MCB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental material
Right arrow Other Versions of this Article:
MCB.00323-06v1
27/1/229    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mohan, R. D.
Right arrow Articles by Tini, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mohan, R. D.
Right arrow Articles by Tini, M.
Molecular and Cellular Biology, January 2007, p. 229-243, Vol. 27, No. 1
0270-7306/07/$08.00+0     doi:10.1128/MCB.00323-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

SUMO-1-Dependent Allosteric Regulation of Thymine DNA Glycosylase Alters Subnuclear Localization and CBP/p300 Recruitment{triangledown} ,{dagger}

Ryan D. Mohan,1,{ddagger} Anita Rao,1,{ddagger} Jason Gagliardi,2,{ddagger} and Marc Tini1,2*

Departments of Physiology and Pharmacology,1 Department of Microbiology and Immunology, Siebens-Drake Medical Research Institute, Schulich School of Medicine and Dentistry, University of Western Ontario, London, Ontario, Canada N6G 2V42

Received 21 February 2006/ Returned for modification 26 April 2006/ Accepted 10 October 2006


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Previous studies have demonstrated that the base excision repair enzyme thymine DNA glycosylase (TDG) mediates recruitment of histone acetyltransferases CREB-binding protein (CBP) and p300 to DNA, suggesting a plausible role for these factors in TDG-mediated repair. Furthermore, TDG was found to potentiate CBP/p300-dependent transcription and serve as a substrate for CBP/p300 acetylation. Here, we show that the small ubiquitin-like modifier 1 (SUMO-1) protein binding activity of TDG is essential for activation of CBP and localization to promyelocytic leukemia protein oncogenic domains (PODs). SUMO-1 binding is mediated by two distinct amino- and carboxy-terminal motifs (residues 144 to 148 and 319 to 322) that are negatively regulated by DNA binding via an amino-terminal hydrophilic region (residues 1 to 121). TDG is also posttranslationally modified by covalent conjugation of SUMO-1 (sumoylation) to lysine 341. Interestingly, we found that sumoylation of TDG blocks interaction with CBP and prevents TDG acetylation in vitro. Furthermore, sumoylation effectively abrogates intermolecular SUMO-1 binding and a sumoylation-deficient mutant accumulates in PODs, suggesting that sumoylation negatively regulates translocation to these nuclear structures. These findings suggest that TDG sumoylation promotes intramolecular interactions with amino- and carboxy-terminal SUMO-1 binding motifs that dramatically alter the biochemical properties and subcellular localization of TDG.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
In vertebrate genomes, methylation of cytosine within CpG dinucleotides constitutes an important mechanism regulating transcription and chromatin structure (35). CpG methylation also contributes to genome instability by promoting spontaneous hydrolytic deamination of methylated cytosines to generate thymine residues (27), which in the absence of DNA repair give rise to cytosine-to-thymine transition mutations believed to have a causative role in cancer (17). For example, these CpG mutations are the most prevalent genetic alterations in the p53 tumor suppressor gene detected in many human tumors (36). The incidence of CpG mutations is also dramatically increased in aging mouse tissues and, therefore, may contribute significantly to cellular aging (11).

Thymine DNA glycosylase (TDG) is one of two enzymes mediating the excision of mispaired thymine (G:T) and uracil (G:U) in the CpG context (23, 32, 33). TDG processes thymine, uracil, 5-hydroxymethyluracil, and 3,N4-ethenocytosine mispaired with guanine (18) to generate an abasic site that is subsequently repaired by other base excision repair (BER) enzymes (41). Interestingly, TDG has also been shown to interact with a number of transcription factors, including Jun and members of the nuclear receptor family, suggesting a link between transcription and BER (7, 8, 28, 46, 47).

Previous studies have revealed a functional association between TDG and transcriptional coactivators CREB-binding protein (CBP) and p300 (46). CBP/p300 are essential proteins that potentiate diverse transcription factor signaling pathways in part by mediating acetylation of chromatin and chromatin-associated proteins (16). Notably, CBP/p300-TDG complexes are recruited to DNA in vitro and have the potential to participate in both transcriptional regulation and DNA repair (46). Accordingly, TDG was shown to be both a potent activator of CBP/p300-dependent transcription and a substrate for CBP/p300 acetylation (46).

TDG is posttranslationally modified by covalent conjugation to SUMO (small ubiquitin-like modifier) proteins (SUMO-1, -2, and -3), resulting in inhibition of DNA binding and altered DNA repair kinetics (20). SUMO-1 is a 97-amino-acid peptide that is covalently attached to proteins at lysine residues (consensus {psi}KXE), thereby affecting subcellular localization and molecular interactions (22). Importantly, SUMO modification plays important roles in transcriptional regulation and maintenance of genomic integrity (22). Sumoylation, in some instances, promotes localization to nuclear compartments, known as promyelocytic leukemia protein (PML) oncogenic domains (PODs) (6, 26, 37). The dynamic association of transcription and DNA repair factors with PODs suggests that these nuclear structures play important roles in regulating gene expression and genome stability (26).

We have investigated the role of sumoylation and noncovalent SUMO-1 binding in the regulation of subcellular localization and biochemical properties of TDG. Our studies have mapped SUMO-1 binding activity to two separate SUMO binding motifs (SBMs) located in the amino- and carboxy-terminal regions. We show that both SBMs are essential for normal POD localization and activation of CBP-dependent transcription. Furthermore, the SBMs are regulated by DNA interactions mediated via an amino-terminal hydrophilic domain. Interestingly, we have shown that sumoylation of TDG promotes intramolecular interactions that dramatically alter the biochemical properties of TDG, thereby preventing association with CBP and POD translocation.


   MATERIALS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Plasmids. Plasmid constructs were verified by sequencing, and details are available on request. GAL-CBP and TDG were expressed in pCMX mammalian expression vectors (46). FLAG-tagged constructs lacking the amino-terminal region of TDG were fused to the simian virus 40 (SV40) nuclear localization signal (NLS) to replace the natural NLS contained within this region. Carboxy-terminal deletions of TDG were constructed by directional cloning of PCR-amplified fragments into the pCMX-FLAG vector. TDG and CBP point mutants were constructed using the QuikChange mutagenesis kit (Stratagene) according to the manufacturer's directions. Cyan fluorescent protein (CFP) and yellow fluorescent protein (YFP) fusions of TDG and PML were constructed using the pCMX-CFP or pCMX-YFP expression vectors. Renilla green fluorescent protein (GFP) fusion constructs were made using the phrGFP-N1 vector from Stratagene. Other expression vectors have been previously described (10, 31, 46).

Cell culture, transfections, and heat shock treatment. MCF-7 cells were maintained in Dulbecco's minimal essential medium containing penicillin-streptomycin and supplemented with 10% fetal bovine serum. Cells were seeded onto 24-well dishes and transfected using Lipofectamine 2000 transfection reagent (Invitrogen). Approximately 250 ng of luciferase-based reporter plasmid, 100 ng of Gal-CBP, and 100 to 500 ng of pCMX-based expression vectors were used per well. Transfection efficiency was normalized by cotransfection of Renilla luciferase reporter vector phRL-SV40 (Promega). Transfection experiments were performed at least three times in duplicate, and results are shown with standard errors. Heat shock treatments (42°C) were performed on MCF-7 cells seeded on six-well dishes transfected with expression vectors for TDG (200 ng) and PML (100 ng). At 0, 15, or 30 min, cells were lysed in 300 µl Laemmli buffer containing 3 units Benzonase (Novagen), and the modification state of TDG was analyzed by immunoblotting with a TDG-specific antibody.

Preparation of whole-cell extracts. MCF-7 whole-cell extracts for glutathione S-transferase (GST)-based interaction assays were prepared from 10-cm dishes of cells transfected with 7.5 µg of TDG expression vector. Cell pellets were resuspended in 500 µl of lysis buffer (50 mM Tris-HCl pH 7.9, 300 mM NaCl, 1 mM EDTA, 1 mM EGTA 10% glycerol, 0.5% Triton X-100, 1 mM dithiothreitol [DTT], proteinase inhibitors) and incubated on ice for 30 min. Subsequently, the cell lysate was diluted with 500 µl of dilution buffer (50 mM Tris-HCl pH 7.9, 1 mM EDTA, 1 mM EGTA, 10% glycerol, 1 mM DTT), and insoluble products were removed by centrifugation. Whole-cell extracts for the analysis of sumoylation were prepared from transfected MCF-7 cells lysed in Laemmli buffer containing Benzonase (Merck).

Protein purification and in vitro interaction assays. Protein purification and GST-based interaction assays using in vitro-translated and recombinant proteins have been previously described (46). For ethidium bromide treatments, in vitro-translated proteins were treated with 100 µM ethidium bromide for 20 min at 4°C prior to use in pull-down experiments. Binding reaction mixtures and washing buffers also contained ethidium bromide. For interaction assays performed in the presence of duplex oligonucleotides containing a G:T mispair, recombinant TDG was preincubated with increasing amounts of the oligonucleotides for 15 min at room temperature. Whole-cell extracts for pull-down experiments were precleared twice with 25 µl (packed bead volume) of glutathione-Sepharose beads (Amersham) for 30 min at 4°C. Total protein concentration of the precleared lysate was determined by bicinchoninic acid assay (Pierce), and the relative expression of transfected proteins was determined by immunoblotting with a TDG-specific antibody. Subsequently, the amount of expressed protein in each lysate used for the pull-down was equalized by addition of untransfected cellular lysates. Pull-downs were performed using 3 µg of GST-SUMO and bound proteins detected by immunoblotting with a TDG-specific antibody. FLAG-epitope-based interaction assays were performed with baculovirus-expressed FLAG-CBP (400 ng) and recombinant sumoylated (400 ng) or mock-sumoylated HIS-TDG (400 ng). Proteins were incubated with 10 µl packed commercial anti-FLAG affinity matrix (Sigma-Aldrich) in NETN buffer (50 mM Tris-HCl pH 7.9, 100 mM NaCl, 1 mM EDTA, 10% glycerol, 0.2% NP-40, 1 mM DTT) in a final volume of 150 µl for 1 h (4°C). The beads were subsequently washed with NETN buffer, and bound proteins were fractionated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and visualized by immunoblotting with an anti-HIS antibody. Recombinant GST-p53 (2.5 µg) was sumoylated as described below and bound to glutathione-Sepharose affinity matrix (25 µl packed volume). Beads were washed three times with NETN buffer, including one wash with NETN containing 500 mM NaCl and subsequently resuspended in 150 µl of NETN. Binding reactions were carried out with 40-µl aliquots as described above. A portion of the slurry was analyzed by immunoblotting with anti-p53 and anti-GMP-1 monoclonal antibodies. Interaction assays with baculovirus-expressed FLAG-CBP (1 µg) were carried out as described above, but bound complexes were immunoprecipitated with CBP polyclonal antibody.

Oligonucleotide cleavage assays. Cleavage assays were performed essentially as previously described (32). Approximately 25 ng of recombinant TDG or 5 µl of in vitro-translated TDG was incubated at 30°C with 5 ng of radiolabeled duplex oligonucleotide containing either a G:T or G:U mispair in 20 µl of cleavage buffer (25 mM HEPES-KOH [pH 7.8], 1 mM EDTA, 0.1 mg/ml bovine serum albumin, and 1 mM DTT). Reactions were carried out for 30 min for recombinant TDG and 2 h for in vitro-translated protein. Subsequently, the DNA was precipitated, resuspended in 100 mM NaOH, and incubated at 90°C for 30 min. The cleavage products were fractionated by denaturing polyacrylamide gel electrophoresis and visualized by autoradiography and phosphorimaging. Assays on sumoylated TDG were carried out using 10 ng of duplex oligonucleotide.

ABCD assays. Duplex oligonucleotides containing either no mispairs or a single G:T or G:U mispair were generated by annealing the following complementary oligonucleotides: 5'-[biotin]-TAGACATTGCCCTCGAGGTACCATGGATCCGATGTCGACCTCAAACCTAGACGAATTCCG-3' and 5'CGGAATTCGTCTAGGTTTGAGGT[C, T, or U]GACATCGGATCCATGGTACCTCGAGGGCAATGTCTA-3'). Approximately 500 ng annealed oligonucleotide was incubated for 30 min at room temperature with 10 µl of streptavidin-coated paramagnetic beads (MagneSphere; Promega) and 500 ng of purified bacterially expressed TDG in avidin-biotin complex DNA (ABCD) buffer (50 mM Tris-Cl [pH 7.9], 150 mM NaCl, 10% glycerol, 5 mM MgCl2, 0.1% NP-40, and 0.5 mM DTT). Total reaction volume was 50 µl. Beads were washed five times with 200 µl of ABCD buffer, and bound proteins were analyzed by immunoblotting. In some experiments, TDG was preincubated on ice with 2 µg GST-SUMO or GST for 30 min prior to analysis.

Protein acetylation assays. Sumoylated or mock-sumoylated TDG (400 ng) was incubated with approximately 100 ng of purified, full-length CBP in a total volume of 30 µl in acetylation buffer (20 mM HEPES [pH 7.8], 1 mM EDTA, 1 mM DTT, 10 mM sodium butyrate, and 10% glycerol) in the presence of 1.5 µM [14C]acetyl coenzyme A (AcCoA) and incubated for 30 min at 30°C followed by electrophoresis on an 8% sodium dodecyl sulfate-polyacrylamide gel. The gel was subsequently fixed with a 30% methanol, 10% acetic acid solution and treated with amplifying solution (Amersham) before exposure to film. Western blotting was performed to confirm equal loading of protein and the maintenance of SUMO modification of TDG.

In vitro sumoylation. Sumoylation was performed as previously described (10). Briefly, recombinant GST-SAE1, GST-SAE2, polyhistidine (His)-tagged UBC9 (1 µg each), and SUMO-1 (1.5 µg) proteins were incubated with 5 µg of His-TDG in SUMO conjugation buffer (20 mM HEPES [pH 7.4], 5 mM MgCl2, 1 mM creatine phosphate, 0.35 units/ml of creatine kinase [Roche], 1 mM ATP). Mock sumoylation reactions were performed in the absence of SUMO-1. Modified TDG was purified using Ni-nitrilotriacetic acid Superflow affinity resin (QIAGEN) and dialyzed against NETN at 4°C for 12 h. Copurified His-UBC9 was removed by centrifugal membrane separation using a 10-kDa molecular weight cutoff cellulose filter (Centricon).

Antibodies and immunostaining. TDG-specific antibody was raised in rabbits immunized with recombinant full-length TDG. Immunoglobulin G (IgG) was purified from immune sera by protein A chromatography. Human PML-specific monoclonal (PG-M3, sc-966), CBP-specific polyclonal (sc-369), and anti-p53 (DO-1, sc-126) monoclonal antibodies were obtained from Santa Cruz Biotechnology. Mouse PML-specific monoclonal antibody (05-718) was from Upstate/Chemicon. SUMO-1-specific monoclonal antibody was purchased from Zymed (clone 21C7). Anti-FLAG monoclonal antibody (M2) was obtained from Sigma-Aldrich. For immunostaining, cells were fixed with 4% formaldehyde for 15 min followed by a 10-min incubation with 0.1 M glycine in phosphate-buffered saline. Cells were then permeabilized with 0.5% Triton X-100. Alternatively, cells expressing FLAG-tagged proteins were fixed with methanol-acetone (1:1) for 1 minute at room temperature. Immunostaining was performed with the appropriate primary antibody and fluorophore-conjugated donkey secondary antibody (CY3 and fluorescein isothiocyanate [FITC]; Jackson ImmunoResearch Laboratories).

Microscopy. Epifluorescence imaging was performed on an Axiovert 200 M inverted microscope equipped with an Apotome (Carl Zeiss) using appropriate fluorophore-specific filter sets. Z-series images (x63 magnification) of 0.5-µm thickness were acquired and processed with Axiovision software and Adobe Photoshop.


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
SUMO-dependent translocation of TDG to PML oncogenic domains. Previous studies have demonstrated that TDG is acetylated by CBP/p300 and can act as a potent activator of CBP-dependent transcription (46). CBP is recruited to PODs by association with PML (12), and there is considerable evidence suggesting that these nuclear structures play important roles in transcription and DNA repair (9, 26, 50). We investigated the subnuclear localization of TDG in human breast carcinoma cells (MCF-7) by indirect immunofluorescence using TDG-specific antibodies and by transient expression of YFP-tagged TDG. YFP-TDG localized throughout the nucleoplasm with the exception of nucleoli (Fig. 1A, panel II); accentuated fluorescence was observed in nuclear PODs, as demonstrated by colocalization of YFP-TDG with PML (Fig. 1A, panel III). Furthermore, coexpression of YFP-TDG and PML dramatically increased POD localization of TDG (panels V to VIII). Since these observations suggested that TDG associates with PML, we determined whether a bacterially expressed GST-PML fusion protein bound TDG in whole-cell lysates derived from transfected MCF-7 cells. While we did not detect binding of TDG to GST-PML, binding to GST-SUMO-1 was readily observed (Fig. 1B). In light of the SUMO-1 binding properties of TDG, we investigated whether this activity is required for POD targeting. Mild hyperthermic stress causes rapid desumoylation of PML and another POD component, SP100, without affecting the structural integrity of the PODs (34). We subjected MCF-7 cells expressing YFP-TDG and CFP-tagged PML (CFP-PML) to heat shock at 42°C for 15 min and monitored protein localization in live cells by fluorescence microscopy. A dramatic loss of POD accumulation of YFP-TDG was observed without detectable changes in CFP-PML localization (Fig. 1C). Immunoblotting analysis of cell lysates, using a TDG-specific antibody, indicated that heat shock did not alter the levels of TDG sumoylation (Fig. 1C). In light of the reported desumoylation of PML and SP100 following hyperthermic stress (34), these observations suggest that the SUMO-1 binding activity of TDG mediates POD targeting.


Figure 1
View larger version (42K):
[in this window]
[in a new window]

  		FIG. 1. SUMO-1-dependent recruitment of TDG to PML oncogenic domains. (A) Subcellular localization of YFP-TDG with and without coexpression of PML. MCF-7 cells were transfected with 300 ng of YFP-TDG expression vector alone or in combination with 300 ng of PML expression vector. PODs were detected by immunostaining fixed cells with PML-specific antibody. Fluorescence microscopy was performed using appropriate filters (CY3, YFP). Representative 0.5-µm optical sections are shown. (B) In vitro interaction of MCF-7-expressed TDG with recombinant GST-SUMO-1 and GST-PML. Cellular lysates were incubated with GST fusion proteins, and bound TDG was detected by immunoblotting. (C) Hyperthermic stress releases TDG from PODs. Live MCF-7 cells expressing YFP-TDG and CFP-PML were imaged initially at 37°C and following incubation at 42°C for 15 min. Lysates of control and heat-shocked cells (15 and 30 min) were immunoblotted with a TDG-specific antibody to reveal unmodified and sumoylated (S) YFP-TDG. (D) TDG-specific antibody recognizes mouse and human TDG. Whole-cell extracts from MCF-7 cells transfected with empty vector (control) or mouse TDG expression vector were immunoblotted with purified TDG-specific rabbit IgG. (E) Nuclear colocalization of endogenous TDG and PML. Untransfected MCF-7 and NIH 3T3 cells were immunostained with TDG- and PML-specific antibodies and fluorophore-conjugated (FITC, CY3) secondary antibodies. The fluorescence intensity plot illustrates the coincidence of peak fluorescence for TDG (CY3, red) and PML (FITC, green). Measurements were obtained by performing a line scan across three PODs using Axiovision software. Representative 0.5-µm optical sections are shown.

 
To rule out the potential influence of overexpression on the subcellular distribution of TDG, we determined whether native TDG is found in PODs by immunostaining MCF-7 cells with purified TDG-specific rabbit IgG and commercial PML-specific mouse monoclonal antibodies. TDG-specific rabbit antibody raised against recombinant mouse TDG also recognizes human TDG (Fig. 1D) but does not cross-react with PML (data not shown). In untransfected MCF-7 cells, endogenous TDG staining was observed in a granular pattern throughout the nucleoplasm: a subpopulation of cells consistently displayed increased staining within the PODs (Fig. 1E, panels I to III). Similar results were obtained with immortalized mouse NIH 3T3 cells (Fig. 1E, panels IV to VI). These findings indicate that a small fraction of endogenous TDG localizes to the PODs, consistent with our transient-expression studies.

Identification of POD targeting and SUMO-1 binding domains. In order to identify protein domains within TDG essential for POD localization, we generated a series of CFP-tagged amino- and carboxy-terminal deletions and examined their cellular localization following coexpression with PML. Our analysis indicated that amino-terminal residues were required for nuclear targeting of TDG (Fig. 2A). Deletion of residues 1 to 156 shifted localization predominantly to the cytoplasm, whereas deletion of residues 1 to 121 resulted in similar levels of nuclear and cytoplasmic fluorescence. In order to assess the contribution of this region to POD localization, we engineered deletion constructs containing the SV40 NLS. FLAG epitope-tagged amino- and carboxy-terminal deletions were coexpressed with YFP-PML, and subcellular localization was examined by immunostaining with anti-FLAG antibody. Remarkably, we found that TDG lacking the first 121 residues (i.e., NLS122-421) accumulated preferentially in the PODs compared to wild-type TDG (Fig. 2B, compare panels I and II). Consequently, in a large fraction of cells (40 to 50%) nuclear fluorescence was predominantly associated with PODs, whereas in the case of wild-type TDG, substantial nucleoplasmic localization was observed. A lysine-rich regulatory domain (LRD; residues 70 to 118) previously shown to be acetylated by CBP/p300 is contained within this deleted region (46). Further removal of residues 123 to 156 (NLS157-421) led to a dramatic decrease in the number of expressing cells, with the majority of the tagged protein being found in large aberrant nucleoplasmic foci that also contained PML. Loss of carboxy-terminal residues 346 to 421 did not affect POD targeting; however, further deletion to residue 307 completely abrogated TDG accumulation in these structures. These data suggest that both amino- and carboxy-terminal regions of TDG contribute to POD localization.


Figure 2
View larger version (30K):
[in this window]
[in a new window]

  		FIG. 2. Deletion mapping of SUMO-1 binding and POD-targeting domains in mouse TDG. (A) Subcellular localization of amino- and carboxy-terminal deletions of TDG. CFP-tagged deletions of TDG were expressed in MCF-7 cells and analyzed by direct fluorescence microscopy. (B) Amino- and carboxy-terminal deletions of TDG (depicted in the upper panel) containing the FLAG epitope were coexpressed with YFP-PML in MCF-7 cells. Approximately 300 ng of TDG and 500 ng of YFP-PML expression vectors were used. TDG was detected by immunostaining with anti-FLAG monoclonal antibody and CY3-conjugated secondary antibody. Representative 0.5-µm optical sections are shown. The location of the LRD is indicated. Note that FLAG-tagged amino-terminal deletions include the SV40 NLS. (C) In vitro interaction of TDG with GST-SUMO-1 is enhanced by deletion of residues 1 to 121. In vitro-translated 35S-radiolabeled full-length TDG and the indicated deletion mutants were used in binding assays with GST-SUMO-1 and GST. Bound proteins were detected by autoradiography.

 
In light of evidence suggesting a SUMO-1-dependent mechanism in POD targeting, we also tested amino- and carboxy-terminal deletions of TDG produced by in vitro transcription/translation for binding to GST-SUMO-1. While full-length TDG displayed only weak SUMO binding activity, removal of residues 1 to 121 dramatically stimulated binding (Fig. 2C). In contrast, deletion of residues 1 to 156 or 307 to 346 resulted in complete loss of SUMO binding activity. These findings suggest that two distinct regions of TDG (residues 122 to 156 and 307 to 346) are essential for SUMO-1 binding, whereas a third region (residues 1 to 121) containing the LRD appears to suppress binding activity. Notably, these domains are also involved in POD targeting, suggesting that the SUMO-1 binding activity of TDG may be required for targeting to these nuclear structures.

DNA interactions regulate SUMO-1 binding activity. In vitro mapping studies suggested that the amino-terminal region (residues 1 to 121) modulated interactions with SUMO-1. In human TDG, this region has been found to be essential for nonspecific DNA binding and interactions with abasic sites (44). Since in vitro translation reaction mixtures contain plasmid DNA, we wanted to establish whether the DNA binding properties of the amino-terminal region could interfere with SUMO-1 binding in vitro; therefore, we performed SUMO binding experiments in the presence of ethidium bromide to effectively prevent DNA binding (25). A marked stimulation in SUMO-1 binding from full-length TDG was observed, while binding of the 122-421 protein was not affected (Fig. 3A). These data suggested that DNA binding by the amino-terminal region of TDG may prevent SUMO-1 recognition.


Figure 3
View larger version (25K):
[in this window]
[in a new window]

  		FIG. 3. The amino-terminal DNA binding domain of TDG regulates SUMO-1 binding activity. (A) Binding of full-length in vitro-translated TDG to GST-SUMO-1 is sensitive to ethidium bromide. In vitro-translated 35S-radiolabeled full-length TDG or the 122-421 truncated protein was bound to GST-SUMO-1 in the presence or absence of ethidium bromide. (B) Recombinant mouse TDG binds normally paired as well as G:T or G:U mispaired duplex oligonucleotides. An electrophoretic mobility shift assay was performed using the indicated radiolabeled duplex oligonucleotides. Approximately 100-fold molar excess of the same unlabeled oligonucleotides was used as competitor DNA. (C) Residues 1 to 121 of mouse TDG are essential for DNA binding. A biotin-tagged duplex oligonucleotide (500 ng) containing a G:T mispair was bound to recombinant full-length TDG or the 122-421 protein (500 ng each). DNA-protein complexes were isolated using streptavidin-Sepharose and analyzed by immunoblotting with a monoclonal antihistidine antibody. (D) DNA binding suppresses the SUMO-1 binding activity of TDG. Recombinant TDG (200 ng) was bound to GST-SUMO-1 in the presence of increasing amounts (70, 210, and 420 ng) of duplex oligonucleotides containing a G:T mispair.

 
Human TDG has been shown to bind both G:T/U-mispaired and normally paired DNA (19). The mouse and human TDG orthologs are highly conserved within the central enzymatic core and less well conserved in the amino- and carboxy-terminal regions. Using an electrophoretic mobility shift assay (13), we confirmed that bacterially expressed mouse TDG has similar DNA binding specificity and could form complexes with normally paired (G:C) as well as G:T/U-mispaired duplex oligonucleotides (Fig. 3B). The requirement of the 1-121 region in DNA interactions was confirmed using the ABCD binding assay (15). In these assays, recombinant full-length TDG bound to a G:T-mispaired oligonucleotide, while an amino-terminal-truncated protein fragment (122-421) did not detectably associate with DNA (Fig. 3C). To determine whether residues 1 to 121 contained a modular DNA binding domain, we assayed a GST fusion protein containing this region for binding to G:T duplex oligonucleotide using the ABCD assay (data not shown). The fact that DNA binding was not observed suggests that the amino-terminal region does not independently associate with DNA. Interestingly, the NLS122-421 protein displayed preferential POD localization upon coexpression with PML (Fig. 2B), suggesting that loss of DNA interactions promotes POD targeting. We tested whether the DNA and SUMO binding activities of TDG are mutually exclusive by performing binding studies with GST-SUMO-1 and recombinant TDG in the presence of increasing amounts of duplex oligonucleotide containing a G:T mispair. A dose-dependent reduction in SUMO binding was consistently observed in the presence of DNA (Fig. 3D); in contrast, preincubation of TDG with SUMO-1 did not affect binding to G:T or G:C duplex oligonucleotides (see Fig. S1 in the supplemental material).

Identification of a novel amino-terminal SUMO-1 binding motif. In vitro mapping studies indicated that residues 122 to 156 are required for SUMO-1 binding; therefore, we examined the amino acid sequence within this region and identified four residues (IVII; amino acids 145 to 148) which are identical to the recently characterized SUMO-1 binding consensus motif (I/V-X-I/V-I/V) (42). Furthermore, this motif is flanked by an aspartic acid (i.e., DIVII) residue also present adjacent to the SUMO-1 binding motifs of the RanBP2/NUP358 and SUMO activating enzyme 2 (SAE2) proteins (42). The DIVII residues are conserved in mammalian, chicken, and Drosophila melanogaster TDG orthologs and are contiguous with the conserved GINPGL glycosylase motif (2, 18) (Fig. 4A). Previous structural studies using a truncated form of human TDG have identified a carboxy-terminal SUMO-1 binding motif (VQEV) (1) that is conserved in mouse, human, and chick TDG, but not in the Drosophila ortholog (Fig. 4A). In order to establish whether the putative amino- and carboxy-terminal SBMs in mouse TDG bind SUMO-1, we generated a series of mutant proteins with single amino acid substitutions and measured their ability to bind to GST-SUMO-1 (Fig. 4B). Alanine substitution mutants were generated for each residue in the DIVII motif, whereas a single glutamic acid-to-glutamine (E321Q) substitution in the VQEV motif was analyzed, as this had been previously reported to abrogate SUMO-1 binding in human TDG (1). Substitution of specific residues within each putative SBM independently abrogated SUMO-1 binding, suggesting that in the context of full-length TDG both motifs are essential for stable SUMO-1 interactions. Specifically, within the DIVII motif, the I145A and V146A substitutions produced small but consistent reductions in binding, whereas the D144A and I147A substitutions displayed more pronounced loss of binding. In contrast, the I148A substitution appeared to stimulate binding. The E321Q substitution in the carboxy-terminal SBM completely abrogated binding. To rule out gross effects of the amino acid substitutions on protein folding, we performed DNA glycosylase assays using radiolabeled duplex oligonucleotides containing a single G:T mispair (see Fig. S2 in the supplemental material). Comparable levels of base excision were observed with in vitro-translated wild-type TDG, D144A, and E321Q, consistent with proper folding. Similar results on SUMO binding were obtained when we expressed the respective TDG mutants in MCF-7 cells and used whole-cell lysates in interaction studies (data not shown) (see Fig. S3 in the supplemental material).


Figure 4
View larger version (36K):
[in this window]
[in a new window]

  		FIG. 4. SUMO-1 binding activity of TDG resides in amino-terminal and carboxy-terminal motifs. (A) Amino acid sequence alignments of the mouse (47), human (32), chicken (51), and Drosophila melanogaster (18) TDG orthologs showing putative SBMs. The coordinates for the mouse sequences are indicated. Complete sequences were aligned using Clustal W software, but only pertinent regions are shown (136 to 155 and 312 to 326 of mouse TDG). The location of the conserved DIVII and VQEV motifs (boxed) as well as the active site glycosylase motif (GINPGL) and the substrate recognition motif (VMPSSSAR) (19) are shown. Asterisks indicate identical residues, while colons indicate conserved residues. The different engineered substitution mutants of the DIVII and VQEV motifs are indicated. (B) Single amino acid substitutions within the conserved DIVII and VQEV motifs abrogate SUMO-1 binding. Radiolabeled in vitro-translated wild-type TDG and the indicated substitution mutants were analyzed for binding to GST-SUMO-1 in the presence of ethidium bromide. Binding was measured by phosphorimaging. (C) In vitro sumoylation of recombinant GST-p53. In vitro sumoylation reactions were performed by incubating GST-p53 with sumoylation enzymes (GST-SAE1, GST-SAE2, and UBC9) and SUMO-1. Reaction products were immunoblotted with p53 or SUMO-1 antibodies. (D) SBMs are required for optimal binding to conjugated SUMO-1. The products of the GST-p53 sumoylation and mock sumoylation reactions were bound to glutathione affinity beads and used for interaction assays with in vitro-translated TDG and SBM mutants (D144A and E321Q). Binding experiments were also carried out with sumoylation reaction mixtures lacking GST-p53 to exclude interactions of TDG with GST-SAE1/SAE2. The results of three independent experiments are plotted, showing the mean percent binding of input proteins to beads containing sumoylated GST-p53. The standard error is shown.

 
In order to determine whether the SBMs are required for binding to SUMO-1 conjugated to a target protein, we employed an in vitro sumoylation system (10) reconstituted with bacterially expressed enzymes (SAE1, SAE2, and UBC9) and SUMO-1 to sumoylate a purified bacterially expressed GST fusion of tumor suppressor p53 protein (39). As a control, mock sumoylation reactions were carried out in the absence of SUMO-1. Analysis of the reaction products by immunoblotting revealed the presence of a protein band reactive with both p53 and SUMO-1 antibodies only in the sumoylation reaction (Fig. 4C). The products of both the mock and sumoylation reactions were bound to glutathione affinity beads and used in interaction studies with in vitro-translated TDG and SBM mutants (D144A and E321Q). We observed appreciable binding of wild-type TDG on beads containing sumoylated GST-p53, while only marginal binding was detected with mock-sumoylated GST-p53 (Fig. 4D). The D144A mutant displayed substantially reduced binding, while the E321Q substitution almost completely abrogated binding. Since the sumoylation reaction mixtures contain GST fusions of the SAE1 and SAE2 enzymes, we also performed binding reactions with the products of sumoylation reactions lacking GST-p53. In this case, binding of TDG was not detected. These findings indicate that the DIVII residues in TDG constitute a bona fide SBM and that the SUMO-1 binding activity of TDG resides within two separate motifs.

To assess whether the amino- and carboxy-terminal SBMs are required for POD translocation, we expressed the SUMO-1 binding-deficient mutants (D144A, I147A, and E321Q) with YFP-PML and monitored cellular localization by immunostaining with an anti-FLAG antibody (Fig. 5A and B). This analysis showed that mutations within the amino-terminal SBM substantially reduced the number of transfected cells displaying POD accumulation of TDG; in contrast, the E321Q substitution completely abolished POD accumulation. These findings indicate that both SBMs in TDG are involved in POD targeting.


Figure 5
View larger version (30K):
[in this window]
[in a new window]

  		FIG. 5. SUMO-1 binding activity of TDG is essential for CBP activation and normal POD recruitment. (A) SUMO-1 binding mutants are defective in POD recruitment. FLAG-tagged wild-type TDG and the indicated mutants were coexpressed with YFP-PML, and POD recruitment was analyzed by indirect immunofluorescence using anti-FLAG antibody. Representative 0.5-µm optical sections are shown. (B) Incidence of POD accumulation for wild-type TDG and mutants coexpressed with YFP-PML. An average of 120 cells were counted to determine POD accumulation of wild-type TDG or SUMO-1 binding mutants. Error bars represent standard deviations of determinations from three independent experiments. (C) Activation of CBP-dependent transcription is abrogated by E321Q substitution in the VQEV motif. (D) Mutations in DIVII motif also abrogate CBP activation. Expression vectors for GAL-CBP, wild-type TDG, and point mutants of TDG were cotransfected into MCF-7 cells. The luciferase reporter plasmid contains five copies of the GAL4 DNA binding site fused to the core ß-globin promoter. The amounts of transfected TDG expression vectors were titrated to obtain approximately equal levels of protein expression (immunoblot, lower panel).

 
Previous studies have demonstrated a potent stimulatory function of TDG on CBP-dependent transcription (46). To assess whether the SUMO binding activity of TDG is involved in stimulating CBP-dependent transcription, we tested the activation potential of the D144A, I147A, and E321Q mutants using a chimeric fusion of CBP and the GAL4 DNA binding domain on a GAL4-responsive reporter gene (Fig. 5C and D). The amounts of transfected expression vectors were adjusted to achieve approximately equivalent levels of expression of wild-type TDG and mutants. In accordance with the SUMO-1 binding and POD localization studies, the D144A, I147A, and E321Q mutants were found to be defective in CBP activation. Previous studies have shown that TDG interacts with the histone acetyltransferase and a carboxy-terminal domain (CH3) of CBP (46). Since the amino acid substitutions may affect interactions with CBP, we tested in vitro-translated wild-type, D144A, and E321Q proteins for binding to full-length recombinant CBP (see Fig. S4 in the supplemental material). However, no differences in binding were detected, suggesting that abrogation of the CBP activation properties in the SBM mutants is likely due to loss of SUMO binding activity. We also investigated whether the previously reported sumoylation sites in CBP (14, 24) are required for activation by TDG. Accordingly, we examined the ability of TDG to activate a CBP mutant containing lysine-to-arginine substitutions at sumoylation sites K999, K1034, and K1057. The transcriptional activity of this mutant was also robustly stimulated by TDG, indicating that CBP sumoylation is not essential for this effect (see Fig. S5 in the supplemental material).

Sumoylation of TDG regulates association with CBP and subnuclear localization. Human TDG has been reported to be sumoylated at a single carboxy-terminal lysine residue (20). Consistent with this, we have observed a higher-molecular-weight band in immunoblots of mouse cell extracts and in transfected human cells expressing mouse TDG (Fig. 6 and data not shown). To determine whether the higher-molecular-weight band corresponds to sumoylated TDG, we cotransfected FLAG-TDG and HA-SUMO-1 expression vectors into MCF-7 cells and analyzed cellular lysates by immunoblotting with anti-FLAG and anti-HA antibodies (Fig. 6A). This analysis indicated that the higher-molecular-weight band detected with the anti-FLAG antibody corresponds to SUMO-1-modified TDG, since it was also detected with the anti-HA antibody. The sumoylated form of TDG is not observed when lysine 341, located within the SUMO consensus conjugation site (VKEE), is mutated to arginine (Fig. 6B).


Figure 6
View larger version (38K):
[in this window]
[in a new window]

  		FIG. 6. TDG sumoylation occurs in the nucleus and does not require noncovalent SUMO-1 binding activity. (A) FLAG-TDG and HA-SUMO-1 expression vectors were cotransfected into MCF-7 cells, and lysates were analyzed by immunoblotting using anti-FLAG or anti-HA antibody. Sumoylated TDG (S-TDG), TDG, and nonspecific (NS) bands are indicated. (B) In vivo sumoylation occurs at lysine 341 of mouse TDG. Expression vectors for TDG and K341R were transfected into MCF-7 cells, and cellular lysates were blotted with TDG-specific antibody. (C) TDG is sumoylated predominantly in the nucleus. Whole-cell lysates of MCF-7 cells transfected with expression vectors for the indicated CFP fusions were analyzed by immunoblotting with TDG-specific antibody. CFP157-421 is localized preferentially in the cytoplasm (Fig. 2A). (D) SUMO-1-binding-deficient mutants are sumoylated efficiently in MCF-7 cells. Cellular lysates from MCF-7 cells transfected with the indicated expression vectors were analyzed by immunoblotting with a TDG-specific antibody.

 
In vitro sumoylation experiments, using recombinant TDG as a substrate, confirmed that the sumoylation machinery is present in both nuclear and cytoplasmic extracts (data not shown). We assessed whether TDG sumoylation occurred within the nucleus by expressing CFP-tagged amino-terminal deletions (CFP122-421, CFP157-421) of TDG defective in nuclear targeting and determining the level of sumoylation by immunoblotting cellular lysates with a TDG-specific antibody (Fig. 6C). The CFP157-421 fusion protein that localized preferentially to the cytoplasm was not efficiently sumoylated, since a higher-molecular-weight band corresponding to sumoylated TDG was not readily detectable. A truncated form of TDG (CFP32-272) lacking the sumoylation site was also not sumoylated. These data suggest that TDG sumoylation takes place in the nuclear compartment and/or at the nuclear membrane. The observation that SUMO-1 binding mutants (E321Q, D144A, and I147A), defective in POD localization, were sumoylated efficiently in vivo indicates that noncovalent SUMO binding is not required for sumoylation (Fig. 6D). Furthermore, these findings also suggest that sumoylation is likely not occurring within the PODs.

Sumoylation has been shown to regulate cellular partitioning of PML and other proteins (49); we therefore compared the subcellular localization of Renilla GFP fusions of the sumoylation-deficient mutant K341R and wild-type TDG. Notably, the K341R mutant protein was found to accumulate exclusively in PODs in approximately 30% of transfected cells, whereas wild-type TDG exhibited its characteristic nucleoplasmic distribution (Fig. 7A). The enhanced ability of the K341R mutant to localize to the PODs suggests that sumoylation of TDG negatively regulates POD translocation.


Figure 7
View larger version (28K):
[in this window]
[in a new window]

  		FIG. 7. Sumoylation of TDG abrogates interaction with CBP in vitro and negatively regulates POD translocation. (A) A sumoylation-defective TDG mutant localizes preferentially to PODs. Expression vectors for GFP-TDG or GFP-TDG(K341R) were transfected into MCF-7 cells, and subcellular localization was analyzed following immunostaining with PML-specific antibody. Representative 0.5-µm optical sections are shown. (B) In vitro sumoylation of TDG using recombinant SUMO-1 conjugation enzymes (UBC9, SAE1, and SAE2). Sixty nanograms of purified reaction products was analyzed by immunoblotting with TDG- and SUMO-1 specific antibodies. Mock-sumoylated TDG was produced by performing sumoylation reactions without SUMO-1. (C) Sumoylated mouse TDG does not bind to DNA. Approximately 400 ng of either mock-sumoylated or sumoylated TDG was analyzed for DNA binding using the ABCD assay. Supernatants of each binding reaction mixture were also immunoblotted to demonstrate the stability of the modification during the assay (SUP-M and SUP-S). (D) TDG sumoylation abrogates CBP-TDG interactions. Mock or sumoylated TDG produced in vitro (400 ng) was incubated with recombinant FLAG-CBP (100 ng). Anti-FLAG resin was used to immunoprecipitate CBP, and the presence of TDG was detected by immunoblotting. Supernatants of each binding reaction mixture were also immunoblotted (SUP-M and SUP-S). (E) Sumoylated TDG is not acetylated efficiently by CBP. Mock-sumoylated or sumoylated TDG (400 ng) was incubated with CBP (100 ng) in the presence of [14C]acetyl-CoA. Reaction products were separated by electrophoresis, and acetylation was detected by autoradiography. (F) Sumoylated TDG does not bind GST-SUMO-1. Mock-sumoylated or sumoylated TDG was analyzed for binding to GST-SUMO-1.

 
The conjugation of SUMO-1 to lysine 341 covalently links a bulky peptide to a region that has been previously shown to interact with CBP/p300 (46). Moreover, the presence of SUMO binding motifs within the amino- and carboxy-terminal regions suggests that sumoylation of TDG may promote intramolecular interactions that drastically alter the conformation of this protein. To investigate the functional consequences of sumoylation, we produced sumoylated recombinant TDG in vitro along with a mock-sumoylated control (Fig. 7B). Mock-sumoylated TDG and sumoylated TDG proteins were analyzed for DNA binding to a duplex oligonucleotide containing a single G:T mispair using the ABCD assay. As reported for the human ortholog, sumoylated mouse TDG failed to interact with DNA (Fig. 7C). We performed interaction studies with purified recombinant baculovirus-expressed CBP bearing a FLAG epitope tag. Protein complexes were captured using anti-FLAG affinity resin. While binding of mock-sumoylated TDG to CBP was readily observed, only weak interactions were observed with sumoylated TDG (Fig. 7D). Consequently, sumoylated TDG was not appreciably acetylated by CBP in the presence of [14C]AcCoA (Fig. 7E). To establish whether intramolecular SUMO binding in sumoylated TDG occludes both SBMs, we tested whether sumoylated TDG can interact with GST-SUMO-1 in vitro. In contrast to mock-sumoylated TDG, sumoylated TDG failed to interact with GST-SUMO-1 (Fig. 7F). To confirm that sumoylated TDG is enzymatically active, we performed DNA glycosylase assays using radiolabeled duplex oligonucleotides containing a single G:U mispair. As previously reported (20), sumoylated TDG displayed enhanced G:U processing activity compared to unmodified TDG and mock-sumoylated TDG (see Fig. S6 in the supplemental material). Therefore, abrogation of the interaction of sumoylated TDG with CBP is unlikely due to aberrant misfolding but likely involves sumoylation-induced conformational changes.


   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
We have investigated the role of sumoylation and noncovalent SUMO-1 binding in the regulation of subcellular localization and biochemical properties of thymine DNA glycosylase. Our studies have mapped SUMO-1 binding activity to two separate motifs (SBMs) located in the amino- and carboxy-terminal regions that are essential for POD localization and activation of CBP-dependent transcription. The activities of the SBMs are regulated by DNA interactions, and uncoupling of TDG from DNA appears to be an essential step in POD translocation. In addition, we have established that TDG sumoylation regulates molecular interactions with CBP as well as translocation to PODs.

PODs contain regulatory proteins involved in different nuclear processes, including DNA repair and transcription (9, 50). Diverse models of POD function have been proposed, including their potential role as sites of storage and modification of nuclear factors (reviewed in reference 26). For example, tumor suppressor p53 acetylation by CBP in PODs constitutes a critical step in p53 activation during RAS-induced premature senescence (38). Interestingly, there is evidence that PODs associate with transcriptionally active genomic loci, and transcription has been detected at the periphery of these structures (5, 48). We have established the presence of endogenous TDG in PODs using a TDG-specific antibody and demonstrated the SUMO-dependent recruitment of exogenously expressed TDG to these nuclear structures (Fig. 1). While recent studies have reported a direct interaction of mouse TDG with PML (45), we have not observed significant binding of these proteins in vitro. It is unlikely that the lack of interaction is due to misfolding of bacterially expressed PML, since we have obtained similar results with baculovirus-expressed FLAG-tagged PML (data not shown). Furthermore, no appreciable binding of sumoylated TDG with PML in vitro was observed (data not shown). Our findings suggest that TDG may be recruited to PODs by sumoylated PML (30) and/or other sumoylated POD components. We base this assertion on the requirement of SUMO-1 binding activity of TDG for POD recruitment and the fact that mild hyperthermic stress, previously shown to cause rapid desumoylation of PML and SP100 (34), abrogates POD accumulation of TDG without affecting PML localization.

Deletion analysis of TDG indicated the presence of two regions (residues 122 to 156 and 308 to 346) essential for SUMO-1 binding in vitro, as well as a third region containing the LRD that exerted an inhibitory function (Fig. 2B). We have identified a novel SUMO-1 binding motif (DIVII) within the 122-156 region conforming to the recently reported consensus SUMO-1 binding site (I/V-X-I/V-I/V) that mediates recognition of SUMO-modified proteins (42). The observation that single amino acid substitutions within the DIVII motif (particularly D144A and I147A) substantially reduce SUMO-1 binding in the context of full-length TDG clearly demonstrates that this is a bona fide SBM (Fig. 4B) (see also Fig. S3 in the supplemental material).

Recent structural analysis of amino- and carboxy-terminal-truncated sumoylated human TDG identified a SUMO-1 binding motif (VQEV) located within the carboxy terminus near the SUMO conjugation site (1). Mutational analysis of mouse TDG revealed that both the amino- and carboxy-terminal SBMs (i.e., DIVII and VQEV) are required for stable interactions with both free and conjugated SUMO-1 (Fig. 4). Consequently, these observations suggest that SUMO-1 may bind concurrently to both SBMs. However, as both motifs have been reported to make nearly identical contacts (1, 42, 43) with residues on SUMO-1, we infer that there is considerable plasticity in these interactions. Accordingly, the I/V-X-I/V-I/V motif has recently been shown to interact bidirectionally with SUMO-1 (43). In contrast to a recent report (45), we have found that the SUMO-1 binding activity of TDG is not required for covalent conjugation of SUMO-1 (Fig. 6). This discrepancy could be explained by the reliance of previous studies on the analysis of a TDG mutant containing an 11-amino-acid deletion, whereas more subtle single amino acid substitutions were employed in our studies.

The essential role of SUMO binding in POD translocation is demonstrated by the observation that single amino acid substitutions that decrease or abolish SUMO-1 binding also disrupt POD targeting (Fig. 4 and 5). Notably, both SBMs, in the context of full-length TDG, appeared to be required for optimal POD translocation. In contrast, similar analysis of deletion mutants provided some discrepant results. Deletion of the 122-156 region, containing the amino-terminal SBM, did not prevent POD localization in vivo. This may result from an inhibitory effect of the amino terminus (residues 1 to 121) on the carboxy-terminal SBM; removal of this region in the 157-421 deletion mutant may relieve inhibition and promote POD targeting. The TDG amino terminus (residues 1 to 121) is required for tight interaction with DNA and abasic sites (44) (Fig. 3C). Our findings suggest that uncoupling of TDG from DNA is necessary to unmask SUMO binding activity and promote POD translocation (Fig. 8). This may occur following excision of base mispairs via apurinic/apyrimidinic endonuclease (APE)-mediated displacement of TDG and/or following sumoylation of TDG (20). In the latter case, our findings suggest that sumoylation would prevent POD targeting by occluding the SBMs via intramolecular interactions; consequently, translocation to the PODs would require removal of SUMO-1 by isopeptidases (22).


Figure 8
View larger version (20K):
[in this window]
[in a new window]

  		FIG. 8. Model for SUMO-1-dependent regulation of TDG subcellular localization and function. TDG associated with transcriptionally active euchromatin (46) initiates repair of G:T/U mispairs within CpG dinucleotides in a process that is likely linked to transcription. TDG-mediated repair may require CBP/p300 acetylase for local chromatin remodeling and/or regulation of repair enzymes via acetylation (3, 21, 46). Transcription has been detected in the periphery of PODs, and there is evidence for association of these structures with transcriptionally active genomic loci (5, 48). POD localization of TDG is dependent on its intrinsic SUMO binding activity and may be required to deliver this enzyme to transcriptionally active loci. For this purpose, TDG would require regulatory switches to control transit to PODs. Chromatin-associated TDG may not translocate to these nuclear structures due to DNA interactions that suppress SUMO binding activity. On the basis of our experimental findings, we propose that POD translocation is contingent upon uncoupling TDG from DNA, which may occur following base excision as a result of displacement by APE and/or sumoylation. Sumoylated TDG may not translocate to the PODs due to occlusion of the SBMs by intramolecular SUMO-1 interactions. In this case, desumoylation would be required to permit POD translocation. Within these structures, TDG may be posttranslationally modified by CBP, as previously demonstrated for p53 (38), and/or assembled into functional complexes for delivery to transcriptionally active genomic loci.

 
A number of DNA repair factors have been shown to transit in PODs prior or following DNA damage (reviewed in reference 9), including enzymes involved in repair of double-stranded DNA breaks (DSB). For example, MRE11 and NBS1 associate with PODs in unirradiated cells and relocate to sites of DNA damage following gamma irradiation (4, 29). These observations suggest that PODs may act as sites of storage and/or assembly of DNA repair protein complexes. Consistent with a role of PODs in genome maintenance, PML null mice display increased susceptibility to tumors following exposure to carcinogens (40). The association of PODs with transcriptionally active genomic loci (5, 48) provides an attractive model for TDG function (Fig. 8). While DSB occur infrequently, it is estimated that several hundred DNA mispairs occur daily per cell as a result of cytosine deamination (27), suggesting that the DNA repair functions of TDG would be required more frequently to maintain genome integrity. The well-documented interactions of TDG with transcriptional coactivators and sequence-specific transcription factors suggest that the genome surveillance functions of this enzyme are linked to transcription (8, 46, 47). In vitro studies have demonstrated that CBP/p300 and TDG form stable ternary complexes with DNA containing G:T/U mispairs (46). The recruitment of CBP/p300 to repair sites in vivo may be required to promote local chromatin remodeling and/or regulate the functions of BER enzymes, such as TDG, APE, and DNA polymerase ß, previously shown to be acetylated by these factors (3, 21, 46). On the basis of the association of PODs with genomic loci (48), it is plausible that transient association of TDG with these nuclear structures is required to deliver this DNA repair enzyme to sites of active transcription, ensuring efficient repair of damaged CpG dinucleotides. Alternatively, as reported for p53 (38), POD localization of TDG may serve as a regulatory step to promote acetylation by CBP (46).

Our studies indicate that the SUMO-1 binding activity of TDG is essential for activation of CBP-dependent transcription. Using reporter gene assays, we have demonstrated that SBM mutants (D144A, I147A, and E321Q), defective in SUMO-1 binding, do not mediate CBP activation (Fig. 5). Given that SUMO-1 binding is also required for POD recruitment, we are not able to resolve whether POD recruitment is required for CBP activation. However, this seems unlikely, since coexpression of TDG with PML does not produce greater levels of CBP activation despite increasing POD localization of TDG (data not shown). Accordingly, we believe that SUMO-1 binding activity per se, and not POD targeting, is essential for CBP activation. The presence of a sumoylation-dependent transcriptional repressor domain in CBP/p300 that recruits histone deacetylases (14, 24) suggests a plausible role of TDG binding in promoting derepression by displacement of histone deacetylases. However, a CBP mutant containing lysine-to-arginine substitutions at characterized sumoylation sites (24) was also robustly activated by TDG (see Fig. S5 in the supplemental material), indicating that CBP sumoylation is not essential.

TDG sumoylation has been shown to abrogate DNA binding activity and has been proposed as a mechanism to promote the turnover of TDG from abasic sites following base excision (20). We have confirmed that sumoylated mouse TDG is also defective in DNA binding. Our studies suggest that TDG sumoylation also plays an important role in regulating POD translocation and protein-protein interactions. Recent structural studies of sumoylated human TDG have revealed important insights on the conformational changes resulting from this covalent modification (1, 20). In addition, partial proteolysis studies have shown that sumoylation induces conformational changes involving interactions between the amino- and carboxy-terminal regions of human TDG (44). We have now identified a conserved amino-terminal SBM that in concert with a carboxy-terminal SBM may account for the observed sumoylation-induced conformational changes. Indeed, binding of the amino- and carboxy-terminal domains to SUMO-1 likely interferes with the DNA binding functions associated with the amino terminus (residues 1 to 121) (44). We have examined the effect of sumoylation on interactions with CBP and intermolecular SUMO-1 recognition in mouse TDG. Remarkably, TDG sumoylation abrogates both CBP interaction as well as intermolecular SUMO binding (Fig. 7). Based on these observations, loss of intermolecular SUMO binding should prevent POD translocation. Corroborating evidence for a role of sumoylation in regulating POD translocation comes from the analysis of the sumoylation-deficient mutant GFP-K341R, which displays exclusive POD localization in a subpopulation of cells. In view of our biochemical studies, we believe that loss of negative regulation (i.e., sumoylation) dramatically favors POD translocation. In view of the documented interactions of CBP/p300 with the TDG amino terminus and the resulting acetylation of the lysine-rich regulatory domain (46), we cannot exclude a role for CBP/p300 and acetylation in regulating POD targeting. Nevertheless, CBP is not likely to be a direct intermediary in POD recruitment since, when coexpressed with PML and TDG, it did not consistently accumulate with TDG in the PODs (data not shown).

In conclusion, we have elucidated the roles of sumoylation and noncovalent SUMO-1 binding in regulating the subcellular localization and biochemical properties of TDG. Although the significance of POD localization remains to be established, our findings suggest a key role for these nuclear structures in regulating the functions of TDG in transcription and/or genome maintenance.


   ACKNOWLEDGMENTS
 
We thank Ron Evans, Thomas Sternsdorf, Pierre Chambon, Kristen Walker, and Ron Hay for the generous gift of plasmids and Lisa Danielczak for technical help. We are grateful to Joe Torchia, Rob Sladek, and David Litchfield for critical reading of the manuscript.

This work was supported by operating grants from the Canadian Institutes of Health Research, The Cancer Research Society, and the Canadian Cancer Society as well as infrastructure funding from the Canadian Foundation for Innovation and Ontario Innovation Trust. A.R. is a recipient of an OGSST award from the Ontario Ministry of Education.

We do not have any financial interests related to this work.


   FOOTNOTES
 
* Corresponding author. Mailing address: Department of Physiology and Pharmacology, Siebens-Drake Medical Research Institute, Schulich School of Medicine and Dentistry, University of Western Ontario, London, Ontario, Canada N6G 2V4. Phone: (519) 850-2942. Fax: (519) 661-3827. E-mail: mtini{at}uwo.ca. Back

{triangledown} Published ahead of print on 23 October 2006. Back

{dagger} Supplemental material for this article may be found at http://mcb.asm.org/. Back

{ddagger} R.D.M., A.R., and J.G. contributed equally to the manuscript. Back


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
1. Baba, D., N. Maita, J. G. Jee, Y. Uchimura, H. Saitoh, K. Sugasawa, F. Hanaoka, H. Tochio, H. Hiroaki, and M. Shirakawa. 2005. Crystal structure of thymine DNA glycosylase conjugated to SUMO-1. Nature 435:979-982.[CrossRef][Medline]

2. Barrett, T. E., R. Savva, G. Panayotou, T. Barlow, T. Brown, J. Jiricny, and L. H. Pearl. 1998. Crystal structure of a G:T/U mismatch-specific DNA glycosylase: mismatch recognition by complementary-strand interactions. Cell 92:117-129.[CrossRef][Medline]

3. Bhakat, K. K., T. Izumi, S. H. Yang, T. K. Hazra, and S. Mitra. 2003. Role of acetylated human AP-endonuclease (APE1/Ref-1) in regulation of the parathyroid hormone gene. EMBO J. 22:6299-6309.[CrossRef][Medline]

4. Bischof, O., S.-H. Kim, J. Irving, S. Beresten, N. A. Ellis, and J. Campisi. 2001. Regulation and localization of the Bloom syndrome protein in response to DNA damage. J. Cell Biol. 153:367-380.[Abstract/Free Full Text]

5. Boisvert, F. M., M. J. Hendzel, and D. P. Bazett-Jones. 2000. Promyelocytic leukemia (PML) nuclear bodies are protein structures that do not accumulate RNA. J. Cell Biol. 148:283-292.[Abstract/Free Full Text]

6. Chalkiadaki, A., and I. Talianidis. 2005. SUMO-dependent compartmentalization in promyelocytic leukemia protein nuclear bodies prevents the access of LRH-1 to chromatin. Mol. Cell. Biol. 25:5095-5105.[Abstract/Free Full Text]

7. Chen, D., M. J. Lucey, F. Phoenix, J. Lopez-Garcia, S. M. Hart, R. Losson, L. Buluwela, R. C. Coombes, P. Chambon, P. Schar, and S. Ali. 2003. T:G mismatch-specific thymine-DNA glycosylase potentiates transcription of estrogen-regulated genes through direct interaction with estrogen receptor alpha. J. Biol. Chem. 278:38586-38592.[Abstract/Free Full Text]

8. Chevray, P. M., and D. Nathans. 1992. Protein interaction cloning in yeast: identification of mammalian proteins that react with the leucine zipper of Jun. Proc. Natl. Acad. Sci. USA 89:5789-5793.[Abstract/Free Full Text]

9. Dellaire, G., and D. P. Bazett-Jones. 2004. PML nuclear bodies: dynamic sensors of DNA damage and cellular stress. Bioessays 26:963-977.[CrossRef][Medline]

10. Desterro, J. M., M. S. Rodriguez, G. D. Kemp, and R. T. Hay. 1999. Identification of the enzyme required for activation of the small ubiquitin-like protein SUMO-1. J. Biol. Chem. 274:10618-10624.[Abstract/Free Full Text]

11. Dolle, M. E. T., W. K. Snyder, D. B. Dunson, and J. Vijg. 2002. Mutational fingerprints of aging. Nucleic Acids Res. 30:545-549.[Abstract/Free Full Text]

12. Doucas, V., M. Tini, D. A. Egan, and R. M. Evans. 1999. Modulation of CREB binding protein function by the promyelocytic (PML) oncoprotein suggests a role for nuclear bodies in hormone signaling. Proc. Natl. Acad. Sci. USA 96:2627-2632.[Abstract/Free Full Text]

13. Fried, M., and D. M. Crothers. 1981. Equilibrium and kinetics of lac repressor-operator interactions by polyacrylamide gel electrophoresis. Nucleic Acids Res. 9:6505-6525.[Abstract/Free Full Text]

14. Girdwood, D., D. Bumpass, O. A. Vaughan, A. Thain, L. A. Anderson, A. W. Snowden, E. Garcia-Wilson, N. D. Perkins, and R. T. Hay. 2003. P300 transcriptional repression is mediated by SUMO modification. Mol. Cell 11:1043-1054.[CrossRef][Medline]

15. Glass, C. K., O. V. Devary, and M. G. Rosenfeld. 1990. Multiple cell type-specific proteins differentially regulate target sequence recognition by the alpha retinoic acid receptor. Cell 63:729-738.[CrossRef][Medline]

16. Goodman, R. H., and S. Smolik. 2000. CBP/p300 in cell growth, transformation, and development. Genes Dev. 14:1553-1577.[Free Full Text]

17. Greenblatt, M. S., W. P. Bennett, M. Hollstein, and C. C. Harris. 1994. Mutations in the p53 tumor suppressor gene: clues to cancer etiology and molecular pathogenesis. Cancer Res. 54:4855-4878.[Free Full Text]

18. Hardeland, U., M. Benetele, J. Jiricny, and P. Schar. 2003. The versatile thymine DNA-glycosylase: a comparative characterization of the human, Drosophila and fission yeast orthologs. Nucleic Acids Res. 31:2261-2271.[Abstract/Free Full Text]

19. Hardeland, U., M. Bentele, J. Jiricny, and P. Schar. 2000. Separating substrate recognition from base hydrolysis in human thymine DNA glycosylase by mutational analysis. J. Biol. Chem. 275:33449-33456.[Abstract/Free Full Text]

20. Hardeland, U., R. Steinacher, J. Jiricny, and P. Schar. 2002. Modification of the human thymine-DNA glycosylase by ubiquitin-like proteins facilitates enzymatic turnover. EMBO J. 21:1456-1464.[CrossRef][Medline]

21. Hasan, S., N. El-Andaloussi, U. Hardeland, P. O. Hassa, C. Burki, R. Imhof, P. Schar, and M. O. Hottiger. 2002. Acetylation regulates the DNA end-trimming activity of DNA polymerase beta. Mol. Cell 10:1213-1222.[CrossRef][Medline]

22. Hay, R. T. 2005. SUMO: a history of modification. Mol. Cell 18:1-12.[CrossRef][Medline]

23. Hendrich, B., U. Hardeland, H.-H. Ng, J. Jiricny, and A. Bird. 1999. The thymine glycosylase MBD4 can bind to the product of deamination at methylated CpG sites. Nature 401:301-304.[CrossRef][Medline]

24. Kuo, H.-Y., C.-C. Chang, J.-C. Jeng, H.-M. Hu, D.-Y. Lin, G. G. Maul, R. P. S. Kwok, and H.-M. Shih. 2005. SUMO modification negatively modulates the transcriptional activity of CREB-binding protein via the recruitment of Daxx. Proc. Natl. Acad. Sci. USA 102:16973-16978.[Abstract/Free Full Text]

25. Lai, J., and W. Herr. 1992. Ethidium bromide provides a simple tool for identifying genuine DNA-independent associations. Proc. Natl. Acad. Sci. USA 89:6958-6962.[Abstract/Free Full Text]

26. Lin, R. J., T. Sternsdorf, M. Tini, and R. M. Evans. 2001. Transcriptional regulation in acute promyelocytic leukemia. Oncogene 20:7204-7215.[CrossRef][Medline]

27. Lindahl, T. 1993. Instability and decay of the primary structure of DNA. Nature 362:709-715.[CrossRef][Medline]

28. Lucey, M. J., D. Chen, J. Lopez-Garcia, S. M. Hart, F. Phoenix, R. Al-Jehani, J. P. Alao, R. White, K. B. Kindle, R. Losson, P. Chambon, M. G. Parker, P. Schar, D. M. Heery, L. Buluwela, and S. Ali. 2005. T:G mismatch-specific thymine-DNA glycosylase (TDG) as a coregulator of transcription interacts with SRC1 family members through a novel tyrosine repeat motif. Nucleic Acids Res. 33:6393-6404.[Abstract/Free Full Text]

29. Mirzoeva, O. K., and J. H. Petrini. 2001. DNA damage-dependent nuclear dynamics of the Mre11 complex. Mol. Cell. Biol. 21:281-288.[Abstract/Free Full Text]

30. Muller, S., M. J. Matunis, and A. Dejean. 1998. Conjugation with the ubiquitin-related modifier SUMO-1 regulates the partitioning of PML within the nucleus. EMBO J. 17:61-70.[CrossRef][Medline]

31. Murphy, M., J. Ahn, K. K. Walker, W. H. Hoffman, R. M. Evans, A. J. Levine, and D. L. George. 1999. Transcriptional repression by wild-type p53 utilizes histone deacetylases, mediated by interaction with mSin3a. Genes Dev. 13:2490-2501.[Abstract/Free Full Text]

32. Neddermann, P., P. Gallinari, T. Lettieri, D. Schmid, O. Truong, J. J. Hsuan, K. Wiebauer, and J. Jiricny. 1996. Cloning and expression of human G/T mismatch-specific thymine-DNA glycosylase. J. Biol. Chem. 271:12767-12774.[Abstract/Free Full Text]

33. Neddermann, P., and J. Jiricny. 1993. The purification of a mismatch-specific thymine-DNA glycosylase from HeLa cells. J. Biol. Chem. 268:21218-21224.[Abstract/Free Full Text]

34. Nefkens, I., D. G. Negorev, A. M. Ishov, J. S. Michaelson, E. T. H. Yeh, R. M. Tanguay, W. E. G. Muller, and G. G. Maul. 2003. Heat shock and Cd2+ exposure regulate PML and Daxx release from ND10 by independent mechanisms that modify the induction of heat-shock proteins 70 and 25 differently. J. Cell Sci. 116:513-524.[Abstract/Free Full Text]

35. Ng, H.-H., and A. Bird. 1999. DNA methylation and chromatin modification. Curr. Opin. Gen. Dev. 9:158-163.[CrossRef][Medline]

36. Olivier, M., R. Eeles, M. Hollstein, M. A. Khan, C. C. Harris, and P. Hainaut. 2002. The IARC TP53 database: new online mutation analysis and recommendations to users. Hum. Mutat. 19:607-614.[CrossRef][Medline]

37. Park, J., T. Seo, H. Kim, and J. Choe. 2005. Sumoylation of the novel protein hRIPß is involved in replication protein A deposition in PML nuclear bodies. Mol. Cell. Biol. 25:8202-8214.[Abstract/Free Full Text]

38. Pearson, M., R. Carbone, C. Sebastiani, M. Cioce, M. Fagioli, I. Saito, Y. Higashimoto, E. M. Appella, S., P. P. Pandolfi, and P. G. Pelicci. 2000. PML regulates p53 acetylation and premature senescence induced by oncogenic Ras. Nature 406:207-210.[CrossRef][Medline]

39. Rodriguez, M. S., J. M. Desterro, S. Lain, C. A. Midgley, D. P. Lane, and R. T. Hay. 1999. SUMO-1 modification activates the transcriptional response of p53. EMBO J. 18:6455-6461.[CrossRef][Medline]

40. Salomoni, P., and P. P. Pandolfi. 2002. The role of PML in tumor suppression. Cell 108:165-170.[CrossRef][Medline]

41. Sancar, A., L. A. Lindsey-Boltz, K. Unsal-Kacmaz, and S. Linn. 2004. Molecular mechanisms of mammalian DNA repair and the DNA damage checkpoints. Annu. Rev. Biochem. 73:39-85.[CrossRef][Medline]

42. Song, J., L. K. Durrin, T. A. Wilkinson, T. G. Krontiris, and Y. Chen. 2004. Identification of a SUMO-binding motif that recognizes SUMO-modified proteins. Proc. Natl. Acad. Sci. USA 101:14373-14378.[Abstract/Free Full Text]

43. Song, J., Z. Zhang, W. Hu, and Y. Chen. 2005. Small ubiquitin-like modifier (SUMO) recognition of a SUMO binding motif: a reversal of the bound orientation. J. Biol. Chem. 280:40122-40129.[Abstract/Free Full Text]

44. Steinacher, R., and P. Schar. 2005. Functionality of human thymine DNA glycosylase requires SUMO-regulated changes in protein conformation. Curr. Biol. 15:616-623.[CrossRef][Medline]

45. Takahashi, H., S. Hatakeyama, H. Saitoh, and K. I. Nakayama. 2005. Noncovalent SUMO-1 binding activity of thymine DNA glycosylase (TDG) is required for its SUMO-1 modification and colocalization with the promyelocytic leukemia protein. J. Biol. Chem. 280:5611-5621.[Abstract/Free Full Text]

46. Tini, M., A. Benecke, S. J. Um, J. Torchia, R. M. Evans, and P. Chambon. 2002. Association of CBP/p300 acetylase and thymine DNA glycosylase links DNA repair and transcription. Mol. Cell 9:265-277.[CrossRef][Medline]

47. Um, S., M. Harbers, A. Benecke, B. Pierrat, R. Losson, and P. Chambon. 1998. Retinoic acid receptors interact physically and functionally with the T:G mismatch-specific thymine-DNA glycosylase. J. Biol. Chem. 273:20728-20736.[Abstract/Free Full Text]

48. Wang, J., C. Shiels, P. Sasieni, P. J. Wu, S. A. Islam, P. S. Freemont, and D. Sheer. 2004. Promyelocytic leukemia nuclear bodies associate with transcriptionally active genomic regions. J. Cell Biol. 164:515-526.[Abstract/Free Full Text]

49. Wilson, V. G., and D. Rangasamy. 2001. Intracellular targeting of proteins by sumoylation. Exp. Cell Res. 271:57-65.[CrossRef][Medline]

50. Zhong, S., P. Salomoni, and P. P. Pandolfi. 2000. The transcriptional role of PML and the nuclear body. Nat. Cell Biol. 2:E85-E90.