To determine whether different agents that enhance the expression of potentially lethal X-ray damage (PLD) interact with the same or different lesions (or spectrum of lesions), cell killing was measured in three kinds of experiments: (1) When cells were irradiated in G 1 phase and treated with caffeine or hydroxyurea at concentrations that yield maximal response, the same survival plateaus were reached. (2) Treatment of cells irradiated in G 1 phase either with caffeine or with hydroxyurea so as to yield survival levels that differed twofold after 4 h incubation, followed by treatment with caffeine to allow expression of PLD in G 2 phase, resulted eventually in the same level of survival. (3) When cells were irradiated and treated with caffeine, hydroxyurea, or 9-β-D-arabinofuranosyladenine (araA) after progressively longer delays, to trace the time course of recovery from the PLD, the responses obtained with caffeine and araA were similar; sensitivity to hydroxyurea was lost more rapidly. The results are consistent with the possibility that these three agents interact with the same lesions, but that different steps in the repair process are inhibited by caffeine or araA than by hydroxyurea.

The duration of the cell cycle in synchronous cultures of HeLa S3 cells that were either irradiated with 3.5 Gy of 220-kV X rays in mid-S phase or treated in early G 1 or mid-S phase for several hours with 1 or 3 μM aphidicolin, or were subjected to both treatments, was measured by time-lapse cinemicrography. When compared with the generation time of untreated cells, the delay in cell progression with the combined treatment was found to be less than the sum of the delays with the individual treatments, but longer than the imposed delay caused by treatment with aphidicolin alone. Because recovery from potentially lethal radiation damage proceeds in the presence of aphidicolin, this finding suggests that a portion of the radiation-induced delay in cell progression may be associated with processes other than those that directly affect cell viability. It was also observed that the incidence of both spontaneous and radiation-induced sister-cell fusion is decreased in cultures incubated in the presence of aphidicolin.

HeLa S3 cells were sensitized to the lethal action of 220-kV X rays by partially replacing the thymidine in their DNA with 5-bromodeoxyuridine (BrdU). To examine the expression of and recovery from potentially lethal radiation damage (PLD), both BrdU-grown and control cells were treated with 4 mM caffeine for increasing times up to 2 days, either immediately after irradiation or after increasing delays up to 28 h. When the same dose of X rays (3 Gy) was applied to BrdU-grown and control cells, the difference in survival that is found in the absence of caffeine disappeared after about 30 h of incubation in its presence; when isosurvival doses were applied (BrdU-grown cells, 2.5 Gy; control cells, 4 Gy), the control cells suffered more killing. When treatment with caffeine was delayed for progressively longer times after both groups of cells received 3 Gy, the control cells achieved a higher level of survival. These results indicate that the increased radiation sensitivity of cells containing BrdU derives from a decreased ability to repair PLD.

Hypothermic enhancement of the lethal effect of 3.5 Gy of 220-kV X rays in the absence of caffeine as well as in its presence (4 mM) was examined at temperatures between 10 and 34°C in monolayer cultures in the G 1 phase of the cell cycle. Correction has been made for the toxicity of low temperatures, and of caffeine at low temperatures, by concomitantly measuring cell killing in unirradiated cells. In the absence of caffeine, incubation of irradiated cells for up to 34 h at temperatures in the range 15 to 30°C (or possibly 34°C) enhances killing compared to that observed at 38°C; the amount of enhancement is about the same throughout this range, but is nil at 10°C. The enhanced killing induced by caffeine at 38°C decreases as the temperature is lowered to 15°C; there is no enhancement at 10°C. Less killing is manifested in the range 15 to 25°C in the presence of caffeine than in its absence. Recovery (loss of sensitivity to caffeine) and fixation of potentially lethal damage were studied in $\text{late-}{\rm S}/{\rm G}_{2}\text{-phase}$ cells at reduced temperatures by delaying treatment with caffeine for increasing times after irradiation. As the temperature is progressively lowered to 20°C, less recovery is manifested after 5 h of incubation; no recovery is detected in the range 10 to 20°C. Despite extensive recovery at 34°C, no fixation is observed at that (or any lower) temperature in G 2 -phase cells: the cells are able to recover essentially fully when returned to 38°C. In addition, responses of unirradiated control series to incubation at low temperatures appear to differ from those reported by others for longer treatment times of different cell systems.

Caffeine-mediated enhancement of the killing of V79 cells by 220-kV X rays at various times in the cell cycle was compared with that of HeLa cells by measuring (i) the dependence of cell survival on the duration of treatment with 5-10 mM caffeine, (ii) the effect of caffeine treatment on the X-ray dose-survival curve, and (iii) the loss of sensitivity to caffeine as a function of time after irradiation. The behavior of V79, while similar in many respects to that of HeLa (reported previously), differs in several ways. Caffeine treatment causes rapid killing immediately after irradiation irrespective of cell age, while HeLa is refractory in S phase and highly sensitive in G 2 . As with HeLa, the (multitarget) dose-survival curve parameters are reduced by caffeine treatment, but the age-dependent fluctuations in D 0 are not eliminated as completely as with HeLa, and the extrapolation number assumes values less than unity in the latter part of the cycle rather than in the early part. Loss of sensitivity to caffeine after irradiation early in the cycle appears to undergo a transient reversal in the middle of the cycle, a phenomenon not observed in HeLa.

Recovery from potentially lethal radiation damage in HeLa S3 cells has been studied by irradiating synchronous cultures with 4 Gy at selected ages in the cell cycle, initiating treatment with 4 mM caffeine, which prevents recovery, at progressively later times up to 24-30 h after irradiation, and determining the plateau level of survival after incubation with the caffeine until 36-40 h after mitotic collection. Cell recovery appears to begin immediately after irradiation at any time during interphase: an accelerating increase in survival gives way after several hours to a linear increase which lasts for an additional several hours. The median recovery time is ∼13 h after irradiation at any time during G 1 , but is markedly shorter (5-7 h) after irradiation in S or G 2 . The rate of recovery is slightly depressed if DNA replication is inhibited with aphidicolin after irradiation and slightly enhanced if protein synthesis is inhibited with cycloheximide. Both the rate and the extent of recovery are dependent on the location of the cells in the cycle at the time of irradiation-both functions increasing with cell age from the beginning of S, but having different age dependencies in G 1 . Blocking cell progression with a DNA-synthesis inhibitor before irradiation halts the age-dependent changes.

HeLa cells irradiated with 2 Gy of 220-kV X rays suffer a 60-70% loss of colony-forming ability which is increased to 90% by postirradiation treatment with 10 mM caffeine for 6 hr. The detailed postirradiation patterns of cell death and sister-cell fusion in such cultures and in cultures in which the colony-forming ability was brought to about the same level by treatment with a larger (4 Gy) X-ray dose alone or by longer (48 hr) treatment with 10 mM caffeine alone were recorded by time-lapse cinemicrography. Because the patterns of cell death and fusion differ radically in irradiated and in caffeine-treated cultures, the response of the additional cells killed by the combined treatment can be identified as X-ray induced rather than caffeine induced. The appearance of cultures after several days of incubation confirms the similarity of the post-treatment patterns of proliferation in cultures suffering enhanced killing to those occurring in cultures treated with larger doses of X rays alone. It is concluded that X rays do not sensitize cells to caffeine, but rather that caffeine enhances the expression of potentially lethal radiation-induced damage.

Postirradiation treatment of synchronous HeLa S3 cultures with 4 mM caffeine until ≥32 hr after mitotic collection, following exposure to 220-kV X rays at various times during interphase, severely damps the fluctuations in the age-survival curve. Not only does the dose-survival curve essentially lose its shoulder, as reported previously, but it becomes steeper and displays a virtually age-independent terminal slope (<tex-math>$D_{0}\simeq 0.5$</tex-math> Gy). It becomes multicomponent, at least early in the cycle. The residual structure in the interphase age-survival curve, if any, appears to reflect mainly an age-dependent fluctuation in the size of a subpopulation of cells having marked sensitivity to X rays (<tex-math>$D_{0}\simeq 0.25$</tex-math> Gy), though there might be small residual fluctuations in the size of the shoulder and the slope. Mitotic cells also respond to postirradiation treatment with caffeine; they yield a dose-survival curve whose slope is similar to that of the sensitive subpopulation seen in interphase. These findings indicate that most of the structure in the unperturbed age-survival function derives from repair of potentially lethal radiation damage.

When HeLa S3 cells are irradiated in early G 1 with 4 Gy of 220-kV X rays and are then incubated in growth medium containing up to 5 mM caffeine, survival is reduced (as reported previously), reaching a concentration-dependent plateau. Cell killing presumably occurs as a result of the fixation of a portion of the potentially lethal damage the cells contain. These cells respond to continued treatment with caffeine at concentrations greater than 2 mM during S, but less so than during G 1 . When they reach G 2 arrest, however, extensive cell killing again occurs (reported previously), presumably also the result of potentially lethal damage fixation. ${\rm G}_{1}\text{-irradiated}$ cultures that are treated with caffeine either continuously at a concentration in the range 1-5 mM, or at 10 mM for 8 hr and subsequently with the lower concentration, achieve the same survival level in G 2 , provided that the potentially lethal damage is not repaired during G 1 and S. Repair seems to be completely inhibited in the presence of 3-4 mM caffeine. The results indicate that fixation of potentially lethal damage occurs in the same sector of cells in G 1 and G 2 , suggesting that the same cellular lesion gives rise to cell killing in the two phases.

Measurements of DNA replication in a line of HeLa S3 cells during the generation (Generation 1) following that in which the cells are irradiated with 500 rad of 220-kV X rays (Generation 0) were carried out according to a number of different experimental protocols. These involved preirradiation labeling of the cells with low levels of [ 14 C]thymidine in Generation -1 to provide a measure of the template DNA, synchronization by mitotic collection in Generation 0, resynchronization by either mitotic recollection or temporary drug-induced blockades in Generation 1, and either labeled-thymidine incorporation or density-label transfer during Generation 1. The results show that those cells that progress through S phase of Generation 1 and divide at the next mitosis replicate a full complement of DNA, within the approximately 10% uncertainty of the experimental methods. However, apparent deficits of as much as 45% are measured if resynchronization in Generation 1 is effected by drug treatment following manipulations of the culture in the presence of particular media and drugs during Generation 0. These are attributed to combined radiation- and drug-induced disturbances in cell progression.

The response of X-irradiated and unirradiated HeLa S3 cells to treatment with caffeine at concentrations between 1 and 10 mM has been examined with respect to both delay in progression through the cell generation cycle and enhancement of the expression of potentially lethal X-ray damage. Progression is delayed in a concentration-dependent fashion; the generation time is doubled at about 4 mM. The duration of G 1 is lengthened, and the rate of DNA synthesis is reduced, although the kinetics are different in the two phases: the rate of DNA synthesis is usually unaffected at 1 or 2 mM, while there is no concentration threshold for the slowing of progression through G 1 . Progression through G 2 appears to be unaffected by concentrations up to at least 10 mM. Killing of irradiated cells in G 2 is somewhat greater after treatment with the higher caffeine concentrations than reported previously for 1 mM. Moreover, an additional mode of killing is observed in irradiated G 1 cells which had been found previously to be only slightly affected by 1 mM caffeine; they suffer extensive killing at concentrations above 5 mM. The time-survival curves for irradiated, caffeine-treated G 1 and G 2 cells have characteristically different shapes. The dose-survival curves for cells treated with the higher caffeine concentration display steeper terminal slopes and narrower shoulders.

The ability of caffeine to enhance the expression of potentially lethal X-ray damage in HeLa S3 cells was examined as a function of the age of the cells in the generation cycle. Synchronous populations were irradiated at different times after mitotic collection and treated for various intervals with 1 mM caffeine, which causes negligible killing of unirradiated cells. The response was thereby determined as a function of cell age at both the time of irradiation and the time of exposure to caffeine. The amount of cell killing depends strongly on when in the cycle caffeine is present and only weakly on when the cells are irradiated. If cells are irradiated in early G 1 , caffeine treatment enhances killing for 2 to 3 hr. No additional enhancement is observed until 16 to 17 hr postcollection, corresponding to G 2 , here they enter a second period of much greater sensitivity. Similarly, fluorodeoxyuridine resynchronized cells irradiated during S and treated with caffeine suffer no enhanced killing until they pass into this sensitive phase in G 2 , approximately 7 hr after release from the fluorodeoxyuridine block. The sensitive period appears to coincide with G 2 arrest. The rate and extent of killing during this period are dependent upon the X-ray dose and the caffeine concentration. In the absence of caffeine, cells irradiated in G 1 lose sensitivity to caffeine in about 9 hr; they do so faster in G 2 . It is concluded that the potentially lethal X-ray damage expressed on treatment with caffeine is retained for many hours in the presence of caffeine and is maximally manifested by ${\rm G}_{2}\text{-arrested}$ cells.

Treatment of HeLa S3 cells with 1 mM caffeine delays progression through G1 by 1.5 hours but causes no other detectable inhibition of cell progression; it sometimes results in a large stimulation of thymidine incorporation. When this concentration is applied to cells that have been irradiated with 1-krad doses of 220-kV X rays, there is a marked suppression of both the inhibition of DNA synthesis and G2 arrest induced by the radiation. Larger doses require higher concentrations of caffeine to suppress the inhibition of DNA synthesis. Delaying addition until the rate of synthesis is at its minimum (1.5 hours after irradiation with 1 krad) results in a slightly accelerated recovery of the rate. Treatment before or during irradiation is without effect on the inhibition. Removal of the caffeine as late as 6 hours after its addition at the time of irradiation results in a prompt inhibition in DNA synthesis that mimics that observed immediately after irradiation in the absence of caffeine. These findings raise the possibility that the depression in rate of DNA synthesis might not result from radiation damage introduced into the replicon initiation system, but rather may be an indirect consequence of damage residing elsewhere in the irradiated cell.

Postirradiation treatment of HeLa S3 cells with 1 mM caffeine results in a marked diminution of the surviving fraction as scored by colony formation. The decrease is dose dependent; the effect of a 24-hour postirradiation treatment of a nonsynchronous population with caffeine is to change the terminal slope of the survival curve and its intercept. D 0 is reduced from 130 to 60 rad; the extrapolation number is increased about twofold. The amount of postirradiation killing is maximal if cells are exposed to caffeine at a concentration of at least 1 mM for 8 hours; less than 10% of unirradiated cells are killed under these conditions. Dose-response curves were also obtained for synchronous cells at various phases of the cell cycle. Similar results were obtained at all cell ages, but the magnitude of the effect is age dependent. This age dependence was further explored in experiments in which mitotically collected cells were exposed to 300 or 500 rad doses at 2-hour intervals throughout the cell cycle. Treatment with caffeine for 24 hours after irradiation enhances the killing of cells late in the cycle more than cells in G1. The sensitivities of two other cell lines, CHO and EMT6, also were examined; both are substantially less sensitive to caffeine. The smaller cell-cycle dependence of CHO cells is qualitatively the same as that of HeLa cells.

After irradiation of randomly dividing cultures of HeLa S3 cells with 220-kV X rays, the rate of DNA synthesis, measured by pulsed incorporation of labeled thymidine, falls nearly exponentially with time ( $t_{{\textstyle\frac{1}{2}}}\sim 1.3$ hr), in a dose-independent fashion. The fall is less rapid than that observed after addition of inhibitors of protein synthesis. With doses up to 8 krad, the rate reaches a minimum and begins to increase after 1-3 hr, the minima occurring at lower values and at slightly later times with increasing dose. The increase appears to be roughly linear for about 6 hr, with the slope an inverse function of dose in the range 1-8 krad. About 7-9 hr after the completion of irradiation, the rate again falls, although no more than 10% of the cells die sooner than 14 hr after irradiation with 8 krad (and later with smaller doses). Fluorodeoxyuridine-mediated delay in expression of the depression, described previously for doses up to 1 krad, occurs also at higher doses. During the period when the rate per culture rises, the rate in the individual cells, measured autoradiographically, appears to increase also, i.e., the rise presumably does not merely reflect populational shifts. The initial descending portion of the rate curve can be at least partially separated from the ascending portion by administering the total dose in suitably spaced fractions. If interpreted in terms of the model that attributes the initial depression in rate of synthesis to a temporary absence of replicon initiation, the results indicate that initiation is halted by an X-ray dose smaller than 1 krad; that it begins again after a dose-dependent delay amounting to about 0.7 hr after 1 krad and 1.5 hr after 7 krad; and that once begun, the rate of synthesis increases in a dose-dependent fashion. The second depression might derive from synchronization and/or from the imminence of cell death.

HeLa S3 cells arrested in G2 by a conditioning dose of 650 rad of 220 kV X rays delivered in early G2, were subjected to a second dose 10 hr after the first. In contrast to the survival response, which changes from that characteristic of late S cells to that of M cells early during arrest, susceptibility to additional delay is retained until the end of the arrest period.

Expression of the radiation-induced depression of the rate of DNA synthesis in HeLa S3 cells can be delayed for at least 6 hr by subjecting cultures to either low temperatures (5°C) or inhibitors of DNA synthesis (1 μM fluorodeoxyuridine or 4 μM cytosine arabino-side) at 38°C immediately after irradiation with 1-krad doses of 220-kV X rays. These findings indicate that full expression of, and recovery from, the damage manifested as depressed synthesis requires not only active metabolism, but DNA replication as well. They imply that the depressed rate does not derive from a decrease in availability of DNA precursors. Confirmatory evidence for this was obtained from measurements of both the rate of total thymidine uptake into the pool under conditions in which the endogenous pathway of thymidylate synthesis is either operating or blocked, and the amounts of thymidine mono-, di-, and triphosphates formed in the pool from thymidine, during the postirradiation period, when incorporation into DNA is depressed. The results are interpreted in terms of the recent identification of replicon initiation as the step in the replication process that is affected by irradiation.

The radiation-induced deficiency in DNA synthesis in Generation 1 was studied as a function of the age of HeLa S3 cells at the time of exposure to 220 kV X rays in the previous generation (Generation 0). The amount of DNA synthesized is dependent on the stage in the generation cycle at which cells are irradiated. The smallest deficiency (20-35% after a dose of 500 rad) is observed in cells irradiated in early G1 or early G2, while the greatest deficiency (55-70% after 500 rad) is found in cells irradiated at mitosis or at the G1/S transition. The high sensitivity of cells at G1/S is also manifested by a steeper dose-response curve. Cells irradiated in late G2, past the point where their progression is temporarily blocked by X rays, synthesize a normal amount of DNA in Generation 1, while cells that are held up in the G2 block exhibit deficient synthesis in the next generation. The extent of the deficiency in early G1 cells can be enhanced by treatment with 1 mM hydroxyurea for several hours immediately following irradiation. The possibility that deficient DNA synthesis is related to cell killing, and the relation between the G2 block and deficient synthesis, are discussed.