The mechanisms of hypoxic cell radiosensitization by oxygen and by the 2-nitroimidazole chemical sensitizer, misonidazole, were studied in CHO cells using rapid-mixing techniques. Both agents display a dose response of sensitivity in steady-state experiments which is satisfactorily fitted by a K function with maximum enhancement ratios (ER) of about 2.8. Rapid-mixing experiments demonstrated that oxygen sensitization in these cells is complete within about 10 msec even at concentrations as low as 2.6% oxygen, with no evidence for two time-resolved components as had been reported with V79 cells [Shenoy, M. A., Asquith, J. C., Adams, G. E., Michael B. D., and Watts, M. E. Radiat. Res. 62, 498-512 (1975)]. Full sensitivity persists for at least 165 msec following an 8.4-fold dilution into hypoxic medium. No large temperature dependence was observed in the range 4 to 37°C. Partial development of sensitization by misonidazole also occurs very quickly, within 25 msec with a 10 mM drug concentration, but only to an ER of about 1.7, much below the level observed after many minutes incubation (∼2.5) even at ice temperature. This development, too, has a very weak temperature dependence. It is suggested that although uptake of the drug is rapid, probably occurring by passive diffusion, an additional mechanism not involving gross metabolism and operating on the time scale of seconds to minutes is required to produce maximal sensitization in these cells.

A direct measurement of oxygen diffusion into mammalian cells in suspension has been made using a rapid-mixing technique and an oxygen-quenchable fluorescent probe. The emission from excited pyrene molecules located throughout the cytoplasm of Chinese hamster ovary cells, as determined by fluorescence photomicrography, is reduced in the presence of oxygen. This reduction in fluorescence intensity was monitored as a function of time following mixing of deoxygenated cells with oxygen-saturated phosphate-buffered saline (PBS) for times from 1 to 30 msec after mixing. When deoxygenated labeled cells were mixed rapidly with oxygen-saturated PBS, the half-time of quenching of the fluorescence was 3.7 msec. Analytical solutions were developed on the basis of diffusion into a spherical model cell and compared with experimental data. No effect of an unstirred sheath of water surrounding the cell could be detected, with a limit of sensitivity of about 1 μm. The data were well fitted on the basis of the diffusion model with a diffusion coefficient of <tex-math>$4\pm 2\times 10^{-6}\ {\rm cm}^{2}\ {\rm sec}^{-1}$</tex-math>. This value is fivefold lower than in water. The same value was inferred for diffusion in the local region of pyrene molecules through the equilibrium quenching constant determined from a Stern- Volmer relation. A comparison was made of these results with cell survival experiments when irradiations were performed within milliseconds of oxygen addition. Our data indicate that oxygen can diffuse to near-equilibrium values generally throughout a cell in only a few milliseconds.

A picosecond pulse radiolysis system has been used to study the electron attachment processes encountered in concentrated solutions (0.1-3 M). These highly concentrated solutions approximate the conditions of a biological cell. The reaction $e{}_{{\rm aq}}{}^{-}$ + scavenging molecule has been observed for a range of biologically important molecules. The rate constants depend upon the concentration of the solute molecules and the pH of the solution. As well, the initial yield of $e{}_{{\rm aq}}{}^{-}$ at 30 psec is reduced markedly in concentrated solutions of scavengers such as amino acids and mononucleotides; the initial process is too fast to be explained by normal chemical kinetics. The scavenger which is unique is the hydrogen ion, ${\rm H}{}_{{\rm aq}}{}^{+}$ , which causes no observable reduction of the $e{}_{{\rm aq}}{}^{-}$ yields, but has a normal rate constant of reaction with $e{}_{{\rm aq}}{}^{-}$ . With other electron scavengers, a general relationship between the efficiency of reducing the initial yield of $e{}_{{\rm aq}}{}^{-}$ and the rate constant, $k_{(e{}_{{\rm aq}}{}^{-}+S)}$ in concentrated solutions is obtained. This relationship indicates that both the $e{}_{{\rm aq}}{}^{-}$ and its precursor react with the same efficiency toward the solute molecules.

The technique of pulse radiolysis has been utilized to study the absorption spectra of the transient free radical adducts which result from reactions of ·OH and $e_{{\rm aq}}{}^{-}$ with the pyrimidine, uracil. The ·OH attaches to the 5,6 bond of uracil to form an adduct (·UOH) which has peaks at 390 nm and ∼240 nm. Changes in the spectrum of ·UOH with pH have led us to the conclusion that at least one isomer of ·UOH has ${\rm pK}_{{\rm a}}\text{'}$ s of ∼7 and 9.5. The ·UOH decays by a second order kinetics with rate constant, $2k=(1.1\pm 0.2)\times 10^{9}\ M^{-1}\ s^{-1}$ which is independent of ionic strength. There is a residual spectrum which remains after the decay of the ·UOH absorption. This residual spectrum is different at pH 5 and 8.5, and shows a similar pH dependence to that of ·UOH itself. The $e_{{\rm aq}}{}^{-}$ adds to uracil to form an adduct which has protonated ( $\cdot {\rm UH}_{1}$ ) and unprotonated ( $\cdot {\rm U}^{-}$ ) forms which are in equilibrium with a pK a of 7. The spectra of both $\cdot {\rm UH}_{1}$ and $\cdot {\rm U}^{-}$ both have peaks near 250 nm with shoulders near 300 nm. The spectrum $\cdot {\rm U}^{-}$ is more intense than that of $\cdot {\rm UH}_{1}$ . The rate constant for protonation of $\cdot {\rm UH}_{1}$ is $(5.7\pm 0.4)\times 10^{10}\ M^{-1}\ s^{-1}$ . Both $\cdot {\rm UH}_{1}$ and $\cdot {\rm U}^{-}$ decay by second order kinetics with rate constants, $2k=(2.5\pm 0.5)\times 10^{9}\ M^{-1}s^{-1}$ and $2k=(1.6\pm 0.3)\times 10^{9}\ M^{-1}s^{-1}$ respectively. Evidence is presented which indicates that a reaction between $\cdot {\rm UH}_{1}$ (or $\cdot {\rm U}^{-}$ ) and an absorbing product may occur. In Ar bubbled solution with no scavengers present, the observed spectrum agrees closely with the calculated sum of the spectra of the individual adducts. This spectrum also decays by second order kinetics with the same halflife at all wavelengths implying that the adducts decay by reaction with themselves and with each other.

The stable products resulting from the radiolysis of uracil in deoxygenated solutions have been characterized, and their yields have been measured. The G-value for the destruction of uracil varies from 4.6 at pH 5 to 1.5 at pH 9 in the presence of N 2 O, and from 3.3 at pH 5 to 0.7 at pH 9 in the presence of N 2 . In both cases, a plot of the G-value against pH follows a titration curve with a $pK_{a}$ of approximately 7. The G-value does not depend on pH if the solutions contain 0.2 M t-butanol or oxygen. We have isolated two classes of stable products: monomeric products and dimers. The first class of products includes uracil glycols, uracil hydrates, and dihydrouracil. The dimeric products, which have not been observed before, have a variety of structures, but they all consist of two saturated pyrimidine rings joined by a single bond. The yield of dimers is more pH dependent than the yields of the monomeric products; however, the dimers are the major products under all conditions studied. In addition, the properties of the dimers depend on the pH of the solution and on the gas present. We present a mechanism which explains (i) how the different types of products arise by reactions of the transient uracil adducts with themselves and each other, and (ii) how the state of dissociation of the transient uracil adducts causes the observed pH dependence of the G-values.

Reactions of the hydroxyl radical with various monomeric components of nucleic acid, and with polynucleotides, were studied by observing the transient absorption of the ·OH adduct species. At neutral pH, the rate constants for the attack of ·OH on monomers were of the order of <tex-math>$5\times 10^{9}\ M^{-1}\ {\rm sec}^{-1}$</tex-math>. The rate constants, per base, for ·OH attack on single-stranded polynucleotides were approximately a factor of 5 lower: these were <tex-math>$1.2\times 10^{9}\ M^{-1}\ {\rm sec}^{-1}$</tex-math>, <tex-math>$1.25\times 10^{9}\ M^{-1}\ {\rm sec}^{-1}$</tex-math>, and <tex-math>$0.9\times 10^{9}\ M^{-1}\ {\rm sec}^{-1}$</tex-math> for poly C, poly U, and poly A, respectively. The double-stranded polynucleotides poly (A + U) and DNA had rate constants, per base, for reaction with ·OH at pH 7, of 0.5 and <tex-math>$0.4\times 10^{9}\ M^{-1}\ {\rm sec}^{-1}$</tex-math>, respectively. Absorption spectra and values of extinction coefficients are presented for the ·OH adducts. It is shown that the absorbing transient product of ·OH attack is the same species in the polymers poly C, poly U, and poly A, as in the corresponding nucleotides CMP, UMP, and AMP. Kinetic and spectral data indicate that ·OH attacks a double-stranded polynucleotide at random base sites; the ·OH adduct spectrum of poly (A + U) is identical in shape and magnitude to a composite average of the poly U and poly A adduct spectra. Secondary and tertiary structure appears to have little effect on ·OH reactivity with polynucleotides.

The rate constants for the attack of ${\rm e}_{{\rm aq}}^{-}$ on adenine, cytosine, and several derivatives were measured in the pH range 6-14. The rate constants at neutral pH are $0.9\times 10^{10}\ M^{-1}s^{-1}$ and $1.3\times 10^{10}\ M^{-1}s^{-1}$ for adenine and cytosine, respectively. The rate constants, per base, for the attack of ${\rm e}_{{\rm aq}}^{-}$ on polynucleotides are much lower. At pH 7 they are $7.5\times 10^{8}\ M^{-1}s^{-1}$ , $2.5\times 10^{8}\ M^{-1}s^{-1}$ , and $1.3\times 10^{8}\ M^{-1}s^{-1}$ for poly U, poly A, and poly (A + U), respectively. It is shown that the low overall reactivity of polynucleotides results from a lowered collision frequency with ${\rm e}_{{\rm aq}}^{-}$ . Secondary structure is an important factor, responsible for the observed differences in the values of the rate constants for the polynucleotides studied. Poly (A + U), which has a double helical structure, is least reactive, and poly U, an unstacked random coil, is most reactive. The inherent reactivity of bases in a polynucleotide also determines the rate constant to some extent.

A screening technique which permits the rapid isolation of bacterial mutants sensitive to ionizing radiation is described. The technique utilizes the radiation sensitivity of the mutants as the property used in their detection. Aliquots of the cell suspension to be screened are plated on agar plates and are incubated at 37°C for approximately 10 hours. The plates are then irradiated with a dose of gamma radiation sufficient to reduce the survival of the parent population to between 10% and 50% and are incubated for a further 18 to 24 hours. Visual inspection of the plates at this time reveals, among the normal clones, the presence of a small percentage of minute clones. Any clones composed of radiosensitive cells will be found among these. A second rapid screening permits the true radiation-sensitive mutants to be isolated with ease. The utility of this screening procedure has been demonstrated by the isolation of a spontaneously arising radiation-sensitive mutant from a culture of Escherichia coli K12 AB1157. Some preliminary radiobiological properties of the isolated mutant are presented.

Gamma-irradiated crystalline ribonuclease has been chromatographically separated into three distinct components classified as denatured, aggregated, and native, on the basis of the physical alterations which have occurred in the conformation of these molecules. Chromatography on G-75 Sephadex failed to reveal the presence of any low-molecular-weight products in the irradiated enzyme. However, after reduction of the disulfide bridges in mercaptoethanol and chromatography on G-75 Sephadex, low-molecular-weight products indicative of polypeptide chain breakage were detected from both the denatured and the aggregated products. The presence of fragments of the reduced protein was confirmed by velocity sedimentation measurements. Fragments were not observed from the "native" component, which was known to contain partially damaged molecules. These results indicated that polypeptide chain breakage occurred on exposure of RNase to ionizing radiation but that it was masked by the presence of disulfide bonds. Evidence was obtained which suggested that free amide groups were produced as a result of the breakage of the polypeptide chain. A high yield of fragments was observed following irradiation in H 2 S and in vacuo. The inability of H 2 S to prevent the main-chain breakage suggested that this cleavage occurred at an early step in the radiation process. It was concluded that the rupture of the polypeptide chain was an important factor, but not the exclusive cause of the inactivation of crystalline ribonuclease by ionizing radiation.

The production of denatured and aggregated products from crystalline RNase was studied following irradiation by 60 Co γ-rays in vacuo, in oxygen and H 2 S, and in the presence of the paramagnetic ion, Cu ++ . The yields of these products were determined by gel filtration of the irradiated enzyme on G-75 Sephadex. The results showed that aggregation occurred through relatively slow free radical reactions which were eliminated or altered by free radical scavengers such as O 2 and H 2 S. Conversely, the yield of denatured products was enhanced from 1.2/100 eV (in vacuo) to 2.4/100 eV by these gases. The yield of both aggregated and denatured molecules decreased in the presence of Cu ++ ions, suggesting that these ions reduce the effective radiation dose. Detailed reaction schemes were proposed to account for the rates of formation of the products, and the yields deduced from these schemes were in reasonable agreement with the yields for inactivation of the enzyme under the same conditions. Two peaks, which eluted in the region of the aggregates, were identified as dimers and trimers of RNase. The trimer was observed even at low radiation doses.

Dry lyophilized RNase, irradiated with γ-rays, has been chromatographed by gel filtration on Sephadex to yield products characterized as "aggregated," "denatured," or "native" from their elution patterns, their sedimentation velocities their relative migration rates during starch gel electrophoresis, and their characteristic UV absorption spectra. The "aggregated" and "denatured" fractions possessed significant enzymatic activity. On reduction and reoxidation of the disulfide bridges, most of the activity of these damaged fractions disappeared. The failure of the "aggregated" and "denatured" molecules to regain enzymatic activity after reduction and reoxidation indicated that disulfide interchange and hydrogen and hydrophobic bond breakage, per se, could not have been the only damage suffered by the molecules. The "native" fraction was indistinguishable from unirradiated RNase as judged by the above physical tests. However, an appreciable fraction of the molecules from the "native" peak were enzymatically inactive following irradiation. After reduction and reoxidation of these molecules, it was found that the specific enzymatic activity returned to that of unirradiated RNase for samples irradiated in vacuo. In contrast to this, an increased level of inactivation was observed following reduction and reoxidation of the "native" fraction produced on irradiation in oxygen.