The Effects of Ionizing Radiation on Mammalian Cells

Transcription

1 The Effects of Ionizing Radiation on Mammalian Cells John E. Biaglow Case Western Reserve University, Cleveland, OH Downloaded via on November 9, 2018 at 21:31:09 (UTC). See for options on how to legitimately share published articles. Human and animal cells, both normal and tumorogenic, can be grown in culture in vitro and studied for their responses to ionizing radiation (for reviews see (i <S)). In vitro studies with ceil cultures may involve cytotoxic and mutagenic effects of radiation. There is also a good deal of current interest in the effects of drug and radiation combinations on cell cytotoxicity. In addition, the effects of hormones and radioprotective agents are being actively pursued. For these studies the cells can be grown under various conditions such as log phase, plateau phase or growth-arrested and in multicellular arrays such as spheroids. Cells may also be grown at lower density and concentrated before irradiation in order to duplicate tissue-like densities. Cells may be collected when they round up for division, or they may be chemically synchronized in order to study the effects of radiation alone or with drugs at various stages of the cell cycle. An important advancement in understanding the nature and severity of radiation damage to mammalian cells was accomplished by the introduction of several techniques to measure the survival of single cells (1-4). These techniques provided the first opportunity to quantify the biological effects of radiation on a cellular basis. The introduction of these survival techniques also hastened the understanding of effects of various biological, chemical, and physical modifications on the survival of cells after radiation. Cell survival techniques are limited to dividing or potentially dividing cells. Although this limitation may seem severe, the dividing cells of an organism are important because they comprise the stem-cell compartment and are more sensitive to radiation than the differentiated cells. In whole body exposure to radiation, the ability of the stem cells to divide determines the life or death status of the individual. The dividing cells are also the ones of interest in the treatment of mammalian tumors with radiation. In this review we are concerned with the effects of radiation on dividing cells and the factors that influence it. In addition, the radical mechanism for radiation damage is briefly reviewed. The mechanism of action of hypoxic cell radiosensitizing drugs is discussed as well as the ability for thioloxidizing agents to enhance the radiation response. A model for the influence of oxidation-reduction reactions on radiation and damage is examined. The ability of oxygen to increase radiation damage, and the effect of oxygen-sparing drugs on the radiation response of multicellular systems is described. A good deal of emphasis is given in this review to oxygen, the oxygen-mimicking drugs, and to the oxygen-sparing drugs because of the central importance of oxygen in radiation effects (1-3, 5, 6). Assays for Radiation Effects In radiation biology, the basic assay for the effect of radiation involves the survival of cells after treatment (the survival curve). Cells may be treated with a single or multiple dose of radiation in the presence or absence of drugs and under different growth conditions. This survival is dependent upon the ability of the initial cell to divide and produce a stainable colony after 7-14 days of growth in a chemically defined me- 144 Journal of Chemical Education PLATING OF SINGLE CELLS FORMATION OF COLONIES Figure 1. The colony-forming assay in vitro cells are plated as single cells and scored 1-2 weeks later as colonies. dium. The in vitro colony determination was made possible by advances in cell culture technique (1-6) to allow for the growth of single cells. The colony formation assay was developed by Puck and Marcus for mammalian celts in 1956 (4). The technique has not changed much since that time. However, some cells that are not easily cultured in liquid medium have been found to produce colonies if agar is included in the growth medium (7, 8). Agar techniques permit the assay of a number of additional cell lines, particularly those derived from human tumor tissue (7) that cannot be assayed in the conventional liquid medium. Both techniques involve plating of single cells into a Petri dish or a flask containing a cell culture media with serum or protein supplement (1-7). The cells are treated either in suspension or after plating onto the assay flask. Different series of replicate plates, or the suspensions, are treated with a particular radiation dose. The survival is measured by scoring the number of single cells that develop colonies (Fig, 1), Since some unirradiated cells fail to form colonies, a correction has to be made in terms of the plating efficiency. Thus survival is calculated by: Number g_ of Colonies Formed Number of Single Cells Plated X (PE)C where (PE)C is the number of colonies formed per number of single cells plated without irradiation. A simplified example showing only one plate for control and 500 rad of radiation is given in Figure 1. In practice it is difficult to obtain pure single-cell populations, so a correction is made generally for higher multiplicities (average number of cells per group at the time of irradiation). Other complications arise from the difficulty of obtaining surviving colonies of uniform size. Therefore, when counting colonies for clonogenic assays an

2 I I I 1-[Till 11 1-[Till 1- irill 1-[TTTTTTT arbitrary number of cells/colonies, usually 7-8 doublings ( cells), is chosen and colonies below that number are excluded or counted as a special group. The effects of irradiation of cells, as initially determined by Puck and Marcus with HeLa cells, have been verified by scores of investigators using a variety of animal and human cell types in vitro (1-7). Figure 2 shows the shape and parameters of a typical survival curve obtained from target theory. The slope of the exponential portions of the curve is given in terms of Do, the dose to give an average of one hit per target and thereby to reduce survival to 0.37 of any survival level of the exponential portion of the curve. The shoulder region of the curve at low doses suggests that the cells have the ability to accumulate sublethal damage before additional radiation causes the sublethal damage to become lethal. In terms of the target theory, the shoulder region is indicative of either several targets in the cell which must be inactivated or target(s) which must be hit several times to kill the cell. The exponential portion of the curve can be extrapolated to zero dose to obtain the extrapolation number n. This number is highly variable and in practice is not a reliable index. Another useful term to describe the shape of the survival curve is Dq, the dose in rad where the exponential portion of the curve intercepts 100% survival. Dq is a convenient number since it measures the shoulder width in rad, not a unitless number like the extrapolation number n. The three parameters are related from derivation of multitarget theory: DQ = Do In n Survival Curve Differences There is a vast amount of literature on dose survival curves for numerous cell lines; however, most of these have been obtained under different growth conditions. Therefore, we have obtained information on a number of cell lines irradiated and assayed under the same conditions (Fig. 3). We have determined the survival curves for the mouse lymphocytic leukemic L5178Y (L suspended culture), L5178Y (L-A grown attached), Chinese hamster ovary (CHO) and CHCL, a morphological variant. We have also determined the radiation response of V hamster lung cells and an isolated variant of this line, the V79 171B. All of these cell lines vary in their response to radiation. The most sensitive lines are the L5178Y (suspended) and the L5178Y (attached). The difference between these two lines is the small shoulder obtained with the attached cell line. The Chinese hamster ovary (CHO) line is more sensitive than its morphological variant, and both have shoulders. The most resistant, cell line was found to be the V Its cloned variant was more sensitive to radiation. The Do values are seen in the lower left-hand corner of the figure. These cell lines demonstrate the variation in radiation response obtainable with different cell lines, the variation in D0 from 80 to 240 is comparable with values found in the literature. Of importance is the difference in the shoulder of the survival cure between the most resistant line (V B), and the most sensitive shoulderless L5178Y lines suggesting a different capacity to accumulate sublethal damage. The large variation in the D0 values as well as the shape of the survival curve makes comparisons between laboratories nearly impossible unless rigorous conditions for growth, irradiation and plating of cells have been maintained. Indirect and Direct Effects of Radiation Cells are usually irradiated in an aqueous environment containing salt or nutrients such as proteins, vitamins and amino adds, and approaching conditions thought to exist in vivo (8). In addition, cells contain considerable amounts of water. Therefore, the radiation chemistry of water is of primary importance in explaining the biological effects of ionizing radiations (5, 6, 9). One of the reactions observed in aqueous solutions is the formation of the free radicals -OH and -H from Figure 3. Dose survival curves for V and 171-B lung cells, Chinese hamster ovary CHO and CH02 and the lymphocytic leukemia lines L5178Y (L) and L5178Y (L-A) grown in McCoy's 5a medium. The cells were irradiated with 220 kv X-rays, 37 C, in McCoy's 5a medium and colony formation was determined 10 days later. [HIM I Volume 58 Number 2 February

3 - - H20 by the ejection of an electron followed by dissociation of the free radical ion : HOH e- + (HOH-)+ H+ + -OH The ejected electron is picked up by water formed: HOH + e~ (HOH)- H- + OH" and H- is These radicals carry no charge but have a strong affinity for electrons (or hydrogen bonds). They can remove hydrogen atoms from other encountered molecules, e.g., from organic molecules in the cell: -OH + RH^R. + HOH Most of the damage to critical molecules such as DNA, RNA, or protein in the living cell occurs by this indirect mechanism. Notedly, -OH and -H radicals will on encounter recombine to form water. Therefore, they rapidly disappear through recombination or other radical-radical interactions. In the presence of 02, recombination of -H and -OH is minimized by formation of the hydroperoxyl radical (-H + 02 H02 ) which is a more stable free radical but is also capable of extracting H atoms from certain organic molecules. This accounts in part for the observed enhancement of damage by irradiation in the presence of oxygen, i.e. the oxygen effect (2, 3, 5, 6). Although the formation of free radicals in critical molecules of the cells occurs largely via the indirect mechanism, it can result also from an ionizing event produced by a direct hit in the molecule. A free radical site is produced by both direct and indirect mechanisms: R-. A variety of reactions can occur with a molecule containing a free radical. Immediate chemical restitution: R- + H- RH -* is minimized in the presence of oxygen but increased in the presence of H-donors (e.g., cysteine, glutathione, etc.). Free radicals can combine with 02 to form RCL" which prevents repair. In this reaction stable oxidative products formed are no longer susceptible to immediate chemical restitution. This accounts for another major part of the oxygen effect (see later). Reactions with other free radicals: R- + -OH - ROH R- + Rj- - R:R, A number of examples of alterations in molecules of biological importance can be cited, such as the splitting of peptide bonds in proteins by opening of heterocyclic ring structures (e.g. tryptophan) and deamination as well as cross-linking. With respect to nucleic acids, there can be splitting of the polynucleotide chain, deaminations, removal of purine or pyrimidine bases, cross-linking, and alterations in purines and pyrimidines. Cellular Effects Many attempts have been made to explain effects of radiation on living cells through interference with specific parts of the metabolic machinery. Studies have been performed with isolated enzymes, enzyme systems, such as those involved in protein synthesis in vivo and in vitro as well as lipid and carbohydrate metabolism. In all cases, the measured effects are too small to account for a specific mechanism for the overall observed damage to the intact cell (2, 3, 5). It is conceivable that breakdown of internal structures (membrane endoplasmic reticulum, etc.) may account for damage which cannot be measured readily by biochemical tests. Loss of ions from the cell nucleus, reduced uptake of metabolites from the cytoplasm by the nucleus and the inhibition of nuclear oxidative phosphorylation have been reported (2, 3). However, their exact importance for altered cell metabolism resulting in cell death has not yet been determined. At present we are unable to explain adequately why some non-dividing cells (such as lymphocytes) disintegrate after exposure to relatively limited radiation doses, whereas others (e.g. liver cells) appear relatively unaltered and continue to function. One clue to this appears to be the large difference in the cytoplasm/nuclear ratio in these cases: cells with large cytoplasmic volumes are generally more radioresistant. The most conspicious cellular effects of radiation are mitotic delay, chromosome aberrations, mutations (including carcinogenic transformation), and cell killing or loss of reproductive ability (2, 3, 5, 6). The most important and certainly the most studied one is cell killing. However, some cells survive irradiation and are not killed. This survival is due to their repair capacity. Repair of Radiation Damage The radiation dose response curve for many of the mammalian cells is not a straight line on semilog scale but has a shoulder (Fig. 2). This shoulder has been explained in at least three ways. 1) The cell may require hits in more than one sensitive area (or volume) to produce lethality. If rt is the number of targets, it may be shown that the number of survivors is represented by: NJN = 1 (1 e~dtd0)n Dose Figure 4. Illustration of the effect of adding two doses of radiation when the survival curve has a shoulder due to fime dependent repair. For D given acutely, survival is Fj; for D given in two equal doses seperated by several hours, survival is F,. Os is sometimes known as the lost dose due to fractionation (10). if one plots log NJN versus D. Do is a measure of the slope of the high dose straight line portion while n is the extrapolation number as shown in Figure 2. This is called the multi-target single hit model. Log NJN was previously defined as the surviving fraction or the number of colonies formed/number of single cells plated X 100 (plating efficiency). 2) The cell may require n hits in the same area. This concept is called single target multi-hit theory, and while it leads to survival curves similar to those in Figure 2, it is a much less popular explanation of the shape of the survival curve for single cells. It has been applied most commonly to colony survival where all of the cells in a group must be killed to suppress colony formation. 3) The cell can repair some of the damage to the radiation (2. 3, 5, 6). Cellular recovery from exposure to ionizing radiation is time dependent, and numerous experiments have shown that it occurs in less than 2 hr following the end of the irradiation. 146 Journal of Chemical Education

4 4 8 Figure 5. The survival of cells treated with 2 conditioning doses of radiation followed in time by a second dose of radiation. The increased survival following the first doses is suggested as the repair of sublethal radiation damage. The dip between 3 and 8 hr indicates movement of cell cycle progression into a more sensitive phase. Figure 6. Cell-cycle dependent variation in sensitivity to radiation. The response ot cells to radiation varies throughout the cell cycle. The major, readily identifiable stages of the division-cycle are shown graphically in this figure. In most mammalian cells, the duration of the different stages are: - M= hr. Gj = 2 hr. S = 6 hr. = Gi variable (0 hr-very long) Cell cycle dependent variations in radiosensitivity has been studied with synchronized populations. Figure 7, The survival of cells irradiated at various stages of the cell cycle. That sublethal recovery is associated with the low dose shoulder of the survival curve may be seen is Figure 4 which illustrates the effect of interrupting the exposure at dose A, waiting an interval longer (say 3 hr), and then delivering the rest of the dose (JO). A new curve with shoulder appears at point A, if radiation repair has occurred. For two equal doses as shown, it is obvious that the number of survivors (Fi at I)) is greater than if all the dose had been delivered at once resulting in F2 at D, If the dose response curve were simply log-linear, there would be no difference in the number of survivors. Ds is a measure of the shoulder and has been considered as the lost or wasted dose from fractionation. Another means for demonstrating repair of sublethal radiation damage is by the use of the split dose experiment originated by Elkind and Sutton {11). If a culture of cells is given an initial dose of irradiation (a conditioning dose) which will bring the survival past the shoulder of the dose-survival curve, and then a second dose is given at various times thereafter, the surviving fraction will vary with the interval between the two doses (Fig. 5). The early increase in survival (for short intervals between the doses) has been interpreted to reflect repair of sublethal damage; the further variations probably result from progression of cells through the cell cycle (cells not killed by the first dose move from a less sensitive state to a more sensitive stage in the celt cycle (Figs. 6 and 7). Other Types of Repair The post-irradiation treatment of cells irradiated with a single dose of X-ray has been found to affect the cell survival. Various ways of slowing down cell metabolism, such as lowered temperature or absence of nutrients in the growth medium, lead to increased survival, indicating that some potentially lethal damage has been repaired. It is not presently clear whether sublethal (see previous discussion) and potentiallylethal damage are related. Other post-irradiation treatments may increase the killing; these are often treatments with drugs that inhibit repair. The effect of elevated temperature (hyperthermia) during or immediately after irradiation has been explained as an inhibition of post-irradiation repair. Radiation-induced cellular damage seems to be maximized at two stages of the cell cycle; during S and during mitosis period. Any treatment that tends to alter the temporal relationship between repair and fixation of damage (Figs. 6, 7) appears to affect lethality. The cell-cycle dependent variation in radiosensitivity discussed later (next section) may be explained on the basis of such interplay between repair and fixation of damage. The term repair has also been used at the tissue level, in cases where repopulation or restoration of cells would be more accurate terms (2, 3, 5, 6). Cell Cycle-Dependent Variation in Sensitivity to Radiation The response of cells to irradiation varies through the cell cycle. The major, readily identified stages of the division cycle are shown graphically in Figure 6. In most mammalian cells, the duration of the different stages are: M = hr, G2 = 2-4 hr, S = 6-8 hr, Gj = variable (0 hr to very long). The cell cycle dependent variations in radiosensitivity has been studied with synchronized cell populations. Cells selected when they round up for division are in M or early Gi and may be considered synchronized. Figure 7 shows schematically the effect in terms of surviving cells (linear scale) following irradiation with a constant dose at different times of the cell cycle. The most sensitive stages (lowest survival) are mitosis (including late G2), late G ] and early S. Mid Gi and late S are relatively resistant stages. It is interesting to note that dose survival curves for cells in mitosis are essentially shoulderless, whereas cells in mid G] and late S' have pronounced shoulders. The dose survival curves for normal asynchronous or randomly dividing cultures are the composite of the dose-survival curves for cells in the different stages of the cell cycle (2, 3, 5,6). Volume 58 Number 2 February

5 Target for Radiation Damage The primary target for radiation-induced cell killing is believed to be the DNA molecule {2-6). The evidence for DNA occupying this central position is manifold and obtained both directly and indirectly, for example: 1) The size of the target is important and because the nucleus, containing cellular genetic material, constitutes a large part of the cell the Iikelybood for a hit on the DNA molecule is statistically quite possible. 2) The DNA molecule is unique and has low redundancy, so that each molecule is more vulnerable than the highly redundant compounds such as lipids; also the central position of DNA in the control of cellular activity makes it critical and the damage gets amplified through transcription and translation. 3) Direct irradiation of the cell nucleus (e.g., with microbeams) is many times more efficient than irradiation of the cytoplasm of the cell. 4) Chromosome damage parallels lethality. 5) Sensitivity of many related cells and organisms parallel the DNA content and to some extent the base composition of DNA. 6) Incorporation of radioactive molecules, such as 3H-thymidine, into DNA leads to very efficient cell killing. 7) Incorporation of base analogues (e.g., 5-bromodeoxyuridine) into DNA increases the sensitivity to ionizing radiation. 8) Cell killing has been shown to be associated with structural damage to DNA (e.g., double-strand breaks; see also 4 above; chromosomal damage). The identification of DNA as the main target for cell killing by radiation makes it possible to visualize repair of sublethal and potentially-lethal damage as being related to repair synthesis of DNA. Also, the fixation of radiation damage during DNA replication might explain the decrease in sensitivity of cells as they progress through the S period (see earlier). The Oxygen Effect with Cells No other chemical has been as widely studied as oxygen and yet continues to be such an intensive subject of study as a sensitizer of mammalian cells (2-6). The early studies of Gray (12), Figure 8, indicate that as the oxygen concentration is lowered there is a corresponding decrease in the radiation response of cells. The relative radiosensitivity of cells increases rapidly between 0 and 0.3% oxygen. Further increases occur until approximately 30 mm oxygen after which additional increases are very small. The OER (Oxygen Enhancement Ratio), or relative radiosensitivity, varies between 2 and 3.5 for the majority of cells (2, 3, 5, 6). Oxygen has been studied as a sensitizer because of the problem thought to occur in vitro with human tumors. Human tumor cells outgrow their blood supply resulting in a decreased availability of oxygen and resulting in hypoxic and anoxic tumor areas which may exhibit, a decreased radiation response. The chief cause of the oxygen effect in vivo is due to the consumption of oxygen by the tumor cells. The metabolic utilization of oxygen decreases the distance to which it may penetrate in the cells more distant from the capillaries. Unlike the physiological situation, most experiments are performed in vitro with cells that have been Figure 9. The effect of different densities of cells on the rate depletion of oxygen in a sealed container. The disappearance of oxygen was monitored with a Clark oxygen electrode. The cells were incubated at 37 C in 20 mm FIEPES buffered physiological saline, ph 7.4. Figure 8. Illustration of the dependence of radiosensitivity on oxygen concentration. If the radiosensitivity under anoxic conditions is arbitrarily assigned a value of unity, the radiosensitivity is about 3 under weil-oxygenated conditions. Most of this change of sensitivity occurs as the oxygen concentration increases from zero to 30 mm of mercury. A further increase of oxygen content to that characteristic of air, or even pure oxygen at high pressure, has little further effect. A sensitivity halt-way between anoxia and full oxygenation occurs at a partial pressure of oxygen of about 3 mm, which corresponds to aboul 0.5 %. This diagram is idealized and does not represent any specific experimental data. Experiments have been performed with yeast, bacteria, and mammalian cells in culture; the results conform to the general conclusions summarized above. Figure 10. The effect of metabolically produced hypoxia on the radiation response in dense suspensions of Chinese hamster ovary cells. In the top curve the 2 X 10 CHO cells/ml in 0.02 MHEPES buffered media were drawn up into a glass syringe, immediately radiated, at 37 C diluted and assayed for clonogenic survival. In the lower curve the same density of cells was spread as a layer of cells on the surface of a T30 flasks, gassed with humidified CO2 and irradiated at 0 C, diluted and assayed for colony forming ability. 148 Journal of Chemical Education

6 equilibrated with a nitrogen-carbon dioxide gas mixture for a period of time that insures the depletion of the dissolved oxygen. These procedures are laborious and require special equipment. A simpler way of demonstrating the oxygen effect is to concentrate cells into a dense suspension approaching in uiuo cell densities. We have found that cells under these conditions exhaust their supply of oxygen within minutes (Fig. 9) and may be irradiated immediately, diluted and plated for survival assays. Figure 10 shows the radiation response of this type of dense cell suspension and compares it to the radiation response obtained for cells at the same density but irradiated at zero degree to inhibit cellular oxygen utilization. The ratio of D0 values for each curve results in a dose modifying factor of 3.3. The OER will vary between cell lines and depends on the previous history and growth conditions of the cells (2, 3, 5, 6). Mechanisms of Oxygen Effect Flanders and Moore (13) proposed that two types of radiation damage are produced in cells, one which is oxygen dependent and the other oxygen independent. For example DNA radicals produced by radiation would he subject to a competitive chemical challenge, with oxygen on the one hand and intracellular donors on the other competing for the radicals (Fig. 26). In the absence of oxygen the first type of oxygen dependent damage may revert chemically to a harmless state, possibly chemically repaired by hydrogen donation from SH-containing compounds or other hydrogen-donating molecules: R- + XSH -* RH + XS- If oxygen reacts the so-called damage may be fixed. The cell is unable to repair itself thus resulting in lethality: R* + O2 * RO2" There is as yet, no direct evidence for this competition in irradiated cellular systems, although there is no doubt that it applies in simple model systems. The competition between chemical repair and oxygen fixation leads to a working hypothesis for the oxygen effect. This approach was further developed by experiments with the reactions of suspected DNA radicals with oxygen, electron affinic sensitizers and sulfhydryl compounds in irradiated bacteriophage and with purified DNA (2,14). A reaction scheme based on the competition between oxygen and sulfhydryl compounds for the oxygen dependent damage was proposed. With the development of alkaline sucrose gradient techniques other workers carried out experiments to study the oxygen dependence on radiation-induced single strand breaks. Strand breakage was enhanced in the presence of oxygen by a factor of 2 to 3 (2,6, 14). Others (15,16) studied DNA strand breakage efficiency as a function of oxygen concentration and concluded that the oxygen effects on strand breakage are similar to those of the radiation killing. These results were interpreted as supporting the viewpoint that the radiation target associated with reproductive death by the oxygen effect is DNA. Agreement on this subject is by no means universal, however, other investigators suggest that other target sites such as the membrane may be important in radiation-induced cellular inactivation, particularly in vivo with lung and heart tissue (17). Despite considerable research in past years, the basic mechanism by which oxygen sensitizes cells to the action of radiation is still not completely understood. For example there are reported deviations from the predicted dependency of OER on oxygen concentration (18, 19). Part of the difficulty in attempting to understand the mechanism of radiosensitization by oxygen is due to the very fast time scale (I0-3 sec) involved in these processes. Oxygen must be present at least 1-2 X 10-3 sec prior to irradiation. This has been demonstrated using rapid mixing systems, a gas explosion method, pulsed irradiation and mechanical mixing techniques (2, 9, 18-20). The fast time scale for the oxygen effect in biological systems, combined with knowledge of the chemical reactions induced by radiation of aqueous systems, strongly implies that the mechanism for the oxygen effect involves free radical reactions. The oxygen fixation hypothesis (2-6, 9) states that O2 reacts with radiation-induced free radical sites on the target molecules, presumably DNA (see Fig. 26), to form peroxides, which are believed to be non-reparable forms of damage (14). Rapid mixing technique (20) demonstrated that irradiation at the shortest time possible after mixing, about 4 X 10~3 sec, resulted in an oxygen enhancement ratio of 1.7, regardless of the oxygen concentration (from 1 to 50%) in the mixed solution. The OER increased to its full value of 2.8 as the time between mixing and irradiation was increased to 4 X 1CT2 sec. The profile of the increase of OER as a function of time after the initial 4 X 10-3 sec level was dependent on the oxygen concentration. From these data it might be concluded that there are two components to the oxygen effect, with the slower one demonstrating a dependence on oxygen concentration. The hypothesis of a two-component oxygen or sensitizer effect has not been widely accepted and there is much current work in this area. Hypoxic Cell Radiosensitizing Drugs (Oxygen Mimicking) In the last decade there has been rapid development in the field of specific radiosensitizers for hypoxic cells, due largely to the work of numerous chemists and radiobiologists (2,21) who defined the desirable criteria for these agents and then set out to select and design appropriate compounds. Most of the initial work was performed with cultured cells (27, 22). Basically there was an attempt to develop drugs that would not be metabolized as rapidly as oxygen, yet would be oxygen-mimicking and possibly of use in vivo in sensitizing hypoxic tumor cells. The electron affinic, hypoxic cell radiosensitizing drugs all contain an aromatic ring and a nitro (NO2) group, which appears to be the critical structural feature. Sensitizer studies to date include nitrobenzenes, nitrofurans, nitroimidazoles, nitropyrroles, and nitropyrazoles (21-24). The most important nitro compounds (Fig. 11) so far studied are the nitroimidazoles, metronidazole (Flagyl) and misonidazole. In Figure 12 are seen the effects of misonidazole on hypoxic cells, There is a small effect with 1 mm, however, the response is greatly improved but not as good as O2 when the concentration is increased to 10 mm. There is no effect of sensitizer in the presence of oxygen. The effect of medium containing dissolved oxygen, (0.2 mm) is also seen. Various nitro aromatic compounds sensitize hypoxic cells to X-rays by increasing the slope of the survival curve (Fig. 2); they do not affect the shoulder except under conditions of prolonged contact prior to irradiation (25). Demonstration of the radiosensitizing ability of a compound in vitro is no guarantee of CHjCK(OH)CH2OCHj CHjCHjOH NITROFURAN N-ETHYLMALEIMIDE <CH3)sNCON=NCON(CH3 2 "DIAMIOE" Figure 11. Structure of some nitro aromatic hypoxic ceil radiosensitizing drugs. Volume 58 Number 2 February

7 Dose (Rad) Figure 12. The effect of misonidazole on the radiation response of hypoxic Chinese hamster ovary cells {20). Figure 14. The closed triangles and circles represent the survival of Chinese hamster cells irradiated with 60Co 7-rays under aerated and hypoxic conditions. The survival of hypoxic cells irradiated in the presence of 2mM Ro is represented by open and closed triangles. Open triangles refer to irradiations carried out immediately after the drug was added and the cells made hypoxic. Closed triangles refer to irradiations carried out after the cells were stored at room temperature for 5 hr following the addition of the drug, a treatment which killed about 80% of the cells (25). Misonidazole Concentration (mm) Figure 13. The effect of misonidazole on the nonprotein and the protein thiol content of Ehrlich ascites tumor cells. Cells (107/ml) were incubated anaerobically in 0.05 M PBS and 10 mm glucose at 37 C, ph 7.3. Thiol determinations were performed 15 min after anaerobic conditions had been achieved (31). favorable results in vivo, since there may be biochemical, physiological, or pharmacological complexities involved in vivo which might interfere with sensitizer action. For example, alterations in cellular biochemicals (26) as well as effects on cellular oxygen utilization (27) may alter the effectiveness of the agents in vivo (see Oxygen-Sparing Drugs). Unfortunately, except for a few isolated cases, the basic studies on the mechanism of action of many of these compounds has been superceded in an attempt to find drugs that will be clinically valuable. With practicality as a guideline the nitroimidazotes are being used for studies with humans. Metronidazole and misonidazole (Ro ) are undergoing clinical trials in several cancer centers. At this time, studies are being carried out to determine appropriate dose and fractionation schemes for both drugs and radiation. This would result in therapeutic gain due to radiosensitization in selected tumor sites, e.g. brain, without the complication of neurotoxicity observed during the early stages of testing these compounds. Another limitation of these drugs is the effective 150 Journal of Chemical Education concentration that must be reached within the tumor to achieve an observable radiosensitization. These concentrations lie in the range of 0.5 to 1 mm, which produces a dose modifying factor of 1.2 to 1.7, Another difficulty is that clinicians use low radiation doses in order to prevent mistakes; consequently, the improvement in the actual radiation response in humans is at present marginal (24, 28). Mechanism of Action of Hypoxic Cell Radiosensitizers There is general agreement that the electron-affinic sensitizers mimic the action of oxygen at the level of DNA (20-24, 28). These contentions are supported by the facts that these sensitizers have no radiation effect on aerated cells, and the ability to sensitize is closely correlated with the one-electron reduction potential, a measure of electron affinity determined by pulse radiolysis (20). Oxygen has the highest one-electron reduction potential and is the most effective radiosensitizer. The electron affinity parameter is obviously important, but other variables have been investigated. Since penetration of the sensitizer to hypoxic cells of a tumor is of fundamental importance for effectiveness in vivo, it is reasoned that high solubility in lipids might be a desirable feature in enhancing the diffusion of the drug. Studies of the octanol: water partition coefficient of sensitizers were therefore conducted (28). It was found that there is no correlation between sensitizer effectiveness and lipophilicity. Side effects such as neurotoxicity may be enhanced if the drug partitions itself in high lipid containing tissue such as brain (28). Additional studies have demonstrated that nitro compounds are far from inert and that upon addition to cells they profoundly alter cellular electron transfer processes involving cellular respiration (27, 29, 30). Intracellular levels of reduced species such as NAD(P)H and glutathione are also affected (26, 29-32). The effect of hypoxic cell radiosensitizing drugs on cellular glutathione (Fig. 13) and protein thiols has been demonstrated and may be the reason that this drug shows an enhanced radiation response when cells are preincubated with the drug for prolonged periods of time (Fig. 14). Cells also show an enhanced radiation response when predated with misonidazole alone, washed and then irradiated under hypoxic conditions. However, cells are more sensitive in the presence of misonidazole (29, 30). Other studies have shown that if the thiols are removed by

8 Figure 16. Titration of GSH and NAD(P)H in EAT cells. Single doses of diamide were added to samples of 4 X 107 cells/3 ml. The recorded values for NAD(P)H represent the points of minimal fluorescence, before regeneration of reduced pyridine nucleotide occurred. GSH was measured in duplicate samples, also at the time of minimal fluorescence. The control value for GSH was 2.5 X 10-6 nmoles/ceil (38). Figure 15. Sensitization of V79-GL1 cells to 250 kvp X-rays by 100 /zm diamide plus 5 mmro (Misonidazole) in deoxygenated MEM +15% serum. The broken lines show the survival observed for the same concentrations of diamide or misonidazole independently as well as the no-drug control (34). the sulfhydryl oxidizing agent diamide there is a synergistic effect on the radiation response that is greater than that obtainable by either agent alone (31, 32). Harris and Power reported a lower survival at both 800 and 1200 rads for anoxic cells irradiated in the presence of diamide and nifuroxime (a nitrofuran) than was observed for either sensitizer alone (32). Similarly, Chapman et al. (33) showed that diamide plus NF-269, another nitrofuran, sensitized to a greater degree than that of either agent alone, and more than oxygen. Watts et al. (34) found that the combination of diamide and misonidazole was more effective than each individual compound with a survival curve showing a reduction in n and a change in slope (Fig. 15). Diamide and misonidazole together sensitized to a greater extent than oxygen and gave results similar to the effect found for diamide in the presence of oxygen. Hydrogen Donors and Chemical Repair of Radiation Damage The pretreatment effects with misonidazole (25, 35) and the combination studies with the thiol oxidant diamide (31-34) suggest that endogenous reducing species are important in the overall mechanism of radiation damage and repair. As mentioned previously, the reaction of the DNA radicals with hydrogen donor or reducing substrates will result in chemical repair of the radiation damage. Some of the intracellular hydrogen donors that are capable of reacting with radicals are reduced flavins, reduced pyridine nucleotides, ascorbate and thiols. The largest concentration of intracellular reducing materials is the protein thiols, the most important tow molecular nonprotein thiols being glutathione. Intracellular glutathione is believed to be an important hydrogen donor under anaerobic conditions for repair of radiation induced radicals in DNA. However, it cannot compete effectively with oxygen in air at low cell densities because the rate of reaction of oxygen with DNA radicals is at least an order of magnitude greater than the reaction rate with thiols (36). In addition, under the experimental conditions usually used for studies on the radiation response of cells, the cell density is usually quite low: 10,000 or less. If we assume that there is 5 nmoles NPSH/109 cells, then 10,000 cells would contribute approximately 5 X 10~,! moles of NPSH. This compares to an oxygen tension near 2 X 1CT4 mm or 2 X 10-7 M. Even if Figure 17. Effect of diamide on the survival of V79-S171 cells irradiated in air or nitrogen in the presence of 20 ^A4and 40 fj.m diamide. The apparent OER is 3.3. Survival curve parameters were estimated by eye; each point is the mean of at least two experiments (32). we allow for a factor of 10 for the contribution of the protein thiols as radioprotectors the thiol concentration is still only 5 X 10_1 M. Oxygen would be in great excess compared to either protein or nonprotein thiols. Cellular Thiols and Radiation Response There have been many attempts to remove the endogenous radioprotecting species with agents such as N-ethy!maleimide (NEM) (37) and diamide (2, 32, 38). In the case of diamide (38) there is a spontaneous chemical reaction with the cellular non-protein thiols as well as with cellular reduced pyridine nucleotide (Fig. 16). This reaction can be utilized to an advantage during hypoxic conditions and has been found to increase the radiation response (32). Diamide (Fig. 17) was found to radiosensitize hypoxic Chinese hamster cells by decreasing the shoulder of the survival curve at low concentrations and increasing the slope at high concentrations. The effect on the shoulder appears to be due to oxidation of endogenous nonprotein sulfhydryls and reduced pyridine nu- Volume 58 Number 2 February

9 cleotides (NADPH + NADH) and a partial oxidation of the protein thiols (39), biochemicals that would normally effect rapid chemical repair of certain single hit-type lesions. The slope effect, on the other hand, may have been due to reactions with DNA similar to those that have been described for the electron-affinic compounds. In addition, cells pretreated with diamide and maintained at 0 C remained sensitized after the removal of diamide. At zero degrees the metabolic regeneration of the thiols and pyridine nucleotides was inhibited (32). Diamide also sensitizes in air (Fig, 18, Refs. 33, 34). Another thiol-binding reactive agent that has been tested as a radiosensitizer of cells is NEM (37). NEM is quite effective at low concentrations as a sensitizer of cells under oxygenated conditions (37). NEM, unlike diamide, is less specific for cellular nonprotein thiols. It reacts with both protein and nonprotein thiols and DNA. The reaction with protein thiols alters enzyme activities believed to be involved in the repair of radiation damage. It has been suggested that the primary effect of NEM on the radiation response of cells is due to the removal of a Q factor involved in the repair of radiation damage. Presumably, this Q factor is a sulfhydryl-containing enzyme (37). It is interesting to note that other workers claim to have a glutathione deficient mutant human cell line that does not show the oxygen enhancement ratio (41) for DNA breaks. This cell line has the same non-proteinthiol content relative to other cells. The deficiency in glutathione appears to be in the final step in biosynthesis of the tripeptide. The reported effects on DNA breakage with these cells may be due to a deficiency of glutathione for a glutathione requring enzyme involved in the repair of the radiation damage (41). Figure 18. Whole survival curves for Chinese hamster cells irradiated under conditions exhibiting near maximum radiosensitivity and near maximum radioprotection (33). Radioprotectors and Radiation Response Addition of chemicals to the culture medium prior to irradiation of the cells has resulted in either protection or sensitization of the cells. In an unique publication, based in part on a long series of observations, Chapman and his colleagues demonstrated such effects (33). Figure 18 shows survival curves for Chinese hamster cells irradiated under conditions exhibiting near maximum and minimum radiosensitivities, obtained with the radiosensitizer diamide under oxic conditions and the radioprotector cysteamine under hypoxic conditions. Survival curves for cells irradiated under air-saturated and acutely hypoxic conditions have been included for comparison. These results show that the radiosensitivity of airsaturated cells is not a maximum and that radiosensitivity of cells made acutely hypoxic is not a minimum. The interpretation of these results according to a free radical model led Chapman to suggest that the cellular target environment is an important component in the expression of potentially lethal free-radical damage in the cellular target(s). Chapman (39) found that dimethylsulfoxide (DMSO) protected against radiation damage by competing with cellular targets for -OH and had no effect on the radical repairing of radical-fixing species within the cell, t-butanol, while not as effective a radioprotector as DMSO, also had no effect on the environment near the target molecules. Radioprotection by cysteamine (Fig. 18) as well as by other thiols (33,40) such as dithiothreitol does not appear to be the result of -OH scavenging although thiols are quite reactive with this radical (Fig. 26) (20-21). It is believed that cysteamine protection reflects the total amount of radical damage in cellular targets which can be chemically repaired by hydrogen donating species (resulting in enhanced cell viability). The thiol protectors increase the hydrogen donating species in the target area (Fig. 26). In his careful studies, Chapman suggests that approximately 82% of the radiation inactivation (cell death) measured for air-saturated cells is due to the fixation of target radicals (62% of the target radicals are produced from -OH and 20% from the direct effect) by oxygen (approximately 65%) and Figure 19. Effect of nitrotoluene on the oxygen consumption ot V79 cells (43). other endogenous electron affinic substances (approximately 17%). The remaining 18% of cellular inactivation or cell death may result from irreparable or lethal damage to cellular targets) by direct action. Moreover, the extent to which the indirect action of eaq~ and -H contribute to the remaining cell inactivation is not known. There is at the present time no comprehensive mechanism that will account for all of the effects on ionizing radiation on mammalian cells (2). Inhibition of Cellular Oxygen Utilization and Radiosensitization of an In Vitro Tumor Model It was realized early in the development of radiosensitizing drugs that the ability to sensitize hypoxic cells to ionizing radiation could not be completely related to their electron affinity or one-electron acceptance capacity (27). The more electron affinic compounds are also more active metabolically (26, 27, 29, 30) and one of the more important metabolic effects is alteration of oxygen utilization (27). Such an effect is obviously unimportant under totally anoxic conditions, as have been used previously to determine radiosensitizing ability (20-23). However, the effects of drugs on oxygen utilization becomes important in tumor-like situations where 152 Journal of Chemical Education

10 Figure 20. (A) Effects of nitrotoluene on plating efficiency {top) and survival {bottom) (after 2100 rad) of cells recovered from treated spheroids. Broken lines show fully reoxygenated cells. (B) Survival of cells from irradiated 582 spheroids pretreated 1 hr with 2.0 mm NT. Survival is expressed relative to untreated spheroid celfs at time zero (43)- Figure 21. Effect of dinitrobenzonitrile on the oxygen utilization of V79 cells (43). Figure 22. (A) Effects of dinitrobenzonitrile on plating efficiency (top) and survival (bottom) (after 2100 rad) of cells recovered from treated spheroids. Broken lines represent control survival when hypoxic cells were reoxygenated. (B) Survival of cells from 582 /^m spheroids pretreated with 0.5 mm DNBN for 1 hr prior to irradiation (43). hypoxic conditions result because the oxygen utilization of the cells closest to the blood supply exceeds the rate of oxygen diffusion to more distant cells (22). One system that can be utilized as a model to demonstrate such respiratory effects in vitro is the multicellular spheroid system as developed by Sutherland and Durand (42). This system is based on the fact that a number of animal as well as human tumor cells can be grown in suspension culture in vitro in clusters consisting of many cells. In these clusters the consumption of oxygen by the outermost cells decreases the availability of oxygen to the innermost cells, thereby limiting the depth to which oxygen can penetrate in the spheroid. This results in an inner core of hypoxic cells that are radioresistant, because of lack of oxygen. These clusters of cells can be trypsinized and radiation survival curves obtained (Fig. 20, 22). With this system Biaglow and Durand (43) were able to demonstrate that relatively poor sensitizers, such as the nitrobenzene derivative nitrotoluene, were more effective sensitizers of the spheroids than the more electron affinic dinitrobenzonitrile known to have a potent effect on the radiation response of hypoxic cells irradiated as monolayer or attached cells (43). Nitrotoluene was a more effective sensitizer because it is a potent inhibitor of cellular oxygen utilization (Fig. 19). A nearly log-linear response was found between nitrotoluene concentration, starting as low as 1.5 X 10-4 M, and inhibition of oxygen utilization (43). Inhibition was 85% with 2 mm and 90% with 10 mm nitrotoluene. No recovery from this inhibition was observed for periods up to 1 hr in the presence of drug. While 10 mm nitrotoluene proved to be slightly toxic when incubated for periods greater than 90 min, 0.4 mm was not toxic. The same concentrations of nitrotoluene were tested for effects on the radiation response of the multicell spheroids (Fig. 20A). At all concentrations tested, the survival after 2100 rads was drastically reduced immediately after addition of the drug. To determine whether this was a true radiosensitizing effect or metabolic phenomena caused by the effects of nitrotoluene on oxygen utilization, complete survival curves for the spheroids in the presence and absence of drug were generated (Fig. 20B). Spheroids exposed to the nontoxic concentration of 2 mm nitrotoluene for 1 hr prior to radiation had survival curves that were essentially indistinguishable from those of fully reoxygenated spheroids, obtained by pre-irradiation trypsinization of the spheroids into single cells. Thus, in the partially hypoxic spheroid system, the inhibition of cellular oxygen utilization by nitrotoluene (Fig. 19) sensitized the spheroid to almost the same degree as did molecular oxygen (Fig. 20B), demonstrating that under simulated in vivo conditions alterations in the radiation response can be produced even by a drug known to be a relatively poor anoxic sensitizer (43) if it is an inhibitor of cellular oxygen utilization (27). Dinitrobenzonitrile, a moderately good anoxic radiosensitizer stimulated oxygen utilization (43) at concentrations as low as 30 um (Fig. 21). An increase in the rate by a factor of 3, or 200%, occurred with DNBN concentrations in excess of 1-2 mm. The pronounced stimulation occurred immediately after addition of drug and declined with time. However, at 45 min, all the cells still consumed oxygen at rates greater than did the controls (Fig. 21). All concentrations of dinitrobenzonitrile applied prior to radiation initially protected against 2100 rad, as seen by the increasing surviving fraction (Fig. 22A). The greatest protection, as well as the largest effect on oxygen utilization occurred after addition of 5 mm dinitrobenzonitrile. The radiation protection with 5 mm drug diminished with time, becoming zero at 45 min; at 90 and 180 min, measurable radiosensitization occurred (43). Sensitization at longer incubation time, may be due to either cytotoxic effects or increased metabolic consumption of thiols known to protect against radiation damage. Complete radiation survival curves were obtained after exposure to nontoxic concentrations of 0.5 mm dinitrobenzonitrile for 1 hr (Fig. 22B). The radiation response measured demonstrated a slight degree of radioprotection when compared to the response observed with untreated spheroids. Even greater protection was observed initially with higher dinitrobenzonitrile concentrations. To compare the consequences of stimulation or inhibition of oxygen utilization within a single population of spheroids, a composite curve was constructed by plotting the survival observed when addition of nitrotoluene or dinitrobenzonitrile immediately preceded a radiation dose of 2100 R, against the Volume 58 Number 2 February

11 b Figure 24, The effect of different concentrations of rotenone, antimycin A, oligomycin, and nitrotoluene on the oxygen consumption of V79 lung cells, 3.0 X 107 cells/ml, in complete growth medium buffered with 0.02 M HEPES. Each point represents a duplicate measurement (from the same preparation) at ph 7.4 and 37 C. DMSO, at the concentrations used, had no effect on respiration (46). Figure 23. Survival of spheroid cells treated with NT or DNBN immediately prior to receiving 2100 rad, as a function of the relative oxygen utilization rate for each drug treatment. Solid line represents the predicted survival for spheroids consuming 02 at the indicated rates (43). As a first-order approximation, the depth to which oxygen will penetrate the spheroid is proportional to the inverse square root of the oxygen consumption rate. Histological preparations allow an estimate of the size of the necrotic region. The relative volume of hypoxic cells can easily be estimated if it is assumed that they are located predominantly near the necrotic region of the spheroid. Therefore, for spheroids with radius a {291 ^im), necrotic core radius b (72 ^tm), and critical" oxygen penetration depth r, the hypoxic fraction Fis given by F = (a r)3 &3]/{a3 b3) for a > r 0. The spheroids, when totally oxygenated, had a surviving fraction Sox = after 2100 rad; if totally anoxic, the survival was SAN = 0.129, Hence, the net survival is a function of the fraction of cells that are anoxic, i.e., S = (1 F)S0X - + F SAN. Since the oxygen penetration (r) varies with rate of consumption K so that r = ro(k)1/2, where ro represents the penetration depth under normal = conditions (observed survival ), substitution into the equation for S defines ro, and hence allows calculation of S as a function of K (43). measured respiratory rate (Fig, 23). A fairly straightforward calculation leads to a prediction of response as a function of cellular respiration rate (dotted line) and enables a comparison to be made between the observed and theoretical responses (22). The consistent overestimate of survival at the various DNBN concentrations may indicate a degree of hypoxic cell sensitization by this compound or some other metabolic effect as already described. For nitrotoluene a similar tendency was only apparent at the lowest concentration tested. The results with spheroids indicate that radiosensitizing chemicals which alter oxygen utilization cannot be tested adequately for potential usefulness in radiotherapy when studied in a hypoxic single-cell systems. When oxygen is available to cells distant from the vascular supply (metabolic barrier) the degree and extent of cellular and tissue hypoxia are critically dependent upon the rate of oxygen utilization. Many electron affinic drugs, including some of the nitrobenzenes and nitrofurans, have been shown to stimulate cellular oxygen utilization (27) or to stimulate oxygen consumption by reaction with ascorbate (45). This property may alter the effectiveness of these drugs as sensitizers in tumors, since they may produce an increased hypoxic region, as they do in the case of spheroids, and thereby reduce any effect that they may have as radiation sensitizers of the hypoxic and anoxic cells (20, 23, 27, 43). In an extension of these studies Durand and Biaglow investigated the radiosensitizing effects of more potent inhibi- 154 Journal of Chemical Education Figure 25. The response of spheroids to rotenone and irradiation. Panel (A) shows the response to rotenone alone (top) or to 2100 rad of 7-rays after exposure to the drug for different times. The dotted lines indicate the response of intact or dissociated spheroids to 2100 rad alone. Complete dose-survival curves for the same drug concentrations present 30 min before and during irradiation are shown in panel (B( (46). tors of respiration. The effects of the classical inhibitors of mitochondrial respiration on the oxygen consumption of V79 lung cells in suspension are shown in Figure 24. The concentrations required for 50% inhibition were rotenone (3 X 10-8 M), antimycin A (6 X 10~7 M), oligomycin (7 X 10-5 M), and nitrotoluene (5 X 10-4 M). Clearly, rotenone is the most effective of these classical inhibitors of cellular oxidation on a molar basis, and all were considerably more effective than nitrotoluene (Fig. 19). All of the drugs that inhibited cellular oxygen consumption were found to be sensitizers for the spheroid system (45). Figure 25 shows the results obtained with rotenone. By varying the time of incubation with drug prior to irradiation (Fig. 25A), the kinetics of reoxygenation of the hypoxic cells in the spheroid can be estimated (46). With the higher concentration of rotenone a large amount of reoxygenation occurred almost immediately, while less sensitization was observed immediately after the lower concentration. Less sensitization was observed with 1.2 X 10-8 M rotenone, as would be expected on the basis of the respiration data. Suspensions of aerobic and anoxic V79 cells from single-cell cultures showed no modification of the radiation response by the same rotenone concentrations (46). These data indicate that oxygen-sparing mechanisms are very efficient in sensitizing hypoxic cells within a spheroid to radiation. However, such potent inhibitors of respiration cannot be used

12 diagnose disease, is a waste by-product from medical and other industries as well as from nuclear power plants. Solar radiation penetrates our bodies everyday, and we have natural radioisotopes in the environment. In addition, space flights may be limited because of the continuous solar bombardment by gamma rays to which astronauts are exposed. The chief problem with human exposure to radiation is the production of mutational events that may lead to malignancies. There- Figure 26. The relationship between metabolism and chemicals with respect to DNA damage and repair. in vivo because they would inhibit the oxygen utilization of every cell in the body, producing death. A more attractive approach is to utilize drugs that concentrate in tumor tissue and are also inhibitors of cellular oxygen utilization (46-49). Another approach that we have been concerned with over the last few years is to utilize physiological controls or tumor enzymes that would release respiratory inhibitors such as cyanide (47) in the tumor in order to produce the oxygen-sparing effects. We have found that exploitation of the Crabtree effect, namely the inhibition of respiration by glucose, is a potential physiological means for producing oxygen-sparing effects. The Crabtree effect is nearly a universal phenomena occurring with most tumor tissue in vitro and has been recently shown to occur in vivo. The glucose effects are innocuous to the surrounding tissue and the metabolic transient alteration is transient with rapid return to normal. The metabolic use of glucose in vivo has shown some promise alone and in combination with breathing oxygen (47-49). Summary Depicted in Figure 26 are the various places where drugs may influence the radiation effects on the target molecule DNA as well as its subsequent chemical and enzymatic repair. We have also indicated the sites where metabolism may be important with respect to the cellular content of endogenous hydrogen-donating species and to the availability of oxygen consumed by respiration. The effects of radiation on DNA can be modified by hydroxyl radical scavengers in air and by hydrogen donating thiols under aerobic and anaerobic conditions. The -OH radical scavengers, such as DMSO, alter the indirect effect of radiation by reacting with -OH before it reacts with DNA. Thiols can react with either the -OH radical or add to the pool of hydrogen donating species within the cell that can chemically repair the damaged DNA. The pool of hydrogen-donating species can be removed by preincubation of cells with nitro compounds or by the oxidizing agent diamide. NEM is also quite effective in reacting with thiols. The major hydrogen donating thiol species within the cell are the nonprotein and protein thiols. Radiation or drug oxidized thiols are regenerated back to the reduced thiol via glucose linked metabolism. Oxidation of protein thiols can result in enzyme inhibition and a decreased capacity to repair DNA damage. On the left-hand side of Figure 26 is seen the ability of oxygen or electron affinic drugs to react with the target radical to produce drug or oxygen adducts as well as DNA breaks. Some of this damage can be repaired enzymatically resulting in an increase in cell survival. Oxygen is the chief radiosensitizer and it is rapidly depleted in vivo and in vitro in multicellular systems by cellular respiration. Many drugs can stimulate or inhibit the oxygen consumption resulting in radiation protection or sensitization. Conclusion The effects of ionizing radiation on mammalian cells remain viable area for investigation because of the influence of radiation on our lives. Radiation is used to treat cancer, to a fore, understanding the effects of radiation on living things is necessary in order to define and limit the unnecessary risks due to high level exposure than may inadvertently occur through misuses of radiation. The understanding of the basic mechanisms for radiation damage and protection may lead to the design of drugs that will greatly reduce the risk from high or low level exposure, improve the response of human tumors to ionizing radiation with respect to this later possibility; the protection of normal tissue during the irradiation of tumor tissue is being actively pursued in order to deliver more radiation to the tumor. Acknowledgment I would like to thank Dr. Marie E. Varnes and Mrs. Birgit Jacobson for reading and correcting this manuscript prior to its completion. The work on this review was supported by Grant No. CA-13747, awarded by the National Cancer Institute, DHEW. Literature Cited (1) Fogh, J., "Human Tumor Celia in Vitro," Plenum Press, New York, (2) Alper, T., "Cellular Radiobiology," Cambridge University Press, Cambridge, (3) Elkind, M. M. and Sinclair, W. K., "Current Topics in Radiation Res., North Holland Publishing Co., Amsterdam, 1965, Vol. 1, p (4) Puck, T, T. and Marcus, P. 1, J,, Exp. Med., 103,653 (1956). (5) Hall, E. J., Radiobiology for the Radiologist," Harper, New York, (6) Grosch, D. S. and Hopwood, L. E., Biological Effects of Radiations," 2nd Ed., Academic Press, New York, 1979, (7) Salmon, S. E., Hamburger, A. W., Soehnlen, B., Brian, B. S., Durie, G. M., Aherts, D. S. and Moon, '1'. E., New Eng. J of Med., 298,1321 (1978). (8) Jakoby, W. B. and Postan, J. H., (Editors, Methods in Enzymology" Vol. LV1II, Cell Culture, Academic Press, New York, (9) Pryor, W. A., (Editor), "Free Radicals in Biology, Vols. 1-5, Academic Press, New York, (10) Gregg, E. C., "Effects of ionizing Radiation in Humans," Chem. Rubber Co. Reviews, 1980, in press. (11) Elkind, M. M. and Sutton, H., Nature, 184,1293 (1959). (12) Gray, L. H., Conger, A. D,, Ebert, M., Hornsey, S. and Scott, O. C. A., Brit. J. Radiol,, 26,638(1958). (13) Flanders, H. P. and Moore, D.,/fad. Res., 9,422 (1958). (14) Ward, J. F,, Ado. Radiation Biol., 5, 181 (1977). (15) Painter, R. B., "Radiation Biology in Cancer Research," R. E. Meyer and H. R. Withers, (Editors), Raven Press, New York, 198ft, p. 59. (16) Roots, R. and Okada, S., Rad. Res., 64, 306 (1975). (17) Epp, E. R., Weiss, H., Ling, C, C,, Current Topics in Radiation Res. Quarterly, 11, 201 (1976). (18) Millar, B. C,, Fielden, E. M., and Steele, L., Int. J. Radiat. Biol., 36, 177 (1979). (19) Okada, S., "Radiation Biochemistry, Vol. 1 and Vol. II, Academic Press, (20) Adams, G. E. and Jameson, D., Radiat. Environ. Biophys., 17,95 (1980), or Adams, G. E. and Wardman, P.T Free Radical and Cancer, Pryor, W. A., (Editor), Academic Press, New York, 1977, Vol. 3, p. 53. (21) Adams, G. E., Radiation Sensitizers for Hypoxic Cells: Problems and Prospects in Treatment of Radioresistant Tumors, M. Abe, K. Sakamoto and T. L. Philips, Editors. Elsevier/North Holland Biomedical Press, (22) Fowler, J. F., Adams, G. E., and Denekamp. J., "Radiosensitizers of Hypoxic Cells in Solid Tumors," Cancer Treatment Reu., 3, 2277 (1976)(23) Becker, F. F., Editor, Cancer, A Comprehensive Treatise," Vol. 6, Radiotherapy, Surgery and Immunotherapy, Plenum Press, New York, (24) Gray, L. H. Conference on Hypoxic Cell Sensitizers in Radiobiology and Radiotherapy, Brit. J. Cancer, Suppl. Ill (1978). (25) Hall, E. J. and Biaglow, J. E., Int. J. Rad. Oncol. Biol. Phys., 2, 521 (1977). (26) Biaglow, J. E., Proceedings of the International Conference on Oxygen and OxyRadicals in Chemistry and Biology, (1980), in press. (27) Biaglow, J. E,, Inter, Encyclo. Pharm. Ther., (1980), in press. (28) Brady, L., "Proceedings of the Conference on Combined Modality Cancer Treatment: Radiation Sensitizers and Protectors," Masson Press, New York, 1980, in press. (29) Biaglow, J. E., Rad. Res., (1980), in press. (30) Biaglow, J. E., Varnes, M. E., Koch, C, J. and Sridhar, R., in Free Radicals and Cancer," R. Floyd, (Editor), Marcel Dekker, New York, 1980, in press. (31) Varnes, M. E., Biaglow, J. E., Koch, C. J. and Hall, E. J., Conference on Combined Modality Cancer Treatment: Radiation Sensitizers and Protectors, Brady, L., (Editor), Masson Press, New York, 1980, in press. (32) Harris, J. W. and Power, J. A., Rad. Res., 56, 97 (1973). (33) Chapman, J. D.,Rad. Res.. 56,291 (1973). (34) Watts, M. E., Whillans, I). W. and Adams, G. E., Int. J. Radiat. Biol., 27, 259 (1975). (35) Whitmore, G. F., Gulyas, S. and Varghese, A., Brit. J. Cancer, 37, Suppl. Ill, 115 (1978). (36) Greenstock, C. L. and Dunlop, 1., Fast Processes in Radiation Chemistry and Biology, Adams, G. E., Fielden, E. M., Michael, B. D., (EdiLors), J. Wiley and Sons, London, 1975, p (37) Sinclair, W. K. Radiation Research, Biomedical, Chemical and Physical Perspectives, Volume 58 Number 2 February