Problems of extrapolation from high dose to low dose in Tradescantia mutation studies.

Inflorescences of several clones of Tradescantia heterozygous for flower color have been treated with ionizing radiation and with the gaseous form of several known or suspected chemical mutagens. Pink somatic mutations were subsequently scored in the stamen hair cells of mature flowers and exposure/dose response curves were constructed. Results indicate clearly that there is no threshold for mutation response following x or neutron irradiation. Results so far obtained for gaseous chemicals are less clear, but also suggest that there is no threshold for mutation response. Following exposure to 0.43-MeV neutrons, the mutation response is directly proportional to dose, forming a + I slope on a log-log plot, from 10 mrads to the area of mutation saturation beginning at ca 10 rads. The x-ray mutation response curve (determined at a dose rate of 30 rads/min) is more complex, forming a + I slope between doses of 0.25 and 5 rads; a slope of + 1.4 may be fitted to the data for doses from 5 to 100 rads. Clearly, an extrapolation from the high-dose portion of the curve to lower doses would have increasingly underestimated the mutagenic potential of doses lower than 5 rads. In addition, a threshold effect would have been predicted. Since neutron and x-ray curves are converging above 5 rads, RBE decreases as dose increases. However, since the curves are parallel below 5 rads, RBE is constant and maximum in that region. Extrapolation of high-dose x-ray data to lower doses would have predicted an ever increasing RBE. Departure of the x-ray dose-response curve