ChIP with anti-Flag antibody. Total DNA and immunoprecipitated DNA were applied to nitrocellulose in a slot blot manifold. Duplicate slot blots were hybridized with a telomeric probe or an Alu probe. Input DNA represents 10% of total DNA. HEK293 cells were co-transfected with Flag-TRF2 or Flag-K293R and together with p300-HA or p300 siRNA and subjected to immunoprecipitation with anti-Flag antibody, followed by immunoblotting with anti-acetylated lysine antibody. The p300 expression was detected with anti-p300 or anti-HA antibodies. (E) Quantification of blots shown in D. Histogram values represent the expressed TRF2 telomeric ChIP signal normalized to input signal. The percentage of telomeric DNA in ChIP is shown with the average and standard deviation from three independent experiments. (F) ChIP of telomeric DNA by endogenous TRF2. HEK293 cells were transfected with p300-HA or p300 siRNA and subjected to ChIP with anti-TRF2 antibody. HEK293 cells expressing p300-HA or p300 siRNA were subjected to immunoprecipitation with anti-TRF2 antibody, followed by immunoblotting with anti-acetylated lysine antibody. The endogenous TRF2 expression was detected with anti-TRF2 antibody. (G) Quantification of blots shown in F. Histogram values represent the endogenous TRF2 telomeric ChIP signal normalized to input signal. The percentage of telomeric DNA in ChIP is shown with the average and standard deviation from three independent experiments.

To examine whether TRF2 is ubiquitinated before proteasome-dependent degradation, HEK293 cells were co-transfected with HA-ubiquitin and either Flag-TRF2 or Flag-K293R. To illuminate ubiquitin-modified proteins, anti-Flag immunoprecipitates were evaluated by immunoblotting with anti-HA antibody. Ubiquitination of K293R was detectable even in the absence of MG132 and consistently higher than that of wild-type TRF2 (Figure 5C). This modification was further enhanced when cells were treated with MG132, suggesting that acetylation of lysine 293 increases the stability of TRF2 by inhibiting its ubiquitin-dependent proteolysis.

Acetylation is required for efficient telomere binding of TRF2

To determine whether p300-mediated acetylation affects telomere-binding activity of TRF2, we performed ChIP experiments. Cross-linked extracts prepared from cells expressing Flag-TRF2 or Flag-K293R were immunoprecipitated with anti-Flag antibody. The percentage of telomeric DNA bound to TRF2 was determined in slot blots using a labeled telomeric probe. When Flag-TRF2 was transfected, anti-Flag antibody immunoprecipitated ∼8.5% of total telomeric DNA in the empty vector-expressing cells (Figure 5D and E). However, telomeric DNA immunoprecipitated by anti-Flag antibody was significantly increased in p300-expressing cells. In marked contrast, the amounts of telomeric DNA bound to TRF2 were greatly reduced in p300 knockdown cells compared with control siRNA cells. When Flag-K293R was transfected, the amounts of telomeric DNA immunoprecipitated by anti-Flag antibody did not get significantly altered by overexpression or knockdown of p300 (Figure 5D and E). We also confirmed that the defect of K293R in associating with telomeric DNA is caused by the lack of acetylation at lysine 293 but not due to its reduced protein stability (Figure 5D).

To determine whether p300-mediated acetylation controls telomere binding of endogenous TRF2, HEK293 cells expressing p300-HA or p300 siRNA were subjected to ChIP with anti-TRF2 antibody. The amounts of telomeric DNA bound to endogenous TRF2 were increased by p300 overexpression but reduced by p300 knockdown (Figure 5F and G). Consistent with the results of Figure 4A and B, the expression levels of endogenous TRF2 were increased by p300 overexpression but decreased by p300 knockdown (Figure 5F). Taken together, these results suggest that acetylation at lysine 293 is required for efficient telomere binding of both endogenous and ectopically expressed TRF2.

Depletion of p300 induces DNA-damage foci at telomeres

Because p300-mediated acetylation is required for efficient telomere binding of TRF2, depletion of p300 may result in telomere dysfunction owing to the lack of TRF2 function. Dysfunctional telomeres resemble damaged DNA and are recognized by the canonical DNA-damage signaling pathway (22,24). The resulting telomere dysfunction-induced foci (TIFs) represent the foci of DNA-damage response factors that coincide with telomeres (21). To determine the role of p300 in telomere-damage pathway, the telomeric foci for 53BP1 were examined in p300 knockdown cells. Depletion of p300 induced numerous 53BP1 foci in the nucleus, whereas these events were rare in the empty vector-expressing cells (Figure 6A). Approximately 70% of p300 knockdown cells contained more than 10 DNA-damage foci per nucleus (Figure 6B). Many of these foci co-localized with telomeres as indicated by dual staining with telomeric TTAGGG-specific FISH probe (Figure 6A). When the TIF response was quantified in 200 nuclei showing ≥10 DNA-damage foci, ∼58% of p300 knockdown cells contained more than four foci at telomeres and were scored as TIF positive (Figure 6C). Depletion of p300 also resulted in 53BP1 foci that were not obviously associated with telomeres, suggesting that p300 is involved in the control of a general DNA-damage response.

Figure 6.

Depletion of p300 induces a DNA-damage signal at telomeres. (A) HT1080 cells expressing p300 siRNA or scrambled control siRNA were analysed by indirect immunofluorescence for co-localization of 53BP1 foci (green) with telomeric sites marked by TTAGGG-specific FISH probe (red). DNA was stained with DAPI (blue). A subset of 53BP1 foci co-localized with TTAGGG probe is indicated by arrows. (B) Quantification of the induction of 53BP1 foci by p300 knockdown. The average percentage of cells with either 6–10 or >10 53BP1 foci is shown. For each condition, at least 200 nuclei were counted. (C) Quantification of the induction of TIFs by p300 knockdown. The average percentage of cells with either 4–6 or >6 TIFs is shown.

Overexpression of K293R induces DNA-damage foci at telomeres

We next investigated the effect of K293R mutation on telomere-damage response. To examine the telomeric foci for 53BP1, we established HT1080 cell lines stably expressing TRF2 or K293R and monitored the PD levels at regular intervals. Multiple independent clones were isolated to rule out the effect of clonal variation. The levels of endogenous and ectopically expressed TRF2 proteins were examined by immunoblotting with anti-TRF2 antibody. The ectopic expression of TRF2 was ∼15-fold higher than the endogenous gene (Figure 7A). TRF2 and K293R proteins were almost equally expressed in HT1080 cell lines throughout the duration of the experiments. The basal levels of 53BP1 foci were observed in the empty vector-expressing cells (Figure 7B). Although 53BP1 function has been shown to be constitutively activated in several tumor cell lines (50), these foci rarely co-localized with telomeres. Overexpression of TRF2 did not affect the basal levels of 53BP1 foci up to PD 52 after antibiotic selection. By contrast, overexpression of K293R induced numerous 53BP1 foci in the nucleus when cells were examined at PD 35 (Figure 7B), resulting in ∼53% cells showing ≥10 DNA-damage foci per nucleus (Figure 7C). When the TIF response was quantified in 200 nuclei showing ≥10 DNA-damage foci, ∼60% of K293R-expressing cells contained more than four foci at telomeres (Figure 7D). Overexpression of K293R also resulted in a massive increase in the number of 53BP1 foci that were not obviously associated with telomeres, demonstrating that K293R has additional function in nontelomeric DNA-damage. Whereas the number of 53BP1 foci greatly increased following K293R overexpression, there is a statistically significant re-localization of those foci to telomeres (Figure 7E), indicating that K293R preferentially affects telomeres.

Figure 7.

Overexpression of K293R induces a DNA-damage signal at telomeres. (A) HT1080 cells were transfected with TRF2-Myc or K293R-Myc, and multiple independent stable clones were isolated, followed by immunoblotting with anti-TRF2 or anti-actin antibodies. (B) HT1080 cells stably expressing TRF2-Myc or K293R-Myc were analysed by indirect immunofluorescence for co-localization of 53BP1 foci (green) with telomeric sites marked by TTAGGG-specific FISH probe (red). Representative fluorescence images of nuclei showing a large number of 53BP1 foci are shown as indicated. DNA was stained with DAPI (blue). A subset of 53BP1 foci co-localized with TTAGGG probe is indicated by arrows. (C) Quantification of the induction of 53BP1 foci by K293R overexpression. The average percentage of cells with either 6–10 or >10 53BP1 foci is shown. For each condition, at least 200 cells were counted. (D) Quantification of the induction of TIFs by K293R overexpression. The average percentage of cells with either 4–6 or >6 TIFs is shown. Cells with four or more DNA-damage foci co-localized with TTAGGG probe were scored as TIF positive. (E) The average percentage of 53BP1 foci located at telomeres was determined in TIF-positive nuclei.

Because the TIFs are induced in response to K293R overexpression, we next evaluated the status of telomeric DNA by telomere FISH. The analysis of metaphase spreads derived from K293R-expressing cells did not show significant levels of telomere aberrations such as end-to-end chromosome fusions, double telomeric signals on single chromatids and extrachromosomal telomeric signals. However, we observed a small but significant increase of chromatid ends without telomeric FISH signal (Figure 8A). Although the nature and origin of these aberrant telomere structures have not been established, they might be formed due to telomere uncapping induced by K293R overexpression. No such telomere aberrations were observed in HT1080 cells expressing the empty vector and TRF2 (Figure 8B).

Figure 8.

Overexpression of K293R increases the occurrence of telomeric signal-free chromatid ends. (A) Representative telomere FISH analysis on metaphase spreads for telomere defects. HT1080 cells stably expressing the empty vector TRF2-Myc or K293R-Myc were processed for telomeric FISH. Telomere signal-free chromatid ends are indicated by arrows. (B) Quantification of telomere signal-free chromatid ends in HT1080 cells stably expressing the empty vector, TRF2-Myc or K293R-Myc.

Overexpression of K293R limits cell proliferation

Because overexpression of K293R induces a DNA-damage response at telomeres, we wanted to determine the effect of K293R overexpression on growth suppression. HT1080 cell lines stably expressing TRF2 exhibited only minor difference in cell proliferation up to PD 52 after antibiotic selection compared with the empty vector-expressing cells (Figure 9A). They grew normally and continued to divide throughout the duration of the experiments. In contrast, the growth rates of two independent K293R-expressing cells gradually slowed down compared with TRF2-expressing cells and almost stopped dividing at ∼PD 39 (Figure 9A). Similar growth curves were obtained for DU145 prostate carcinoma cells (Figure 9B).

Figure 9.

Overexpression of K293R limits cell proliferation. (A) Cell growth curves of HT1080 cell lines stably expressing the empty vector, TRF2-Myc or K293R-Myc. Stable cells were replated every 3 days to maintain log-phase growth and calculate the growth rate, with day 0 representing the first day after blasticidin selection. (B) Cell growth curves of DU145 cell lines stably expressing the empty vector, TRF2-Myc or K293R-Myc. (C) HT1080 cell lines stably expressing the empty vector, TRF2-Myc or K293R-Myc, at the indicated PDs were photographed after staining for SA-β-Gal activity. (D) The percentage of total cells that are positive for SA-β-Gal activity is shown in four HT1080 cell lines. For quantitation of senescent cells, a total of 200 cells were counted in each field and four fields were examined. (E) Flow cytometric analysis of HT1080 cell lines stably expressing the empty vector, TRF2-Myc or K293R-Myc. Cell lines at the indicated PDs were stained with propidium iodide, followed by fluorescence-activated cell sorter analysis. The percentage of total cells in each phase of the cell cycle is shown. Results are representative of three separate experiments. (F) The percentage of total cells in the aneuploidy fraction is shown with the average and standard deviation from three experiments being present in E.

We next examined whether the growth arrest of K293R-expressing cells at later PDs was accompanied by cellular senescence. About 62% of K293R-expressing cells were stained intensely for SA-β-Gal activity and displayed a flattened and enlarged morphology (Figure 9C and D). However, no such changes were found in cells expressing the empty vector and TRF2 throughout the duration of the experiments. To determine whether the growth arrest correlates with an altered cell cycle distribution, TRF2- and K293R-expressing cells were subjected to flow cytometric analysis by propidium iodide staining. K293R-expressing cells exhibited a substantial decrease in G₁ and S fractions and a concomitant increase in G₂/M fraction compared with cells expressing the empty vector and TRF2 (Figure 9E). In addition, K293R-expressing cells showed increased DNA content (Figure 9F). These results suggest that p300-mediated acetylation of TRF2 plays an important role in the control of cell proliferation and cell cycle by maintaining functional telomeres.

TRF2 has been shown to be phosphorylated in response to DNA damage by an ATM-dependent pathway (32,33). Therefore, we determined whether DNA damage induced by K293R overexpression is dependent on the ATM kinase. Phosphorylation of ATM on serine 1981, an essential step in the activation of this kinase, was detected after ionizing radiation (Supplementary Figure S5) (51). In contrast, ATM was not phosphorylated in cells expressing TRF2 and the empty vector when DNA damage was not induced. However, autophosphorylation of ATM was detected in K293R-expressing cells. Furthermore, DNA-damage–induced phosphorylation of TRF2 at threonine 188 was detected in K293R-overexpressing cells but not in cells expressing TRF2 and the empty vector (Supplementary Figure S5) (32,33). These results suggest that the observed growth arrest and cellular senescence in K293R-expressing cells directly result from the telomeric damage induced by the TRF2 mutation.

DISCUSSION

Here we provide compelling evidence that TRF2 interacts with and is acetylated by p300. Identification of p300 as a TRF2-interacting partner raises the possibility that the function of human telomeres is regulated by TRF2 acetylation. The results presented in this work demonstrate that p300-mediated acetylation stabilizes the TRF2 protein by inhibiting ubiquitin-dependent proteolysis and is required for efficient telomere binding of TRF2. We also show that overexpression of the acetylation-deficient mutant, K293R, induces DNA-damage foci at telomeres, thereby resulting in impaired cell growth and cellular senescence. On the basis of our collective results, we propose that TRF2 acetylation is required to maintain functional telomeres and regulate the telomere-associated DNA-damage response.

TRF2 plays an important role in protecting chromosome ends by facilitating their organization into the t-loop structure (19,20). Loss or functional inhibition of TRF2 induces telomere dysfunction, leading to cellular senescence through the ATM- and p53-mediated pathway (52–54). These findings suggest that the abundance of TRF2 at telomeres should be maintained and tightly regulated to ensure proper function of telomeres. We show that overexpression of p300 increases the half-life of endogenous TRF2, whereas depletion of p300 results in a decrease of TRF2 expression. In addition, acetylation enhances the ability of TRF2 to bind telomeric DNA. These observations suggest that p300-mediated acetylation is required to maintain the levels of TRF2 at telomeres. Under conditions in which p300 concentration is kept below the threshold level, TRF2 may exist as an unacetylated form, which in turn leads to a decrease in its telomere binding activity. The resulting telomere-unbound form of TRF2 could be rapidly ubiquitinated. Consistent with this idea, we found that the K293R mutation reduced the half-life and telomere-binding activity of TRF2. These results support the idea that sequential post-translational modifications, including acetylation and ubiquitination, play important roles in determining the abundance of TRF2 at telomeres. Thus, p300-mediated acetylation would serve as a molecular switch for conversion of TRF2 from degradation to telomere binding. Interestingly, lysine 293 was identified as an ubiquitination site in a proteome-wide study (55). Therefore, it is likely that acetylation at lysine 293 interferes with ubiquitin-dependent proteolysis by competing for the same lysine residue.

When TRF2 is unacetylated (or deacetylated) by inhibition of p300 (or by activation of deacetylase), TRF2 is dissociated from telomeres and subsequently degraded through ubiquitin-dependent proteolysis. Ubiquitin-mediated degradation of TRF2 was efficiently inhibited by the MG132 treatment, suggesting that proteasomal degradation of TRF2 occurs in a separate step from its dissociation from telomeres. TRF2 has been shown to be ubiquitinated by p53-inducible E3 ubiquitin ligase Siah 1 (34). During replicative senescence, telomeric damage signaling from uncapped telomeres induces activation of p53, which in turn reduces the levels of TRF2 at telomeres through Siah 1-mediated ubiquitination. On the basis of these findings, it is possible that Siah 1 might preferentially act on the unacetylated and telomere-dissociated form of TRF2 to rapidly ubiquitinate this protein. Thus, degradation of unacetylated TRF2 might be coupled with Siah 1-mediated ubiquitination. We propose that telomeres, as in the case of TRF1 (56,57), seem to be assembled with acetylated TRF2 rather than the unacetylated telomere-unbound form to establish functional telomeres. Therefore, it will be of immediate interest to examine whether acetylation/deacetylation of TRF2 functions upstream of Siah 1-mediated ubiquitination and degradation.

The functional significance of TRF2 acetylation was investigated by overexpression of the acetylation-deficient mutant. Overexpression of K293R resulted in telomere dysfunction as indicated by an increase in 53BP1-associated telomere foci. In contrast, overexpression of wild-type TRF2 did not affect the basal levels of TIFs in HT1080 cells. The ability of TRF2 to participate in t-loop formation suggests that this protein is required for repression of a DNA-damage response at telomeres (19,25). Because the expression levels of wild-type and mutant TRF2 were almost identical, the TIF phenotype associated with K293R overexpression was not due to reduced expression of the protein. The K293R mutation resulted in a failure of TRF2 to prevent the formation of DNA-damage foci, suggesting that K293R overexpression interferes with the accumulation of endogenous TRF2 at telomeres. These results indicate that K293R functions as a dominant-negative for telomere function, and that p300-mediated acetylation at lysine 293 is essential for repression of a telomeric damage phenotype. In addition, K293R overexpression also resulted in many 53BP foci that were not obviously associated with telomeres. These findings suggest that TRF2 acetylation also has a nontelomeric function in DNA-damage signaling pathway, thereby contributing to global genome integrity.

Overexpression of K293R induced a growth arrest in the human fibrosarcoma and prostate carcinoma cells used in this study. This growth arrest was accompanied by several features consistent with the induction of cellular senescence, including a specific cellular morphology, expression of a SA-β-Gal activity and an altered cell cycle distribution (58,59). Thus, it is possible that this senescent-like phenotype in transformed cells is closely related to the replicative senescence previously described in primary human cells. The critical question that remains to be answered is how K293R overexpression induces the growth arrest. As noted above, unacetylated TRF2 has reduced telomere-binding activity and is rapidly degraded by ubiquitin-mediated proteolysis. As a result, the lack of TRF2 on telomeres might cause dysfunctional telomeres, which in turn induce DNA-damage foci at telomeres. Because the growth arrest in K293R-expressing cells is accompanied by induction of TIFs, this arrest may be a response of DNA damage arising from dysfunctional telomeres. Thus, the dominant-negative activity of K293R would be expected to induce a growth arrest signal by displacing the endogenous TRF2 function.

In conclusion, our data provide evidence for a previously unknown function of TRF2 acetylation. p300-mediated acetylation is essential for efficient telomere binding of TRF2 and functions in concert with ubiquitination in modulating TRF2 activity. It is likely that TRF2 acetylation is determined by the relative activities of acetyltransferase and deacetylases, both of whose activities are tightly regulated by diverse signaling pathways in different cellular environments. Although important questions about the physiological role of TRF2 acetylation and how this modification is regulated under diverse pathophysiological conditions remain to be resolved, our results suggest that p300-mediated acetylation of TRF2 represents a new route for regulation of telomere maintenance pathway.

SUPPLEMENTARY DATA

Supplementary Data are available at NAR Online: Supplementary Figures 1–5.

FUNDING

Funding for open access charge: World Class University Fund from the Korean Ministry of Education, Science, and Technology [R31-2009-000-10086-0 to I.K.C.].

Conflict of interest statement. None declared.