A method to detect DNA glycosylase activity is described. The substrate used was an oligodeoxyribonucleotide with a unique hypoxanthine base, but has general application to any DNA glycosylase or endonuclease. The oligodeoxyribonucleotide was labeled at the 5' end with 32P and at the 3' end with a biotin linkage and annealed to a complementary oligodeoxyribonucleotide. The hypoxanthine base was excised in solution using the MPG protein, a human DNA glycosylase. Following cleavage of the phosphodiester linkage by NaOH, the oligodeoxyribonucleotide was attached to streptavidin-coated, paramagnetic beads. Binding of the labeled oligodeoxyribonucleotide to the beads was indicative of the kinetics of the reaction. As a control, half of the reaction products were loaded on to a denaturing polyacrylamide gel. Comparable values for steady-state kinetic constants were obtained using both methods. This nonelectrophoretic technique is a rapid assay of DNA glycosylase activity for both purified proteins and crude extracts. This method can be directly adapted for high-throughput techniques.

The progression of a normal cell to senescence in vivo and in vitro is accompanied by a reduction in the length of the telomeres, the chromosome capping segments at the end of each linkage group. However, overexpression of the reverse transcriptase subunit (HTERT) of the ribonucleoprotein telomerase restores telomere length and delays cellular senescence. Although some data exist in the literature with respect to survival, no molecular data have shown that DNA repair in telomerase-immortalized cells is normal. Several telomerase-immortalized human skin fibroblast cell lines were constructed from a primary human fibroblast cell line. The primary line and the telomerase-immortalized cell lines were treated with either ultraviolet (UV) radiation or dimethylsulfate (DMS). UV radiation principally produces cyclobutane pyrimidine dimers that are repaired by nucleotide excision repair, whereas DMS introduces mainly N-methylpurines repaired by base excision repair. Here, we show that repair of both types of damage in the telomerase-immortalized human skin fibroblast cell lines is identical to repair observed in normal skin fibroblasts. Thus, telomerase expression and consequent immortalization of skin fibroblasts do not alter nucleotide or base excision repair in human cells.

Human 3-methyladenine-DNA glycosylase (MPG protein) is involved in the base excision repair (BER) pathway responsible mainly for the repair of small DNA base modifications. It initiates BER by recognizing DNA adducts and cleaving the glycosylic bond leaving an abasic site. Here, we explore several of the factors that could influence excision of adducts recognized by MPG, including sequence context, effect of APE1, and interaction with other proteins. To investigate sequence context, we used 13 different 25 bp oligodeoxyribonucleotides containing a unique hypoxanthine residue (Hx) and show that the steady-state specificity of Hx excision by MPG varied by 17-fold. If APE1 protein is used in the reaction for Hx removal by MPG, the steady-state kinetic parameters increase by between fivefold and 27-fold, depending on the oligodeoxyribonucleotide. Since MPG has a role in removing adducts such as 3-methyladenine that block DNA synthesis and there is a potential sequence for proliferating cell nuclear antigen (PCNA) interaction, we hypothesized that MPG protein could interact with PCNA, a protein involved in repair and replication. We demonstrate that PCNA associates with MPG using immunoprecipitation with either purified proteins or whole cell extracts. Moreover, PCNA binds to both APE1 and MPG at different sites, and loading PCNA onto a nicked, closed circular substrate with a unique Hx residue enhances MPG catalyzed excision. These data are consistent with an interaction that facilitates repair by MPG or APE1 by association with PCNA. Thus, PCNA could have a role in short-patch BER as well as in long-patch BER. Overall, the data reported here show how multiple factors contribute to the activity of MPG in cells.

DNA-protein cross-links (DPCs) present a formidable obstacle to cellular processes because they are "superbulky" compared with the majority of chemical adducts. Elimination of DPCs is critical for cell survival because their persistence can lead to cell death or halt cell cycle progression by impeding DNA and RNA synthesis. To study DPC repair, we have used DNA methyltransferases to generate unique DPC adducts in oligodeoxyribonucleotides or plasmids to monitor both in vitro excision and in vivo repair. We show that HhaI DNA methyltransferase covalently bound to an oligodeoxyribonucleotide is not efficiently excised by using mammalian cell-free extracts, but protease digestion of the full-length HhaI DNA methyltransferase-DPC yields a substrate that is efficiently removed by a process similar to nucleotide excision repair (NER). To examine the repair of that unique DPC, we have developed two plasmid-based in vivo assays for DPC repair. One assay shows that in nontranscribed regions, DPC repair is greater than 60% in 6 h. The other assay based on host cell reactivation using a green fluorescent protein demonstrates that DPCs in transcribed genes are also repaired. Using Xpg-deficient cells (NER-defective) with the in vivo host cell reactivation assay and a unique DPC indicates that NER has a role in the repair of this adduct. We also demonstrate a role for the 26 S proteasome in DPC repair. These data are consistent with a model for repair in which the polypeptide chain of a DPC is first reduced by proteolysis prior to NER.