Construction of a functional cDNA clone of the hamster ERCC2 DNA repair and transcription gene

APLF (C2orf13) is a novel human protein involved in the cellular response to chromosomal DNA strand breaks

Aprataxin and polynucleotide kinase (PNK) are DNA end processing factors that are recruited into the DNA single- and double-strand break repair machinery through phosphorylation-specific interactions with XRCC1 and XRCC4, respectively. These interactions are mediated through a divergent class of forkhead-associated (FHA) domain that binds to peptide sequences in XRCC1 and XRCC4 that are phosphorylated by casein kinase 2 (CK2). Here, we identify the product of the uncharacterized open reading frame C2orf13 as a novel member of this FHA domain family of proteins and we denote this protein APLF (aprataxin- and PNK-like factor). We show that APLF interacts with XRCC1 in vivo and in vitro in a manner that is stimulated by CK2. Yeast two-hybrid analyses suggest that APLF also interacts with the double-strand break repair proteins XRCC4 and XRCC5 (Ku86). We also show that endogenous and yellow fluorescent protein-tagged APLF accumulates at sites of H(2)O(2) or UVA laser-induced chromosomal DNA damage and that this is achieved through at least two mechanisms: one that requires the FHA domain-mediated interaction with XRCC1 and a second that is independent of XRCC1 but requires a novel type of zinc finger motif located at the C terminus of APLF. Finally, we demonstrate that APLF is phosphorylated in a DNA damage- and ATM-dependent manner and that the depletion of APLF from noncycling human SH-SY5Y neuroblastoma cells reduces rates of chromosomal DNA strand break repair following ionizing radiation. These data identify APLF as a novel component of the cellular response to DNA strand breaks in human cells.

The Protein Kinase CK2 Facilitates Repair of Chromosomal DNA Single-Strand Breaks

CK2 was the first protein kinase identified and is required for the proliferation and survival of mammalian cells. Here, we have identified an unanticipated role for CK2. We show that this essential protein kinase phosphorylates the scaffold protein XRCC1 and thereby enables the assembly and activity of DNA single-strand break repair protein complexes in vitro and at sites of chromosomal breakage. Moreover, we show that inhibiting XRCC1 phosphorylation by mutation of the CK2 phosphorylation sites or preventing CK2 activity using a highly specific inhibitor ablates the rapid repair of cellular DNA single-strand breaks by XRCC1. These data identify a direct role for CK2 in the repair of chromosomal DNA strand breaks and in maintaining genetic integrity.

A requirement for PARP-1 for the assembly or stability of XRCC1 nuclear foci at sites of oxidative DNA damage

The molecular role of poly (ADP-ribose) polymerase-1 in DNA repair is unclear. Here, we show that the single-strand break repair protein XRCC1 is rapidly assembled into discrete nuclear foci after oxidative DNA damage at sites of poly (ADP-ribose) synthesis. Poly (ADP-ribose) synthesis peaks during a 10 min treatment with H2O2 and the appearance of XRCC1 foci peaks shortly afterwards. Both sites of poly (ADP-ribose) and XRCC1 foci decrease to background levels during subsequent incubation in drug-free medium, consistent with the rapidity of the single-strand break repair process. The formation of XRCC1 foci at sites of poly (ADP-ribose) was greatly reduced by mutation of the XRCC1 BRCT I domain that physically interacts with PARP-1. Moreover, we failed to detect XRCC1 foci in Adprt1-/- MEFs after treatment with H2O2. These data demonstrate that PARP-1 is required for the assembly or stability of XRCC1 nuclear foci after oxidative DNA damage and suggest that the formation of these foci is mediated via interaction with poly (ADP-ribose). These results support a model in which the rapid activation of PARP-1 at sites of DNA strand breakage facilitates DNA repair by recruiting the molecular scaffold protein, XRCC1.

Development and characterization of CHO repair-proficient cell lines for comparative mutagenicity and metabolism of heterocyclic amines from cooked food

In order to understand the role of repair and metabolism in the mutagenicity of heterocyclic amines from cooked foods, we previously developed the nucleotide excision repair-deficient CHO 5P3NAT2 cell line engineered to coexpress the mouse CYP1A2 and human N-acetyltransferase genes. In the present study, we have made a matched repair-competent cell line by mutagenizing 5P3NAT2 cells with ethyl methanesulfonate and selecting for resistance to cytotoxicity by 2-amino-3-methylimidazo[4,5-f]quinoline (IQ). In the differential cytotoxicity (DC) assay, 4 out of 15 clones showed no cytotoxic effect with IQ at the highest dose (30 microg/ml) tested, in contrast to repair-deficient 5P3NAT2 cells, which showed approximately 100% cytotoxicity at 0.3 microg/ml. Subsequently, these IQ-resistant clones were examined for resistance to killing by UV irradiation. All four IQ-resistant clones, which show resistance to UV similar to that of repair-proficient AA8 cells, still express both the CYP1A2 and N-acetyltransferase genes. Sequence analysis of CXPD cDNA from the 5P3NAT2R9 clone revealed an A:T-->G:C reversion event at the site of the UV5 mutation. This base change results in reversion of the codon 116 tyrosine in UV5 cells back to the original cysteine in AA8 cells, thereby restoring wild-type CXPD activity and repair function. In contrast to 5P3NAT2 cells, the repair-proficient 5P3NAT2R9 revertant cell line shows little IQ-induced cell killing, and dramatically lower levels of induced mutation at the adenine phosphoribosyltransferase (Aprt) gene locus over the range of 2-40 microg/ml IQ. This matched pair of repair-proficient/deficient cell lines can provide insight not only into the genotoxicity of heterocyclic amine dietary carcinogens such as IQ and PhIP, but also into the effects of nucleotide excision repair on the ultimate mutagenicity of these compounds.

Restoration of nucleotide excision repair in a helicase-deficient XPD mutant from intragenic suppression by a trichothiodystrophy mutation

The UV-sensitive V-H1 cell line has a T46I substitution mutation in the Walker A box in both alleles of XPD and lacks DNA helicase activity. We characterized three partial revertants that curiously display intermediate UV cytotoxicity (2- to 2.5-fold) but normal levels of UV-induced hprt mutations. In revertant RH1-26, the efficient removal of pyrimidine (6-4) pyrimidone photoproducts from both strands of hprt suggests that global-genomic nucleotide excision repair is normal, but the pattern of cyclobutane pyrimidine dimer removal suggests that transcription-coupled repair (TCR) is impaired. To explain the intermediate UV survival and lack of RNA synthesis recovery in RH1-26 after 10 J of UV/m(2), we propose a defect in repair-transcription coupling, i.e., the inability of the cells to resume or reinitiate transcription after the first TCR event within a transcript. All three revertants carry an R658H suppressor mutation, in one allele of revertants RH1-26 and RH1-53 and in both alleles of revertant RH1-3. Remarkably, the R658H mutation produces the clinical phenotype of trichothiodystrophy (TTD) in several patients who display intermediate UV sensitivity. The XPD(R658H) TTD protein, like XPD(T46I/R658H), is codominant when overexpressed in V-H1 cells and partially complements their UV sensitivity. Thus, the suppressing R658H substitution must restore helicase activity to the inactive XPD(T46I) protein. Based on current knowledge of helicase structure, the intragenic reversion mutation may partially compensate for the T46I mutation by perturbing the XPD structure in a way that counteracts the effect of this mutation. These findings have implications for understanding the differences between xeroderma pigmentosum and TTD and illustrate the value of suppressor genetics for studying helicase structure-function relationships.

Codominance associated with overexpression of certain XPD mutations

Mutations in the XPD gene are associated with three complex clinical phenotypes, namely xeroderma pigmentosum (XP), XP in combination with Cockayne syndrome (XP-CS), and trichothiodystrophy (TTD). XP is caused by a deficiency in nucleotide excision repair (NER) that results in a high risk of skin cancer. TTD is characterized by severe developmental and neurological defects, with hallmark features of brittle hair and scaly skin, and sometimes has defective NER. We used CHO cells as a system to study how specific mutations alter the dominant/recessive behavior of XPD protein. Previously we identified the T46I and R75W mutations in two highly UV-sensitive hamster cell lines that were reported to have paradoxically high levels of unscheduled DNA synthesis. Here we report that these mutants have greatly reduced XPD helicase activity and fully defective NER in a cell-extract excision assay. We conclude that the unscheduled DNA synthesis seen in these mutants is caused by abortive "repair" that does not contribute to cell survival. These mutations, as well as the K48R canonical helicase-domain mutation, each produced codominant negative phenotypes when overexpressed in wild-type CHO cells. The common XP-specific R683W mutation also behaved in a codominant manner when overexpressed, which is consistent with the idea that this mutation may affect primarily the enzymatic activity of the protein rather than impairing protein interactions, which may underlie TTD. A C-terminal mutation uniquely found in TTD (R722W) was overexpressed but not to levels sufficiently high to rigorously test for a codominant phenotype. Overexpression of mutant XPD alleles may provide a simple means of producing NER deficiency in other cell lines.

A cell cycle-specific requirement for the XRCC1 BRCT II domain during mammalian DNA strand break repair

XRCC1 protein is essential for viability in mammals and is required for efficient DNA single-strand break repair and genetic stability following DNA base damage. We report here that XRCC1-dependent strand break repair in G(1) phase of the cell cycle is abolished by mutations created within the XRCC1 BRCT domain that interact with DNA ligase III. In contrast, XRCC1-dependent DNA strand break repair in S phase is largely unaffected by these mutations. These data describe a cell cycle-specific role for a BRCT domain, and we conclude that the XRCC1-DNA ligase III complex is required for DNA strand break repair in G(1) phase of the cell cycle but is dispensable for this process in S phase. The S-phase DNA repair process can remove both strand breaks induced in S phase and those that persist from G(1) and can in part compensate for lack of repair in G(1). This process correlates with the appearance of XRCC1 nuclear foci that colocalize with Rad51 and may thus function in concert with homologous recombination.

Molecular analysis of ERCC2 mutations in the repair deficient hamster mutants UVL-1 and V-H1

The cDNA sequence of the Chinese hamster ERCC2 nucleotide excision repair and transcription gene from the UVL-1 Chinese hamster ovary (CHO) mutant cell line and the V-H1 Chinese hamster V79 mutant line was analyzed. ERCC2 encodes a presumed ATP-dependent DNA helicase and is single copy in CHO lines due to the structural hemizygosity of chromosome 9. Both UVL-1 and V-H1 have intermediate levels of (6-4) photoproduct repair but are as highly UV sensitive as the group 2 mutants that have no detectable repair. Deficiency in cyclobutane dimer removal has also been shown for V-H1. In UVL-1, a single base substitution resulting in an Arg75-->Trp substitution in helicase domain Ia was identified. The equivalent amino acid position is also Arg in the human, mouse, Xiphophorus maculatus, Saccharomyces cerevisiae, and Schizosaccharomyces pombe homologs. In V-H1, a single base substitution resulting in a Thr46-->Ile substitution in helicase domain I (the ATP-binding domain) was identified in both alleles. The equivalent amino acid position is also Thr in the five homologs. Analysis of three V-H1 partial revertants revealed that they still have the original V-H1 mutation in both alleles, indicating that these are second site reversion events. Site-specific mutagenesis was used to introduce the Thr46-->Ile, Arg75-->Trp, and Lys48-->Arg (helicase domain I) mutations into a hamster ERCC2 expression plasmid. These plasmids each failed to confer UV resistance to group 2 mutant cells, further demonstrating that the changes identified are the causative mutations in V-H1 and UVL-1. Correlations between specific mutations, biochemical activities, and repair phenotype are discussed.