Radiation-Induced DNA Damage in Tumors and Normal Tissues: I. Feasibility of Estimating the Hypoxic Fraction

It is well known that the type and quantity of DNA damage produced by ionizing radiation depend on the oxygen concentration around the DNA. For example, in irradiated mammalian cells, both a decrease in the DNA strand break efficiency and the induction of DNA-protein crosslinks (DPCs) occur as the extracellular oxygen concentration is decreased below 1%. In the study reported here, the feasibility of estimating the hypoxic fraction of irradiated tumors and normal tissues was investigated by measuring the single-strand scission factor, the DNA-protein crosslink factor, and the amount of DNA remaining on polycarbonate filters after elution with ∼24 ml of tetrapropylammonium hydroxide at pH 12.3 without proteinase K (PK) in the lysis solution. In anesthetized air-breathing Fisher 344 rats, no radiation-induced DPCs were detected in either cerebellar neurons or cells of subcutaneous (sc) 9L tumors when the DNA was assayed at approximately one half-time of repair after doses ≤15 Gy. Within 10 min after anesthetized rats were killed, the maximum decrease in the radiation-induced strand break efficiency and the maximum formation of radiation-induced DPCs occurred in both cerebellar neurons and sc 9L tumors. When irradiated cerebellar neurons or sc 9L tumor cells from air-breathing and dead rats were mixed to simulate hypoxic fractions of 0, 10, 25, 50, 75, and 100%, only the percentage of the DNA retained on the filter after ∼24 ml of elution without PK in the lysis solution was a linear function of the simulated hypoxic fraction after doses of both 15 and 2 Gy. At 15 Gy, the linear function was identical for 9L cells in tissue culture, sc 9L tumor cells, and cerebellar neurons. In addition, the slope, but not the intercept, of the linear function appeared to be independent of dose from 2 to 15 Gy. Consequently, if the dose and the amount of strand break repair are kept relatively constant, the linear function appears to depend primarily on radiation chemistry events, rather than the biological properties of the irradiated cells. Moreover, the data suggest that this assay can measure a hypoxic fraction of ≤10% after a conventional radiotherapy dose of 2 Gy, provided sufficient material is available for analysis. Advantages of this assay are that it (1) measures the intracellular rather than the extracellular O₂ concentration, (2) does not require labeling the DNA with radioisotopes, (3) requires only 1-2 × 10⁷ nuclei (intact cells are not required), (4) can be performed on previously frozen samples, and (5) requires only 24-48 h to obtain an answer. The major disadvantage of the assay is that it requires biopsies which are representative of the whole tumor and must be taken rapidly after irradiation. We believe the advantages of this assay for determining the presence and quantity of hypoxic cells in situ outweigh the disadvantages in most laboratory situations, but its application in the clinic will require further research.