Radiosensitivity of T47D cells.

Under the assumption that only clonogenic cells are radiosensitive, the survival probability of cell populations is given by (see Materials and Methods, Eq 2) P = 1 − P0 = 1 − e−S(D)εμ, where S(D) is the fraction of clonogenic cells surviving a dose D of radiation, ε is the fraction of clonogenic cells in the population and μ is the mean number of cells in the sample. a: estimate of S(D) using standard clonogenic assays. Data were fitted with the linear-quadratic model (Eq 1, see Materials and methods): α = 0.389±0.13 Gy-1, β = 0.047±0.01 Gy-2, α/β = 8.22 Gy (χ2/df = 4.97). Thus S(8 Gy)~0.0022. b: estimate of ε in limiting dilution assays with non irradiated cells. Cells were seeded either in F-bottom (blue circles and line) or V-bottom (red squares and line) wells at the indicated mean number μ of cells/well and the fraction of wells showing no proliferation F0 was determined. Data were fitted with Eq 3 (see Materials and methods): ε = 0.228±0.0097 (r2 = 0.99, cells in F-bottom wells), ε = 0.212±0.0081 (r2 = 0.99, cells in V-bottom wells). c: survival probability P of independent cells populations grown in F-bottom (blue circles and lines) or V-bottom (red squares and lines) wells starting from a mean initial number of cells/well μ and irradiated with a dose of 8 Gy. Dashed lines: model predictions with Eq 2 and experimentally determined parameters. In this case the model can be exploited as the null hypothesis for statistical testing. Data from cell populations grown in F-bottom wells did not significantly deviate from model predictions (χ2 = 6.69, p = 0.35, goodness-of-fit test), whereas data obtained with cell populations grown in V-bottom wells did not agree with model predictions and the null hypothesis could be rejected (χ2 = 489.4, p~10−102, goodness-of-fit test). Continuous lines: nonlinear fits of experimental data with Eq 2. and with only S(D) as a free parameter. The estimated surviving fraction was ~4.7 fold higher for cell populations grown in V-bottom wells (S(D) = 0.011±0.0012) than for cells grown in F-bottom wells (S(D) = 0.00236±0.0002). Overall the data show that the sensitivity of cell populations to radiation treatment does not only depend on the number of clonogenic cells in the population, and that tumor cell populations grown under experimental conditions that favour cell contacts are more radioresistant.