R.E. Meyn

In this study we used HeLa cells transfected with a conditional Bcl-2 expression construct to study the effects of Bcl-2 on reduced glutathione (GSH) metabolism. Our previous work demonstrated that depletion of GSH by culturing cells in tissue culture medium lacking the amino acids cysteine and methionine, essential for GSH biosynthesis, caused cells overexpressing Bcl-2 to become sensitized to apoptotic induction. Here we report that Bcl-2 also dramatically alters GSH compartmentalization. Cellular distribution of GSH, assayed by confocal microscopy, revealed that when Bcl-2 expression was suppressed GSH was uniformly distributed primarily in the cytosol, whereas overexpression of Bcl-2 led to a relocalization of GSH into the nucleus. Isolated nuclei readily accumulated radiolabeled GSH and maintained higher nuclear GSH concentration in direct relation to Bcl-2 nuclear protein levels. Moreover, exogenous GSH blocked apoptotic changes and caspase activity in isolated nuclei exposed to the pro-apoptotic protease granzyme B. Our results indicate that one of the functions of Bcl-2 is to promote sequestration of GSH into the nucleus, thereby altering nuclear redox and blocking caspase activity as well as other nuclear alterations characteristic of apoptosis. We speculate that this mechanism contributes to the suppression of apoptosis in cells with elevated Bcl-2 levels.

Fifteen different murine tumours were evaluated with respect to the degree of apoptosis development that occurs in the tumour tissue in the first few hours following irradiation in vivo. Animals were killed at 3 or 6 h following irradiation with 0, 2.5, 10 or 25 Gy. Apoptosis was scored as percent aberrant nuclei by microscopic examination of histological sections made from the tumour specimens. Results showed that three of four mammary adenocarcinomas, one ovarian adenocarcinoma, and one lymphoma displayed at least 10% apoptotic cells after 25 Gy, whereas five sarcomas, three squamous cell carcinomas, and a hepatocarcinoma did not. The time courses and dose responses were similar in those tumours that responded. These data were compared with the known response of these same tumours when analysed using conventional assays. The tumours that did respond by significant apoptosis had longer specific growth delays and lower TCD50 (dose to cure 50% of animals) doses, thus suggesting that an acute apoptotic response following irradiation may be a feature of certain tumours that respond well to irradiation. Additionally, this analysis revealed heterogeneity in the apoptotic response both within an individual tumour specimen and among different tumour types. These observations of intra and intertumour heterogeneity are consistent with the idea that the propensity for apoptosis in tumours is genetically regulated.

Early radiation responses of transplantable murine ovarian (OCaI) and hepatocellular (HCaI) carcinomas were examined at 6, 24, 48, 96, and 144 h after single photon doses of 25, 35, or 45 Gy. Previous studies using tumor growth delay and tumor radiocurability assays had shown OCaI tumors to be relatively radiosensitive and HCaI tumors to be radioresistant. At 6 h, approximately 20% of nuclei in OCaI tumors showed aberrations characteristic of cell death by apoptosis. This contrasted to an incidence of 3% in HCaI tumors. Mitotic activity was eliminated in OCaI tumors but was only transiently suppressed in HCaI tumors. At 24-96 h, OCaI tumors continued to display apoptosis and progressive necrosis, whereas HCaI tumors responded by exhibiting marked pleomorphism. Factors other than mitotic activity may influence tumor radiosensitivity, and one of these may be susceptibility to induction of apoptosis (programmed cell death), because this was a prominent early radiation response by the radiosensitive OCaI tumors.

Wild-type p53 gene transfer into the SW620 colorectal carcinoma cell line was performed using the replication-defective adenovirus Ad5/CMV/p53 to evaluate the effect of wild-type p53 expression on radiation sensitivity. The results indicated that infection with Ad5/CMV/p53 sensitized the cells. The survival at 2 Gy was reduced from 55 to 23%. Flow cytometric analysis of terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling (TUNEL) assay-labeled cells and in situ TUNEL staining of xenograft tumors demonstrated an increase in labeled cells with combination treatment, indicating increased apoptosis in cells treated with Ad5/CMV/p53 before irradiation. A significant enhancement of tumor growth suppression by this combination strategy was observed in a s. c. tumor animal model compared to p53 gene therapy alone. The delay in regrowth to control tumor size of 1000 mm3 was 2 days for 5 Gy, 15 days for Ad5/CMV/p53, and 37 days for Ad5/CMV/p53 + 5 Gy, indicating synergistic interactions. These data indicate that the delivery of wild-type p53 to cells with p53 mutations increases their radiation sensitivity, and this may be accomplished by adenoviral-mediated gene therapy.

In a previous paper (Radiat. Res. 127, 308-316, 1991), we reported that a moderately radiosensitive, transplantable murine ovarian carcinoma (OCaI) displayed apoptosis after irradiation whereas a radioresistant hepatocellular carcinoma (HCaI) did not. These initial observations have been followed up in this detailed analysis of the development of apoptosis in these two tumors as a function of time and dose. Histological sections of OCaI and HCaI carcinomas were scored at various times between 0.5 and 24 h after single doses of 2.5 or 25 Gy gamma radiation for the incidence of apoptosis. The percentage of nuclei undergoing apoptosis in untreated tumors was 5% in OCaI and 0.6% in HCaI. The peak in the number of apoptotic bodies occurred in the OCaI tumors 3-5 h after either dose. After 2.5 Gy, the peak incidence was about 20% and after 25 Gy it was about 30%. Irrespective of dose, HCaI tumors had an incidence of apoptosis of less than 3%. Based on the results of this time course, 4 h after irradiation was chosen for the determination of the dose response, over doses ranging from 2.5 to 25 Gy. The dose response for the OCaI tumors reached a plateau at 25-30% apoptotic nuclei after doses of about 7.5 Gy and above. Autoradiographic analysis of histological sections from mice injected with [3H]thymidine showed that some apoptotic bodies in the OCaI tumors arose from cycling cells. These results confirm that the apoptotic mode of cell death may represent an important response in some irradiated tumors.