Summary DNA methylation is a major epigenetic mechanism for gene silencing. While methyltransferases mediate cytosine methylation, it is less clear how unmethylated regions in mammalian genomes are protected from de novo methylation and whether an active demethylating activity is involved. Here we show that either knockout or catalytic inactivation of the DNA repair enzyme Thymine DNA Glycosylase (TDG) leads to embryonic lethality in mice. TDG is necessary for recruiting p300 to retinoic acid (RA)-regulated promoters, protection of CpG islands from hypermethylation, and active demethylation of tissue-specific, developmentally- and hormonally-regulated promoters and enhancers. TDG interacts with the deaminase AID and the damage-response protein GADD45a. These findings highlight a dual role for TDG in promoting proper epigenetic states during development and suggest a two-step mechanism for DNA demethylation in mammals, whereby 5-methylcytosine and 5-hydroxymethylcytosine are first deaminated by AID to thymine and 5-hydroxymethyluracil, respectively, followed by TDG-mediated thymine and 5-hydroxymethyluracil excision repair.

Eukaryotic DNA polymerase beta (pol beta) can catalyze DNA synthesis during base excision DNA repair. It is shown here that pol beta also catalyzes release of 5'-terminal deoxyribose phosphate (dRP) residues from incised apurinic-apyrimidinic sites, which are common intermediate products in base excision repair. The catalytic domain for this activity resides within an amino-terminal 8-kilodalton fragment of pol beta, which comprises a distinct structural domain of the enzyme. Magnesium is required for the release of dRP from double-stranded DNA but not from a single-stranded oligonucleotide. Analysis of the released products indicates that the excision reaction occurs by beta-elimination rather than hydrolysis.

DNA damage frequently leads to the production of apurinic/apyrimidinic (AP) sites, which are presumed to be repaired through the base excision pathway. For detailed analyses of this repair mechanism, a synthetic analog of an AP site, 3-hydroxy-2-hydroxymethyltetrahydrofuran (tetrahydrofuran), has been employed in a model system. Tetrahydrofuran residues are efficiently repaired in a Xenopus laevis oocyte extract in which most repair events involve ATP-dependent incorporation of no more than four nucleotides (Y. Matsumoto and D. F. Bogenhagen, Mol. Cell. Biol. 9:3750-3757, 1989; Y. Matsumoto and D. F. Bogenhagen, Mol. Cell. Biol. 11:44414447, 1991). Using a series of column chromatography procedures to fractionate X. laevis ovarian extracts, we developed a reconstituted system of tetrahydrofuran repair with five fractions, three of which were purified to near homogeneity: proliferating cell nuclear antigen (PCNA), AP endonuclease, and DNA polymerase 8. This PCNA-dependent system repaired natural AP sites as well as tetrahydrofuran residues. DNA polymerase ,I was able to replace DNA polymerase b only for repair of natural AP sites in a reaction that did not require PCNA. DNA polymerase a did not support repair of either type of AP site. This result indicates that AP sites can be repaired by two distinct pathways, the PCNA-dependent pathway and the DNA polymerase ,8-dependent pathway.Base excision repair is a major pathway for repair of damaged bases in DNA (8). In this pathway, a DNA-Nglycosylase initiates repair by removing a specifically modified base, leaving an apurinic/apyrimidinic (AP) site. AP sites are also generated by spontaneous or induced base loss. Since AP sites arise so frequently, it is reasonable to expect that cells should have evolved very efficient mechanisms to repair this sort of damage.To study the mechanism of repair of AP sites, we have taken advantage of the ability to insert a single copy of a synthetic analog of an AP site, 3-hydroxy-2-hydroxymethyltetrahydrofuran (tetrahydrofuran), at a specific position in a covalently closed circular DNA (cccDNA) (17,23,29 (215) 728-4333. ever, several lines of evidence suggested that DNA polymerase ,B might not be the only polymerase involved in repair of AP sites. First, our early experiments showed that repair of the tetrahydrofuran residue was inhibited by aphidicolin, as expected if repair involves a high-molecular-weight DNA polymerase (unpublished data). Second, biochemical analyses of yeast mutants indicated that DNA polymerase E is responsible for this reaction (33). While a discrepancy concerning the identity of the repair polymerase may be the result of intrinsic differences in the experimental systems, some uncertainties could be eliminated if a reconstituted system which catalyzes the repair reaction with purified factors were available. In this paper, we present results obtained by fractionating the X. laevis ovarian extract, leading to development of a reconstituted system. This system consists of five fractions, including proli...

DNA ligase I belongs to a family of proteins that bind to proliferating cell nuclear antigen (PCNA) via a conserved 8-amino-acid motif [1]. Here we examine the biological significance of this interaction. Inactivation of the PCNA-binding site of DNA ligase I had no effect on its catalytic activity or its interaction with DNA polymerase beta. In contrast, the loss of PCNA binding severely compromised the ability of DNA ligase I to join Okazaki fragments. Thus, the interaction between PCNA and DNA ligase I is not only critical for the subnuclear targeting of the ligase, but also for coordination of the molecular transactions that occur during lagging-strand synthesis. A functional PCNA-binding site was also required for the ligase to complement hypersensitivity of the DNA ligase I mutant cell line 46BR.1G1 to monofunctional alkylating agents, indicating that a cytotoxic lesion is repaired by a PCNA-dependent DNA repair pathway. Extracts from 46BR.1G1 cells were defective in long-patch, but not short-patch, base-excision repair (BER). Our results show that the interaction between PCNA and DNA ligase I has a key role in long-patch BER and provide the first evidence for the biological significance of this repair mechanism.

The DNA mismatch repair (MMR) is a specialized system, highly conserved throughout evolution, involved in the maintenance of genomic integrity. To identify novel human genes that may function in MMR, we employed the yeast interaction trap. Using the MMR protein MLH1 as bait, we cloned MED1. The MED1 protein forms a complex with MLH1, binds to methyl-CpG-containing DNA, has homology to bacterial DNA repair glycosylases͞lyases, and displays endonuclease activity. Transfection of a MED1 mutant lacking the methyl-CpG-binding domain (MBD) is associated with microsatellite instability (MSI). These findings suggest that MED1 is a novel human DNA repair protein that may be involved in MMR and, as such, may be a candidate eukaryotic homologue of the bacterial MMR endonuclease, MutH. In addition, these results suggest that cytosine methylation may play a role in human DNA repair.