XPA versus ERCC1 as chemosensitising agents to cisplatin and mitomycin C in prostate cancer cells: role of ERCC1 in homologous recombination repair.

Nucleotide excision repair is the principal mechanism for the removal of bulky DNA adducts caused by a range of chemotherapeutic drugs, and contributes to cisplatin resistance. In this study, we used synthetic siRNAs targeted to XPA and ERCC1 and compared their effectiveness in sensitising mismatch repair deficient prostate cancer cell lines to cisplatin and mitomycin C. Downregulation of ERCC1 sensitised DU145 and PC3 cells to cisplatin and mitomycin C. In contrast, XPA downregulation did not sensitise either cell line to mitomycin C, and only sensitised DU145 cells to cisplatin. The effects of ERCC1 downregulation may be due to its role in homologous recombination repair. Excision repair of cisplatin adducts in PC3 cells was attenuated to a similar extent by XPA and ERCC1 downregulation. Downregulation of XPA but not ERCC1 caused an increase in the number of cisplatin-induced RAD51 foci in PC3 cells, suggesting that HRR is able to substitute for NER in these cells. We observed co-localisation of ERCC1 and RAD51 in cisplatin treated PC3 cells by immunofluorescence and co-immunoprecipitation, which may represent recruitment of ERCC1/XPF to sites of recombination repair. These results indicate that ERCC1 is a broader therapeutic target than XPA with which to sensitise cancer cells to chemotherapy because of its additional role in recombination repair.

Regulation of DNA repair gene expression in human cancer cell lines.

Although most advanced cancers are incurable, the majority of testicular germ cell tumors can be cured using cisplatin-based combination chemotherapy. The nucleotide excision repair (NER) pathway removes most DNA adducts produced by cisplatin, and the low levels of NER in testis tumor cells may explain why these cancers are curable. Three NER proteins: ERCC1, XPF, and XPA, are present at low levels in testis tumor cell lines, and addition of these proteins to protein extracts of testis tumor cells increases their in vitro DNA repair capacity to normal levels. The aim of this study was to identify the mechanism responsible for the low levels of these DNA repair proteins. The levels of the mRNA transcripts for ERCC1, XPF, and XPA were measured in a panel of 14 different human cancer cell lines, using real-time PCR. Three ERCC1 splice variants were identified and quantitated. Three alternative transcription start points (TSPs) were identified for ERCC1 but none were testis-specific. The significantly lower levels of ERCC1, XPF, and XPA protein in testis tumor cell lines cannot be explained solely by differences in transcriptional efficiency or mRNA stability. For ERCC1, post-transcriptional control by alternative splicing does not account for the testis-specific low levels of protein expression. Pulse-chase experiments showed that the half-life of ERCC1 protein in a testis tumor cell line was not significantly different to that in a prostate cancer cell line. Taken together, these results suggest that constitutive levels of these DNA repair proteins are controlled at the level of translation.

Sequence-selective interaction of the minor-groove interstrand cross-linking agent SJG-136 with naked and cellular DNA: footprinting and enzyme inhibition studies.

SJG-136 (3) is a novel pyrrolobenzodiazepine (PBD) dimer that is predicted from molecular models to bind in the minor groove of DNA and to form sequence-selective interstrand cross-links at 5'-Pu-GATC-Py-3' (Pu = purine; Py = pyrimidine) sites through covalent bonding between each PBD unit and guanines on opposing strands. Footprinting studies have confirmed that high-affinity adducts do form at 5'-G-GATC-C-3' sequences and that these can inhibit RNA polymerase in a sequence-selective manner. At higher concentrations of SJG-136, bands that migrate more slowly than one of the 5'-G-GATC-C-3' footprint sites show significantly reduced intensity, concomitant with the appearance of higher molecular weight material near the gel origin. This phenomenon is attributed to interstrand cross-linking at the 5'-G-GATC-C-3' site and is the first report of DNA footprinting being used to detect interstrand cross-linked adducts. The control dimer GD113 (4), of similar structure to SJG-136 but unable to cross-link DNA due to its C7/C7'-linkage rather than C8/C8'-linkage, neither produces footprints with the same DNA sequence nor blocks transcription at comparable concentrations. In addition to the two high-affinity 5'-G-GATC-C-3' footprints on the MS2 DNA sequence, other SJG-136 adducts of lower affinity are observed that can still block transcription but with lower efficiency. All these sites contain the 5'-GXXC-3' motif (where XX includes AG, TA, GC, CT, TT, GG, and TC) and represent less-favored cross-link sites. In time-course experiments, SJG-136 blocks transcription if incubated with a double-stranded DNA template before the transcription components are added; addition after transcription is initiated fails to elicit blockage. Single-strand ligation PCR studies on a sequence from the c-jun gene show that SJG-136 binds to 5'-GAAC-3'/5'-GTTC-3' (preferred) or 5'-GAGC-3'/5'-GCTC-3' sequences. Significantly, adducts are obtained at the same sequences following extraction of DNA from drug-treated K562 cells, confirming that the agent reaches the cellular genome and interacts with the DNA in a sequence-selective fashion. Finally, SJG-136 efficiently inhibits the action of restriction endonuclease BglII, which has a 5'-A-GATC-T-3' motif at its cleavage site.

PCR-based methods for detecting DNA damage and its repair at the sub-gene and single nucleotide levels in cells.

Three PCR-based methods are described that allow covalent drug-DNA adducts, and their repair, to be studied at various levels of resolution from gene regions to the individual nucleotide level in single copy genes. A quantitative PCR (QPCR) method measures the total damage on both DNA strands in a gene region, usually between 300 and 3,000 base pairs in length. Strand-specific QPCR incorporates adaptations that allow damage to be measured in the same region as QPCR but in a strand-specific manner. Single-strand ligation PCR allows the detection of adduct formation at the level of single nucleotides, on individual strands, in a single copy gene in mammalian cells. If antibodies to the DNA adducts of interest are available, these can be used to capture and isolate adducted DNA for use in single-strand ligation PCR increasing the sensitivity of the assay.