DNA damage in human whole blood caused by radiopharmaceuticals evaluated by the comet assay.

Radiopharmaceuticals used for diagnosis or therapy induce DNA strand breaks, which may be detectable by single-cell gel electrophoresis (called comet assay). Blood was taken from patients before and at different time points after treatment with radiopharmaceuticals; blood cells were investigated by the comet assay using the percentage of DNA in the tail as the critical parameter. Whereas [225Ac]Ac-prostate-specific membrane antigen (PSMA)-617 alpha therapy showed no difference relative to the blood sample taken before treatment, beta therapy with [177Lu]Lu-PSMA-617 3 h post-injection revealed a small but significant increase in DNA strand breaks. In blood of patients who underwent positron emission tomography (PET) with either [18F]2-fluor-2-deoxy-D-glucose (FDG) or [68Ga]Ga-PSMA-11, an increase of DNA migration determined by the comet assay was not found when analysed at different time points (2-70 min) after intravenous tracer injection. Human whole blood was incubated with the targeted clinically relevant therapeutic radiopharmaceuticals [225Ac]Ac-PSMA-617, [177Lu]Lu-PSMA-617 and [90Y]Y-DOTA(0)-Phe(1)-Tyr(3)-octreotide (DOTA-TOC) at different activity concentrations (kBq/ml) for 5 days and then analysed by the comet assay. DNA damage increased with higher concentrations of all radiolabeled compounds tested. [177Lu]Lu-PSMA-617 caused higher blood cell radiotoxicity than equal activity concentrations of [90Y]Y-DOTA-TOC. Likewise, whole human blood was exposed to the positron emitters [18F]FDG and [68Ga]Ga-PSMA-11 in vitro for 24 h with activity concentrations ranging between 5 and 40 MBq/ml. The same activity concentration dependent elevated DNA migration was observed for both compounds although decay energies are different. This study demonstrated that the amount of DNA damage detected by the comet assay in whole human blood is similar among different positron emitters and divergent by a factor of 200 between alpha particles and beta radiation.

Modelling mutational landscapes of human cancers in vitro.

Experimental models that recapitulate mutational landscapes of human cancers are needed to decipher the rapidly expanding data on human somatic mutations. We demonstrate that mutation patterns in immortalised cell lines derived from primary murine embryonic fibroblasts (MEFs) exposed in vitro to carcinogens recapitulate key features of mutational signatures observed in human cancers. In experiments with several cancer-causing agents we obtained high genome-wide concordance between human tumour mutation data and in vitro data with respect to predominant substitution types, strand bias and sequence context. Moreover, we found signature mutations in well-studied human cancer driver genes. To explore endogenous mutagenesis, we used MEFs ectopically expressing activation-induced cytidine deaminase (AID) and observed an excess of AID signature mutations in immortalised cell lines compared to their non-transgenic counterparts. MEF immortalisation is thus a simple and powerful strategy for modelling cancer mutation landscapes that facilitates the interpretation of human tumour genome-wide sequencing data.

p53 mutations in benzo(a)pyrene-exposed human p53 knock-in murine fibroblasts correlate with p53 mutations in human lung tumors.

Human p53 mutation spectra differ significantly from one cancer type to another. One possible reason is that carcinogenic risk factors differ, and these factors elicit distinct mutation patterns. There has been no mammalian assay, however, with which to generate mutation patterns in human p53 sequences experimentally, hampering interpretation of the human tumor spectra. We have designed a new mammalian cell assay using gene targeting technology that selects and scores human p53 gene sequence mutations in human-p53 knock-in (Hupki) murine embryonic fibroblasts (HUF) that have undergone immortalization. With the Hupki assay we examined here whether benzo(a)pyrene (BaP), a major tobacco smoke carcinogen could elicit p53 mutation patterns characterizing the human lung tumor p53 mutation spectrum. We found that, in contrast to unexposed HUFs or HUFs exposed to other carcinogenic agents, HUFs exposed to BaP acquire mutations that display major features of the human lung tumor p53 mutation spectrum: (a) predominance of G-to-T mutations, (b) unequivocal strand bias of the transversions, and (c) a mutation hotspot at codons 157 to 158. These data are consistent with the hypothesis that BaP has a direct role in causing smokers' lung tumor p53 mutations. The assay can be used to examine various hypotheses on the endogenous or exogenous factors responsible for the p53 mutations in human tumors arising in other tissues.