Radiosensitization of hypoxic tumor cells by depletion of intracellular glutathione.

Depletion of glutathione in Chinese hamster ovary cells in vitro by diethyl maleate resulted in enhancement of the effect of x-rays on cell survival under hypoxic conditions but not under oxygenated conditions. Hypoxic EMT6 tumor cells were similarly sensitized in vivo. The action of diethyl maleate is synergistic with the effect of the electron-affinic radiosensitizer misonidazole, suggesting that the effectiveness of misonidazole in cancer radiotherapy may be improved by combining it with drugs that deplete intracellular glutathione.

Role of glutathione in the hypoxic cell cytotoxicity of misonidazole.

Misonidazole (MIS) is a hypoxic cell radiosensitizer currently undergoing Phase III clinical trials in the treatment of cancer by radiation. It is also a cytotoxic agent with specificity toward hypoxic cells, and consequently has a tumoricidal effect in laboratory animals. This tumoricidal effect has not been clinically applicable, in part because the initial resistance to the cytotoxic action of MIS (the nonexponential portion, or shoulder, of the semilogarithmic plot of the surviving fraction of the cell population versus the time of exposure to MIS, referred to hereafter as the shoulder of the survival curve) for cells treated with MIS under hypoxic conditions is too large to be overcome at clinically tolerable doses of MIS. We report here that pretreatment of Chinese hamster ovary cells in vitro with diethylmaleate to deplete intracellular glutathione results in a substantial decrease in the shoulder of the survival curve for MIS-treated hypoxic cells. Restoration of glutathione results in restoration of the shoulder of the survival curve and a slight extension beyond that seen with control cells. These results demonstrate that glutathione protects against the cytotoxic effect of MIS. However, glutathione depletion does not significantly affect the rate of binding of MIS metabolites to cellular macromolecules, indicating that the cytotoxicity of MIS is not simply a reflection of massive binding of MIS metabolites to cellular constituents. We propose that the cytotoxicity of MIS toward hypoxic cells is a result of hydrogen abstraction from target molecules by free radicals formed in the reduction of the nitro group.

Quantitative analysis of cellular glutathione by flow cytometry utilizing monochlorobimane: some applications to radiation and drug resistance in vitro and in vivo.

An assay using a bimane derivative has been developed to detect free glutathione (GSH) in individual viable cells by flow cytometry. Monochlorobimane [syn-(ClCH2CH3)-1,5-diazabicycla[3.30]acta-3,6-diene-2,8-dio ne], itself nonfluorescent, reacts with GSH to form a highly fluorescent derivative. High pressure liquid chromatography analysis showed that, using specific staining conditions, the only low molecular weight fluorescent derivative formed in Chinese hamster ovary cells was that formed with GSH. Very little reaction with protein sulfhydryls was observed. Rates of GSH depletion in Chinese hamster ovary cells exposed to diethylmaleate were essentially the same, whether measured by relative fluorescence intensity, by flow cytometry or by enzymatic assay on cellular extracts. This method was shown to be useful for measurement of GSH resynthesis, uptake, and depletion by prolonged hypoxia and misonidazole treatment. Since measurements are made on individual cells, cell-to-cell variation and populational heterogeneity in GSH content are revealed by flow cytometry. Although under most conditions in vitro GSH content is relatively homogeneous, under certain circumstances, such as release from hypoxia, heterogeneity in populational GSH levels was observed. The significance of this heterogeneity is discussed in regard to the induction of gene amplification and drug resistance by transient hypoxia. Numerous subclones of Chinese hamster ovary cells selected by growth in Adriamycin or methotrexate-containing medium express elevated levels of GSH per cell. The method was extended to quantitate the GSH content of cells excised from EMT-6/SF mouse tumors that had been treated in vivo with L-buthionine-S-R-sulfoximine, an inhibitor of GSH synthesis. The bivariate analysis (forward angle light scatter versus monochlorobimane fluorescence) of cells derived from these tumors gave excellent resolution of normal and tumor cells and demonstrated extensive heterogeneity in the tumor cell population with respect to GSH content per cell.

Chemical modifiers of cancer treatment.

Chemical modification is a concept in cancer therapy in which the state of tumor cells or normal tissues is modified such that a therapeutic gain can be achieved using conventional therapeutic modalities. Hypoxic zones targeted as cells within them may be radiation resistant, poorly perfused by chemotherapeutic agents, and possibly drug resistant due to hypoxia-related gene amplification. Nitroimidazoles have gained particular attention as chemical modifiers because they can increase the radiation sensitivity of hypoxic cells, are cytotoxic to hypoxic cells, can increase sensitivity to chemotherapeutic agents, and are useful for imaging hypoxic cells. While both radiosensitization and chemosensitization require hypoxia, the mechanism of the enhancement of each of the modalities is different. The 2-nitroimidazole hypoxic sensitizers SR 2508 and Ro-03-8799, which are less toxic than the prototype misonidazole (Miso), are in clinical trials, and dual function molecules that include a hypoxic sensitizer and alkylating function are being developed. The presence of both acutely and chronically hypoxic cells in animal tumors has been demonstrated by new imaging techniques. Oxygen delivery to tumors is being altered by the use of perfluorocarbons, and agents that alter hemoglobin affinity for oxygen. Compounds that are selectively toxic to hypoxic cells are being developed. Nonhypoxic modifiers are also being investigated. Thiol modification, particularly the alteration of glutathione concentration, has complex effects on the cell's biochemistry, in addition to affecting the competition between oxygen and thiol groups for the restoration and fixation of radiation-induced radicals. WR-2721 is being studied as a means of reducing the normal tissue toxicity of radiation and chemotherapy. Increased thiol concentration may be a mechanism of cross-resistance between certain chemotherapeutic agents and radiation.

Role of glutathione in the radiation response of mammalian cells in vitro and in vivo.

Radiation interacts with biological systems to produce many types of molecular lesions. Much of the molecular damage is of little consequence with regard to cell killing. The lesions that are most likely to contribute to cell killing are DNA lesions produced by clusters of radicals. The formation of clusters of radicals is characteristic of ionizing radiation and accounts for its high efficiency as a cytotoxic agent. The mechanism by which these lesions kill cells is probably the formation of DNA double-strand breaks, ultimately resulting in chromosomal breaks. There is a possibility that some of the other types of molecular lesions produced by radiation may participate in more subtle mechanisms of cell damage. For instance, radiation induces a self-destructive process (apoptosis) in certain cell types, and the molecular lesions that initiate this process have not been identified. Glutathione (GSH) is a versatile protector. Several distinct mechanisms of radioprotection by GSH can be identified. These include radical scavenging, restoration of damaged molecules by hydrogen donation, reduction of peroxides and maintenance of protein thiols in the reduced state. Of these mechanisms, hydrogen donation to DNA radicals is probably the most important. Since competing reactions are very rapid, this mechanism requires a high concentration of GSH. Radioprotection by hydrogen donation to DNA radicals is not effective in oxygenated cells because the normal intracellular GSH concentration is not sufficient for effective competition with oxygen. Consequently, moderate depletion of GSH has no effect on the radiosensitivity of oxygenated cells. Under hypoxic conditions GSH becomes more competitive, and GSH depletion can markedly affect radiosensitivity. The radiosensitivity of hypoxic cells is most affected by GSH depletion in the presence of low concentrations of radiosensitizers. Since hypoxic cells are a characteristic feature of tumors, moderate depletion of GSH in combination with treatment with hypoxic cell radiosensitizers appears to be a promising strategy for selective tumor sensitization in radiation therapy. Oxidation of GSH can result in radiosensitization of both hypoxic and oxygenated cells. The mechanism of this effect appears to involve oxidation of protein thiols which are important for DNA repair. In principle, modification of DNA repair could have a greater impact on radiation therapy than modification of the number of lesions produced by radiation. However, a strategy for modification of GSH or protein thiol redox state in vivo has not yet been devised.

Combined antitumor effect of radiation and ibuprofen in human prostate carcinoma cells.

Recent clinical observations indicate that ibuprofen may alleviate the radiation-induced dysuria that almost invariably occurs during radiation therapy for prostate cancer. Because the use of ibuprofen could consequently become common during radiation therapy for prostate cancer, we have been interested in the potential interactions between ibuprofen and ionizing radiation on prostate tumor cells. The effects of gamma-irradiation and/or ibuprofen on PC3 and DU-145 human prostate carcinoma cells were evaluated in vitro using three model systems. Clonogenic survival was determined by plating cells 24 h after treatment of nearly confluent monolayers. Analysis of cell growth, cell detachment, and apoptotic cell death was carried out over a period of up to 9 days after treatment of PC3 and DU-145 monolayers. The effect of ibuprofen and/or radiation was also probed by observing the inhibition of growth of established PC3 and DU-145 colonies that were treated on the 14th day of colony growth. Ibuprofen enhanced the radiation response of prostate cancer cells in all three in vitro models. Both the cytotoxic and radiosensitizing effects of ibuprofen seem to require concentrations that are higher than those reported to inhibit prostaglandin synthesis, suggesting that other molecular mechanisms may be responsible for ibuprofen cytotoxicity.

Oxidative stress-induced apoptosis of endothelial cells.

Endothelial cells (ECs) are subjected to oxidative stress during many pathological processes, including ischemia/reperfusion and general inflammation. In the present study, we examined the effects of oxidative stress on rates of apoptosis in EC cultures. We treated large and microvessel ECs with menadione for 1 h in vitro to simulate the most common physiological form of oxidative stress, exposure to O2*-. Capillary ECs were resistant to menadione-induced apoptosis when compared with large-vessel ECs. Treatment with 35 microM menadione resulted in an apoptotic rate of approximately 5% in capillary EC cultures compared with approximately 45% in large-vessel EC cultures. At higher concentrations of menadione (35-75 microM), both types of ECs exhibited a concentration-related increase in apoptosis. Necrotic cell death only became evident at menadione concentrations ranging from 75-100 microM for both cell types. The timing of the apoptotic response to a 1 h menadione exposure was very specific. For both EC types, peaks of apoptosis occurred in two distinct waves, at 6-8 and 18-22 h after treatment. Analysis of the events leading up to the first peak of apoptosis indicated that specific matrix metalloproteinases (MMPs) were activated, suggesting that MMPs may be involved in initiating the apoptotic process.

Novel fluorescein-based flow-cytometric method for detection of lipid peroxidation.

The novel property of fluorescein to detect peroxyl radicals is demonstrated. On the basis of this observation, a fluorescein-based, flow-cytometric method to directly and continuously detect free radicals generated in cell membranes during lipid peroxidation has been developed. 5- and 6-Carboxyfluorescein (5-/6-CF) free in solution and fluorescein-labeled polylysine lose their fluorescence gradually upon addition of a peroxyl-radical-generating system (thermal decomposition of 2,2'-azobis(2-amidinopropane) [AAPH]). 5-/6-CF retains its fluorescence when exposed to AAPH in the presence of the peroxyl radical scavenger Trolox. When 5-/6-CF free in solution is incubated with red blood cells exposed to cumene hydroperoxide (CH), a similar loss of fluorescence occurs due to lipid peroxidation on RBC membranes, which is preventable by pretreatment of the cells with Trolox or vitamin E. Undecylamine-fluorescein (C11-fluor), a lipophilic fluorescein conjugate, has been incorporated into the membranes of RBC. Upon addition of CH, a decrease in fluorescence is fluorometrically observed that is proportional to the amount of hydroperoxide added and inhibited by preincubation with Trolox or vitamin E. Flow-cytometric studies are then performed to demonstrate that C11-fluor can monitor free radicals generated during lipid peroxidation on a cell-by-cell basis. When exposed to CH, a time-dependent shift of the flow-cytometric profile toward lower values is observed that is inhibited by Trolox or vitamin E. This approach in conjunction with multiparametric flow cytometry may allow examination of the biologic significance of lipid peroxidation by correlation to other cellular end points on single cells.

The use of drugs which deplete intracellular glutathione in hypoxic cell radiosensitization.

Diethylmaleate (DEM) is a thiol-binding reagent with specificity toward glutathione. Treatment of Chinese hamster ovary (CHO) cells in vitro with 2 x 10(-4) M DEM for one hour results in a decrease in glutathione content to less than 5% of control, without cytotoxicity. This treatment results in dose-modifying sensitization to radiation under hypoxic conditions, with no effect on the shoulder of the radiation survival curve. No effect on the radiation sensitivity of oxygenated cells was seen. DEM pretreatment enhances the radiosensitization of hypoxic cells by misonidazole, as well. Similar results were obtained in vivo with EMT6 tumors in BALB/c mice. Analysis of DNA damage by the alkaline elution assay indicates that DEM enhances radiation-induced single-strand breaks, but does not significantly affect repair, while diamide and N-ethylmaleimide inhibit repair, in addition to enhancing radiation-induced single-strand breaks.

Studies on the mechanism of chemosensitization by misonidazole in vitro.

Misonidazole (MISO) depletes intracellular glutathione and is more toxic in glutathione depleted cells. The depletion is time, temperature, drug concentration and cell line dependent. The role of glutathione depletion in the chemosensitization to alkylating agents obtained after hypoxic pretreatments with MISO was investigated using diethylmaleate (DEM), a thiol-removing agent with specificity for glutathione, to simulate the effects of MISO on intracellular glutathione levels. Melphalan cytotoxicity and binding to macromolecules were measured after pretreatments with MISO or DEM in vitro. From these studies we found that glutathione depletion could account for only a part of the chemosensitization to melphalan and that this component of sensitization was quantitatively related to an increased rate of melphalan binding. Assessments of DNA damage by the alkaline elution assay suggest that DNA strand breaks and DNA crosslinks are enhanced by sublethal pretreatments with MISO.

Radiosensitization by diamide analogs and arsenicals.

Several analogs of the glutathione (GSH) oxidizing reagent diamide [diazenedicarboxylic acid bis(N,N'-diethylamide)] were tested as radiosensitizers of aerobic cells. In general, radiosensitization correlates with the rate of reaction with cellular reducing agents and occurs only when the reductive capacity of the cell is exceeded. SR-4077, [diazenedicarboxylic acid bis(N,N-piperidide)], is particularly suitable for mechanistic studies, because it is less cytotoxic than diamide, but is equally reactive toward cellular GSH. SR-4077 sensitizes CHO cells to X rays under aerobic conditions, even when the drug is added after irradiation. Radiosensitization is expressed both as a change in the exponential slope of the radiation cell survival curve and as a decrease in the shoulder of the survival curve. Phenylarsine oxide, a dithiol-binding reagent, sensitizes aerobic CHO cells to X rays by modification of the shoulder of the survival curve. The results are consistent with the hypothesis that the shoulder-modifying effect of GSH oxidants is caused by the loss of protein thiols, which might be involved in the repair of X ray-induced DNA damage.

Sensitization of Chinese hamster ovary cells to melphalan by etanidazole under intermittent hypoxia.

Incubation of Chinese hamster ovary cells with etanidazole under hypoxic conditions increases cell killing by subsequent treatment with melphalan under aerobic conditions. We report here that this sensitization can be achieved even if periods of hypoxia are interrupted by reoxygenation. Preincubation under hypoxia in the absence of etanidazole also results in sensitization to melphalan. Intermittent hypoxia is less effective than continuous hypoxia as a single sensitizing factor. Glutathione depletion does not appear to have more than an additive effect on chemosensitization by etanidazole. These results suggest that prolonged treatment with low doses of etanidazole may be a more effective strategy for clinical chemosensitization than a short exposure to a higher dose, in order to target intermittently as well as chronically hypoxic cells.

Role of the adenylate energy charge in the response of Chinese hamster ovary cells to radiation.

Steady-state modification of the adenylate energy charge (EC; ATP + 1/2ADP/ATP + ADP + AMP) in aerobic Chinese hamster ovary (CHO) cells was achieved with a combination of rotenone and 2-deoxy glucose (2dG). Neither agent significantly affected the EC by itself. All incubations were in Eagle's minimum essential medium containing 15% fetal bovine serum. The radiation response of these cells was not significantly affected by this treatment when cells were irradiated either one or three hours after addition of the drugs, and held in the presence of the inhibitors for one hour after irradiation. The ability of cells to repair radiation-induced single-strand breaks was studied by the alkaline elution method. Energy depleted cells repaired single-strand breaks at a slightly slower rate than the controls. However, thymidine incorporation was also inhibited, suggesting that repair may still have preceded events leading to the fixation of that damage (e.g., DNA replication).

Pharmacodynamics of prolonged treatment with L,S-buthionine sulfoximine.

PURPOSE: To develop dosing criteria for the use of L-buthionine-S-sulfoximine (active diastereoisomer) as a glutathione depletor in the clinic, using a pharmacodynamic and pharmacokinetic in vitro-in vivo approach. METHODS AND MATERIALS: In vitro: L-buthionine-S-sulfoximine uptake was determined in human glioblastoma cells (T98G) and NIH-3T3 cells using 35S-labeled drug. Dose response relationships were derived for inhibition of glutathione synthesis in CHO cells, and for depletion of glutathione in exponentially growing T98G and CHO cells, as a function of extracellular L-buthionine-S-sulfoximine concentration. Steady-state glutathione levels for CHO and NIH-3T3 cells were measured using an enzymatic assay, while glutathione synthesis rates in CHO cells were determined using a flow cytometric assay. In vivo: L-buthionine-S-sulfoximine biodistribution was determined in male nude mice carrying human glioblastomas (T98G) intracranially, using 35S-labeled drug infused subcutaneously by osmotic pump. Tissue glutathione levels were measured using an enzymatic assay. RESULTS AND CONCLUSION: The observed cellular uptake t1/2 of approximately 55 min, coupled with a previously reported, rapid in vivo clearance of buthionine sulfoximine, suggest that continuous infusion would be preferable to bolus dosing. Effective concentrations of L-buthionine-S-sulfoximine (24 h exposure), required to lower cellular glutathione content to 50% of control (EC50), were under 1 mM for both cell lines. The amount of L-buthionine-S-sulfoximine in tissues (estimated from 35S drug disposition) reached steady state within 8 h and was proportional to the rate of infusion. Brain tumors were depleted to approximately 50% of control glutathione by a infusion rate of 0.25 mumoles/h (25 g mice). At lower infusion rates an increase in glutathione content was noted in certain nude mouse tissues including brain tumor xenografts.

Novel concepts in modification of radiation sensitivity.

PURPOSE: To determine whether biological effects of radiation, such as apoptosis, that differ from classical clonogenic cell killing, can be modified with agents that would not be expected to modify classical clonogenic cell killing. This would expand the range of potential modifiers of radiation therapy. METHODS AND MATERIALS: EL4 murine lymphoma cell apoptosis was determined by electrophoretic analysis of deoxyribonucleic acid (DNA) fragmentation. DNA was extracted 24 h after irradiation or addition of inducing agents. Modifiers of radiation-induced apoptosis were added immediately after irradiation. The effects of radiation on wounded endothelial monolayers were studied by scraping a line across the monolayer 30 min after irradiation. Cell detachment was used as an endpoint to determine the protective effect of prolonged exposure to retinol prior to irradiation. RESULTS: EL4 cell apoptosis can be induced by tert-butyl hydroperoxide or the glutathione oxidant SR-4077. Radiation-induced EL4 cell apoptosis can be inhibited with 3-aminobenzamide, an agent that sensitizes cells to classical clonogenic cell killing. Radiation-induced endothelial cell detachment from confluent monolayers can be modified by pretreatment with retinol. CONCLUSION: These results raise the possibility that radiation could induce apoptosis by an oxidative stress mechanism that is different from that involved in classical clonogenic cell killing. These and other recent findings encourage the notion that differential modification of classical clonogenic cell killing and other important endpoints of radiation action may be possible.

Modification of the aerobic cytotoxicity of etanidazole.

PURPOSE: To determine the feasibility of modifying the aerobic cytotoxicity of etanidazole without interfering with the tumoricidal action of radiation plus etanidazole. METHODS AND MATERIALS: The aerobic cytotoxicity of etanidazole was studied using two different models: (1) Induction of apoptosis in EL4 cells: apoptotic DNA fragmentation was analyzed by agarose gel electrophoresis following 24 h treatment with etanidazole alone or in combination with various modifiers. (2) Spinal cord neuronal loss in organotypic roller tube cultures: Survival of acetylcholinesterase positive ventral horn neurons was analyzed morphometrically following 72 h treatment with etanidazole alone or in combination with vitamin E succinate. RESULTS: Etanidazole (10 mM) induced apoptosis in EL4 cells. This effect was suppressed by 24 h treatment with TPA, IBMX, the free radical scavenger TEMPOL or vitamin E succinate. Vitamin E succinate also protected spinal cord cultures from etanidazole-induced neuronal loss. CONCLUSION: These results suggest that it might be possible to modify the neurotoxicity of etanidazole with agents that would not be expected to interfere with the tumoricidal action of radiation plus etanidazole.

Elevation of mouse kidney thiol content following administration of glutathione.

We have found that kidney glutathione and cysteine content in C3H mice can be increased by intraperitoneal administration of either glutathione (GSH) or glutathione disulfide (GSSG). Kidney thiol content is maximal 20-60 min after administration of 1000 mg/kg glutathione and returns to normal values by 2 h. The same time-course of thiol perturbation was observed when acivicin, an inhibitor of gamma-glutamyl transpeptidase, was administered 15 min prior to GSSG administration. The increase in kidney thiols after GSSG administration appears to saturate, with little additional increase as the administered dose is increased above 750 mg/kg. There was no significant change in liver GSH or cysteine after GSSG administration. We suggest that glutathione administration may provide a strategy for selective radioprotection or chemoprotection of specialized cells which can effectively utilize systemic GSH precursors.

An oxidative stress-mediated death pathway in irradiated human leukemia cells mapped using multilaser flow cytometry.

OCI/AML-2 acute myeloid leukemia cells were found to undergo apoptosis after treatment with y rays from a 137Cs source. Multilaser flow cytometry techniques using probes for live cell function were used to monitor the biochemical changes that occurred prior to the loss of surface membrane integrity. These showed increases in the generation of reactive oxygen species (ROS) and in the glutathione (GSH) content of irradiated cells. An additional population of cells that showed a further increase in ROS and depletion of GSH was seen in irradiated cells but not in controls. This population showed loss of mitochondrial membrane potential (deltapsim), indicative of the mitochondrial permeability transition, and exposure of phosphatidylserine on the cell surface. Increases in intracellular calcium were observed in a proportion of these low-deltapsi(m)/high-ROS cells. Similar findings were seen using the antileukemia drug cytosine arabinoside (ara-C), although cell cycle analysis showed that the loss of deltapsi(m) occurred mainly in G1 phase with ara-C treatment, and mainly in G2 phase with irradiation. Furthermore, the protective effect of overexpression of BCL2 was more pronounced after ara-C treatment than with radiation. Cells of the TP53 (formerly known as p53)-null human AML line OCI M2 showed growth arrest in G2 phase after radiation treatment, with no loss of deltapsi(m) or morphological changes indicative of apoptosis. The flavine-dependent oxidoreductase inhibitor diphenylene iodonium failed to inhibit generation of ROS in irradiated OCI/AML-2 cells, indicating that the mechanism is unlikely to involve the TP53-induced gene PIG3. These results show that oxidative stress can occur in irradiated human leukemia "blasts", and may play a direct role in radiation-induced apoptosis.

The effect on the Km for radiosensitization at 0 degree C of thiol depletion by diethylmaleate pretreatment: quantitative differences found using the radiation sensitizing agent misonidazole or oxygen.

Pretreatment of V79- WNRE cells with 150 microM diethylmaleate for 1 hr at 37 degrees C caused a decrease in intracellular glutathione levels to approximately 10-15% of control levels (0.5 vs 5.0 nmol/10(6) cells). The cells could be washed free of diethylmaleate and held at 0 degree C for several hours without toxicity and with no increase in glutathione concentration, although the glutathione concentration rapidly increased to normal levels at higher temperatures. Survival curves were determined as a function of oxygen or misonidazole concentration (the latter in the absence of oxygen). A new "thin-film" technique was used to avoid changes in oxygen concentration because of radiochemical or cellular oxygen consumption. Glutathione depletion itself caused a small but consistent radiosensitization of hypoxic cells (dose enhancement ratio of 1.2). However, glutathione depletion caused a profound change in the radiosensitizing efficiency of misonidazole, with a decrease in Km of about sevenfold from 0.6 to 0.09 mM. In contrast, only a 2.5-fold decrease was found in the Km for radiosensitization by oxygen with diethylmaleate pretreatment. These results suggest a fundamental problem with the conventional theory of radiosensitivity whereby one considers a first-order competition for reaction with target radicals between radical-fixing versus radical-repairing species. It also suggests difficulties in the interpretation of glutathione as the only endogenous protective species.

Modulation of radiation-induced apoptosis and G2/M block in murine T-lymphoma cells.

Radiation-induced apoptosis in lymphocyte-derived cell lines is characterized by endonucleolytic cleavage of cellular DNA within hours after radiation exposure. We have studied this phenomenon qualitatively (DNA gel electrophoresis) and quantitatively (diphenylamine reagent assay) in murine EL4 T-lymphoma cells exposed to 137Cs gamma irradiation. Fragmentation was discernible within 18-24 h after exposure. It increased with time and dose and reached a plateau after 8 Gy of gamma radiation. We studied the effect of several pharmacological agents on the radiation-induced G2/M block and DNA fragmentation. The agents which reduced the radiation-induced G2/M-phase arrest (caffeine, theobromine, theophylline and 2-aminopurine) enhanced the degree of DNA fragmentation at 24 h. In contrast, the agents which sustained the radiation-induced G2/M-phase arrest (TPA, DBcAMP, IBMX and 3-aminobenzamide) inhibited the DNA fragmentation at 24 h. These studies on EL4 lymphoma cells are consistent with the hypothesis that cells with radiation-induced genetic damage are eliminated by apoptosis subsequent to a G2/M block. Furthermore, it may be possible to modulate the process of radiation-induced apoptosis in lymphoma cells with pharmacological agents that modify the radiation-induced G2/M block, and to use this effect in the treatment of patients with malignant disease.