Role of the MUTYH protein in the response to oxidative damage to DNA

Abstract

The oxidized base 7,8-dihydro-8-hydroxyguanine (8-oxo-dG) is one of the most deleterious injuries induced by oxidative stress. Multiple DNA repair proteins have evolved to protect the genome against the detrimental effects of this promutagenic lesion. One of the major ones is the Base Excision Repair (BER) MUTYH DNA glycosylase that removes adenine from 8:oxoG-containing mispairs originated by DNA polymerases and via the OGG1 DNA glycosylase contributes to 8-oxodG repair. Germline mutations in the MUTYH gene lead to MUTYH-associated polyposis (MAP), an autosomal recessive syndrome characterized by colorectal polyposis and cancer predisposition. Although several reports characterized MUTYH variants using purified proteins, relatively few mutations have been investigated from the biochemical point of view. In addition no information is available on the mutator phenotype associated with MUTYH inactivation in humans. In our previous study accumulation of 8-oxo-dG and hypersensitivity to killing by the oxidizing agent KBrO3 were identified as a common phenotype among the investigated MAP-associated variants. These results were based on an assay in which single mutant MUTYH proteins were expressed in mouse embryonic fibroblasts (MEFs) derived from Mutyh-null mice (Molatore et al., 2010). This approach however cannot be exploited to analyze compound heterozygous MUTYH mutations, a very common situation among Italian MAP patients. With the aim of studying these particularly complex variants, we derived human lymphoblastoid cell lines (LCLs) from MAP patients harbouring missense and truncating mutations. RT-PCR and Western blotting analyses revealed that while in some cell lines there is a good correlation between transcripts and protein levels, in other instances no MUTYH protein is detectable. When basal levels of 8-oxo-dG were measured in these cell lines, increases were detected in DNA of six LCLs expressing MUTYH variants when compared to two wild-type cell lines. Interestingly the only two exceptions were cells in which no detectable expression of the MUTYH protein could be identified. To determine whether this increase in steady-state levels was due to a defective repair of 8-oxo-dG in these mutants, repair kinetics of this oxidized base were determined following exposure to KBrO3. All the LCLs harbouring MUTYH mutations showed a defective repair of 8-oxo-dG when compared to a cell line from a healthy donor. Results of a novel assay where both MUTYH and OGG1 activity could be evaluated indicate that all these variants were defective in removing adenine from an 8-oxoG:A DNA substrate, but retained wild-type OGG1 activity. Mutation frequency measurements at the PIG-A gene identified a four-fold increase in spontaneous mutagenesis in six LCLs from MAP patients when compared to three LCLs from healthy donors. Finally KBrO3 hypersensitivity was accompanied by a hypermutable phenotype in a MUTYH mutant cell line. These observations support the pathogenic role of these MUTYH mutations and identify accumulation of 8-oxo-dG and a mutator phenotype as relevant factors for a better clinical assessment of MUTYH variant pathogenesis. In the second part of this thesis the results of a parallel study on the effects of MUTYH loss in response to a different type of oxidative damage are reported. This study addresses the toxicity and carcinogenicity of the anti-cancer immunosuppressant drug Azathioprine (Aza) combined with UVA radiation. Systemic treatment with Aza causes the incorporation of 6-thioguanine (6-TG) into DNA. 6-TG is a chromophore which generates reactive oxygen species (ROS) on exposure to UVA and is itself highly susceptible to oxidation. DNA damage induced by the Aza/UVA combination is thought to contribute to the huge incidence of skin cancer in immunosuppressed organ transplant patients. Several studies have shown indeed that Aza combined with low doses of UVA may cause mutagenic damage in human cells. In particular, 6-TG/UVA generates a novel DNA lesion, the guanine-6-sulfonate that blocks DNA replication and is potentially mutagenic. Here we examined the role of the MUTYH DNA glycosylase and the mismatch repair (MMR) MSH2 protein in the cellular response to 6-TG/UVA-induced DNA damage. 6-TG and UVA were synergistically toxic to wild-type mouse embryo fibroblasts (MEFs) while neither 6-TG or UVA alone detectably affected survival. Mutyh- or Msh2-defective cells were more resistant than wild-type MEFs to killing by 6-TG/UVA. Nevertheless, the combined treatment significantly increased the levels of DNA 8-oxo-dG irrespectively of the genotype. Interestingly, we also found in wild-type cells that the deoxynucleoside triphosphates (dNTP) pool contributed to both the increased levels of DNA 8-oxo-dG and the enhanced toxicity of a combined 6-TG/UVA treatment. To better understand the mechanism of 6-TG/UVA toxicity and the relative role of the MUTYH protein we also analysed cell cycle progression by flow cytometry. The data suggest that the MUTYH protein is involved in the S phase arrest induced by the combined 6-TG/UVA treatment We also reported a difference between the two cell lines in the timing of the checkpoint activation: phosphorylation of the Chk1 protein occurred immediately after 6-TG/UVA treatment in wild-type cells, while a later and weaker appearance of phosphorylated Chk1 occurred in Mutyh-/- cells. These data suggest that MUTYH might be involved in the activation of an S phase checkpoint following this type of oxidative damage. We also confirmed that in this mouse model, as in human cells, treatment with 6-TG/UVA produces an increase in double strand breaks (DSBs). These DSBs, as measured by γH2AX foci, were formed immediately and maintained up to 48 hr from the end of the treatment. The comparable increase in DSBs observed in Mutyh-defective cells indicates that the initial level of DNA damage is similar in the two genotypes. However analysis of the homologous recombination (HR) protein RAD51 foci following the combined treatment indicates that the DSBs resolution differs between the two genotypes. Mutyh-/-MEFs showed a significantly lower induction of RAD51 foci in comparison to wild-type cells with a gradual decrease in the levels of foci at late post-treatment times. This contrasts with the persistence in wild-type cells of high levels of these foci indicating the presence of unresolved DSBs. Finally we report the results of in vivo experiments in which the long-term toxicity of single or combined Aza/UVA exposures was analysed in wild-type and Mutyh-defective animals. Mice were treated with Aza given by intraperitoneal injection and/or UVA for 12 months. A differential toxicity between WT and Mutyh-defective animals was observed as consequence of the Aza plus UVA exposure. In fact, a high level of toxicity was identified in the group of wild-type mice (only 20% of the animals survived this exposure), while the 80% of Mutyh-/- animals survived this treatment. Survival was 100% in both UVA treated groups, while immunosuppression in Azatreated groups was associated with some mortality, which was unaffected by the genotype. A significant increase of DNA oxidation in the skin of animals exposed to the combined treatment was also observed, irrespectively of the genotype. Intriguingly when histopathological examination of the skin was performed two squamous cell carcinomas were identified only in Mutyh-/- mice exposed to Aza plus UVA, revealing a possible skin cancer proneness conferred by loss of this protein.