br Materials and methods br Results br Discussion In this

Materials and methods
Results
Discussion In this study, the supF forward mutation assay showed that 5OHU induced mutations in human cells and that a C→T mutation was the predominant mutation among 5OHU-induced mutations. In addition, the DNA cleavage activity assay revealed that 5OHU was excised from DNA by NTHL1 as well as four other DNA glycosylases (SMUG1, NEIL1, TDG, and UNG2). When human cells overexpressing each of these five DNA glycosylases were established, all of the five DNA glycosylases, including NTHL1, were found to show the ability to suppress 5OHU-induced mutations. Based on the above results, we next performed functional evaluation of eight NTHL1 variants using 5OHU-containing DNA material to contribute to accurate diagnosis of NAP. The DNA cleavage activity assay showed that the NTHL1 variants Q90X, Y130X, R153X, and Q287X, but not R19Q, V179I, V217F, or G286S, showed defective ability for the repair of the DNA substrates containing 5OHU as well as two other damaged bases. Moreover, the supF forward mutation assay showed that ability to suppress 5OHU-induced mutations in human cells was reduced in all the four truncated-type NTHL1 variants. Thus, our results suggested that the NTHL1 variants Q90X, Y130X, R153X, and Q287X, but not R19Q, V179I, V217F, or G286S, showed defective ability for 5OHU repair and that the NTHL1 Valproic acid encoding the truncated variants would be pathogenic for NAP, revealing a new and important link between oxidative damage repair and hereditary cancer. In the present study, the NTHL1 variants Q90X, Y130X, R153X, and Q287X showed reduced DNA glycosylase activity for 5OHU in the DNA cleavage activity assay and a reduced ability to suppress 5OHU-induced mutations in the supF forward mutation assay. Besides 5OHU, the repair activities of these four NTHL1 variants for DHU and Tg were also reduced. Therefore, we judged that Q90X, Y130X, R153X, and Q287X are functionally deleterious. According to previous reports, Q90X and Q287X-type NTHL1 have been found in NAP families [4], [5], [6], [41], and it was experimentally shown that NTHL1 mRNA expression was reduced in homozygous Q90X carriers, probably due to nonsense-mediated mRNA decay (NMD) [4]. Thus, based on its identification in NAP families and the reduced mRNA expression levels shown in previous reports [4], [6], and the reduced repair activity level shown in the present study, Q90X is considered to be a pathogenic allele for NAP. In regard to the other three variants (Y130X, R153X, and Q287X), it has not been experimentally proven whether expression of the NTHL1 variants is down-regulated or not by NMD. Since papers reporting exception to the rules governing susceptibility to NMD have been accumulated [42], [43], at present, the expression statuses of the NTHL1 variants Y130X, R153X, and Q287X are unclear. Thus, our results of the functional analysis and a previous report on identification of Q287X in an NAP family [6] strongly suggest that the Q287X allele is pathogenic. With regard to Y130X and R153X, they have not been found in NAP families. This might be due to the relatively low allelic frequency of Y130X and R153X; their allelic frequencies are approximately one-quarter of the allelic frequency of Q287X, according to the data obtained in the Grand Opportunity Exome Sequencing Project (GO-ESP), which are listed in the dbSNP homepage. Besides GO-ESP, Y130X and R153X alleles have also been detected in two other projects [Exome Aggregation Consortium (ExAC) and Trans-Omics for Precision Medicine (TOPMed)] according to the dbSNP homepage, indicating that the alleles do not represent sequencing error. Since our study showed that Y130X and R153X showed defective repair capacity for oxidatively damaged bases like Q90X and Q287X, individuals with homozygous NTHL1 mutations having a combination of Q90X, Y130X, R153X, and Q287X alleles would be susceptible to NAP.