Abstract
Purpose : This study was performed to standardize experimental conditions for the quantification by pulsed-field gel electrophoresis (PFGE) of radiation-induced DNA double-strand breaks (dsb) and rejoining in a human malignant brain tumour xenograft model. Materials and methods : Because no correlation was found between DNA dsb induction or rejoining and clinical radiation response for six fresh glioblastoma (GBM) specimens, assay conditions were examined in detail. SF-767 human GBM cells were implanted into the flanks of athymic mice. Resulting tumours were irradiated in vivo, dissociated mechanically or using an enzyme cocktail, and assayed for DNA dsb induction and repair. In other experiments, excised tumour portions were irradiated and allowed to repair either as chunks (50mm3 pieces), as minced tumour (1mm3 pieces), or as single-cell suspensions. Finally, the effect of holding excised tumours in vitro for times of up to 72h before irradiation and assay for DNA dsb and cell survival was studied. Results and conclusions : The method of tumour dissociation had no effect on results; however, both the configuration of specimens during irradiation and the in vitro maintenance time markedly affected the experimental outcome. Chunks irradiated in vitro had DNA dsb results that were very similar to those obtained when tumours were irradiated in situ, while minced pieces or single-cell suspensions resulted in steeper dose-response curves. When tumour chunks were maintained at 4 C in medium, DNA dsb induction was not affected for 24h and DNA dsb rejoining remained constant for 48h but then decreased. Cell survival, however, decreased continually during the 72h in vitro maintenance time.